Lortie, CJ (2022): Experiment sandbox. figshare. Book. https://doi.org/10.6084/m9.figshare.20442801.v3
Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.
Learning objectives.
Experiments are an excellent data collection strategy for social workers wishing to observe the effects of a clinical intervention or social welfare program. Understanding what experiments are and how they are conducted is useful for all social scientists, whether they actually plan to use this methodology or simply aim to understand findings from experimental studies. An experiment is a method of data collection designed to test hypotheses under controlled conditions. In social scientific research, the term experiment has a precise meaning and should not be used to describe all research methodologies.
Experiments have a long and important history in social science. Behaviorists such as John Watson, B. F. Skinner, Ivan Pavlov, and Albert Bandura used experimental design to demonstrate the various types of conditioning. Using strictly controlled environments, behaviorists were able to isolate a single stimulus as the cause of measurable differences in behavior or physiological responses. The foundations of social learning theory and behavior modification are found in experimental research projects. Moreover, behaviorist experiments brought psychology and social science away from the abstract world of Freudian analysis and towards empirical inquiry, grounded in real-world observations and objectively-defined variables. Experiments are used at all levels of social work inquiry, including agency-based experiments that test therapeutic interventions and policy experiments that test new programs.
Several kinds of experimental designs exist. In general, designs considered to be true experiments contain three basic key features:
Some true experiments are more complex. Their designs can also include a pre-test and can have more than two groups, but these are the minimum requirements for a design to be a true experiment.
In a true experiment, the effect of an intervention is tested by comparing two groups: one that is exposed to the intervention (the experimental group , also known as the treatment group) and another that does not receive the intervention (the control group ). Importantly, participants in a true experiment need to be randomly assigned to either the control or experimental groups. Random assignment uses a random number generator or some other random process to assign people into experimental and control groups. Random assignment is important in experimental research because it helps to ensure that the experimental group and control group are comparable and that any differences between the experimental and control groups are due to random chance. We will address more of the logic behind random assignment in the next section.
In an experiment, the independent variable is receiving the intervention being tested—for example, a therapeutic technique, prevention program, or access to some service or support. It is less common in of social work research, but social science research may also have a stimulus, rather than an intervention as the independent variable. For example, an electric shock or a reading about death might be used as a stimulus to provoke a response.
In some cases, it may be immoral to withhold treatment completely from a control group within an experiment. If you recruited two groups of people with severe addiction and only provided treatment to one group, the other group would likely suffer. For these cases, researchers use a control group that receives “treatment as usual.” Experimenters must clearly define what treatment as usual means. For example, a standard treatment in substance abuse recovery is attending Alcoholics Anonymous or Narcotics Anonymous meetings. A substance abuse researcher conducting an experiment may use twelve-step programs in their control group and use their experimental intervention in the experimental group. The results would show whether the experimental intervention worked better than normal treatment, which is useful information.
The dependent variable is usually the intended effect the researcher wants the intervention to have. If the researcher is testing a new therapy for individuals with binge eating disorder, their dependent variable may be the number of binge eating episodes a participant reports. The researcher likely expects her intervention to decrease the number of binge eating episodes reported by participants. Thus, she must, at a minimum, measure the number of episodes that occur after the intervention, which is the post-test . In a classic experimental design, participants are also given a pretest to measure the dependent variable before the experimental treatment begins.
Let’s put these concepts in chronological order so we can better understand how an experiment runs from start to finish. Once you’ve collected your sample, you’ll need to randomly assign your participants to the experimental group and control group. In a common type of experimental design, you will then give both groups your pretest, which measures your dependent variable, to see what your participants are like before you start your intervention. Next, you will provide your intervention, or independent variable, to your experimental group, but not to your control group. Many interventions last a few weeks or months to complete, particularly therapeutic treatments. Finally, you will administer your post-test to both groups to observe any changes in your dependent variable. What we’ve just described is known as the classical experimental design and is the simplest type of true experimental design. All of the designs we review in this section are variations on this approach. Figure 8.1 visually represents these steps.
An interesting example of experimental research can be found in Shannon K. McCoy and Brenda Major’s (2003) study of people’s perceptions of prejudice. In one portion of this multifaceted study, all participants were given a pretest to assess their levels of depression. No significant differences in depression were found between the experimental and control groups during the pretest. Participants in the experimental group were then asked to read an article suggesting that prejudice against their own racial group is severe and pervasive, while participants in the control group were asked to read an article suggesting that prejudice against a racial group other than their own is severe and pervasive. Clearly, these were not meant to be interventions or treatments to help depression, but were stimuli designed to elicit changes in people’s depression levels. Upon measuring depression scores during the post-test period, the researchers discovered that those who had received the experimental stimulus (the article citing prejudice against their same racial group) reported greater depression than those in the control group. This is just one of many examples of social scientific experimental research.
In addition to classic experimental design, there are two other ways of designing experiments that are considered to fall within the purview of “true” experiments (Babbie, 2010; Campbell & Stanley, 1963). The posttest-only control group design is almost the same as classic experimental design, except it does not use a pretest. Researchers who use posttest-only designs want to eliminate testing effects , in which participants’ scores on a measure change because they have already been exposed to it. If you took multiple SAT or ACT practice exams before you took the real one you sent to colleges, you’ve taken advantage of testing effects to get a better score. Considering the previous example on racism and depression, participants who are given a pretest about depression before being exposed to the stimulus would likely assume that the intervention is designed to address depression. That knowledge could cause them to answer differently on the post-test than they otherwise would. In theory, as long as the control and experimental groups have been determined randomly and are therefore comparable, no pretest is needed. However, most researchers prefer to use pretests in case randomization did not result in equivalent groups and to help assess change over time within both the experimental and control groups.
Researchers wishing to account for testing effects but also gather pretest data can use a Solomon four-group design. In the Solomon four-group design , the researcher uses four groups. Two groups are treated as they would be in a classic experiment—pretest, experimental group intervention, and post-test. The other two groups do not receive the pretest, though one receives the intervention. All groups are given the post-test. Table 8.1 illustrates the features of each of the four groups in the Solomon four-group design. By having one set of experimental and control groups that complete the pretest (Groups 1 and 2) and another set that does not complete the pretest (Groups 3 and 4), researchers using the Solomon four-group design can account for testing effects in their analysis.
Group 1 | X | X | X |
Group 2 | X | X | |
Group 3 | X | X | |
Group 4 | X |
Solomon four-group designs are challenging to implement in the real world because they are time- and resource-intensive. Researchers must recruit enough participants to create four groups and implement interventions in two of them.
Overall, true experimental designs are sometimes difficult to implement in a real-world practice environment. It may be impossible to withhold treatment from a control group or randomly assign participants in a study. In these cases, pre-experimental and quasi-experimental designs–which we will discuss in the next section–can be used. However, the differences in rigor from true experimental designs leave their conclusions more open to critique.
You can imagine that social work researchers may be limited in their ability to use random assignment when examining the effects of governmental policy on individuals. For example, it is unlikely that a researcher could randomly assign some states to implement decriminalization of recreational marijuana and some states not to in order to assess the effects of the policy change. There are, however, important examples of policy experiments that use random assignment, including the Oregon Medicaid experiment. In the Oregon Medicaid experiment, the wait list for Oregon was so long, state officials conducted a lottery to see who from the wait list would receive Medicaid (Baicker et al., 2013). Researchers used the lottery as a natural experiment that included random assignment. People selected to be a part of Medicaid were the experimental group and those on the wait list were in the control group. There are some practical complications macro-level experiments, just as with other experiments. For example, the ethical concern with using people on a wait list as a control group exists in macro-level research just as it does in micro-level research.
exam scientific experiment by mohamed_hassan CC-0
Foundations of Social Work Research Copyright © 2020 by Rebecca L. Mauldin is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License , except where otherwise noted.
Click through the PLOS taxonomy to find articles in your field.
For more information about PLOS Subject Areas, click here .
Loading metrics
Open Access
Peer-reviewed
Research Article
Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing – original draft, Writing – review & editing
* E-mail: [email protected]
Affiliation Department of Ecology and Evolutionary Biology, Cornell University, Ithaca, NY, United States of America
Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – review & editing
Affiliation Laboratory of Atomic and Solid State Physics, Cornell University, Ithaca, NY, United States of America
Roles Conceptualization, Investigation, Methodology, Writing – review & editing
Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing
Roles Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – review & editing
Critical thinking is the process by which people make decisions about what to trust and what to do. Many undergraduate courses, such as those in biology and physics, include critical thinking as an important learning goal. Assessing critical thinking, however, is non-trivial, with mixed recommendations for how to assess critical thinking as part of instruction. Here we evaluate the efficacy of assessment questions to probe students’ critical thinking skills in the context of biology and physics. We use two research-based standardized critical thinking instruments known as the Biology Lab Inventory of Critical Thinking in Ecology (Eco-BLIC) and Physics Lab Inventory of Critical Thinking (PLIC). These instruments provide experimental scenarios and pose questions asking students to evaluate what to trust and what to do regarding the quality of experimental designs and data. Using more than 3000 student responses from over 20 institutions, we sought to understand what features of the assessment questions elicit student critical thinking. Specifically, we investigated (a) how students critically evaluate aspects of research studies in biology and physics when they are individually evaluating one study at a time versus comparing and contrasting two and (b) whether individual evaluation questions are needed to encourage students to engage in critical thinking when comparing and contrasting. We found that students are more critical when making comparisons between two studies than when evaluating each study individually. Also, compare-and-contrast questions are sufficient for eliciting critical thinking, with students providing similar answers regardless of if the individual evaluation questions are included. This research offers new insight on the types of assessment questions that elicit critical thinking at the introductory undergraduate level; specifically, we recommend instructors incorporate more compare-and-contrast questions related to experimental design in their courses and assessments.
Citation: Heim AB, Walsh C, Esparza D, Smith MK, Holmes NG (2022) What influences students’ abilities to critically evaluate scientific investigations? PLoS ONE 17(8): e0273337. https://doi.org/10.1371/journal.pone.0273337
Editor: Dragan Pamucar, University of Belgrade Faculty of Organisational Sciences: Univerzitet u Beogradu Fakultet organizacionih nauka, SERBIA
Received: December 3, 2021; Accepted: August 6, 2022; Published: August 30, 2022
Copyright: © 2022 Heim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All raw data files are available from the Cornell Institute for Social and Economic Research (CISER) data and reproduction archive ( https://archive.ciser.cornell.edu/studies/2881 ).
Funding: This work was supported by the National Science Foundation under grants DUE-1909602 (MS & NH) and DUE-1611482 (NH). NSF: nsf.gov The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Critical thinking and its importance.
Critical thinking, defined here as “the ways in which one uses data and evidence to make decisions about what to trust and what to do” [ 1 ], is a foundational learning goal for almost any undergraduate course and can be integrated in many points in the undergraduate curriculum. Beyond the classroom, critical thinking skills are important so that students are able to effectively evaluate data presented to them in a society where information is so readily accessible [ 2 , 3 ]. Furthermore, critical thinking is consistently ranked as one of the most necessary outcomes of post-secondary education for career advancement by employers [ 4 ]. In the workplace, those with critical thinking skills are more competitive because employers assume they can make evidence-based decisions based on multiple perspectives, keep an open mind, and acknowledge personal limitations [ 5 , 6 ]. Despite the importance of critical thinking skills, there are mixed recommendations on how to elicit and assess critical thinking during and as a result of instruction. In response, here we evaluate the degree to which different critical thinking questions elicit students’ critical thinking skills.
Across STEM (i.e., science, technology, engineering, and mathematics) disciplines, several standardized assessments probe critical thinking skills. These assessments focus on aspects of critical thinking and ask students to evaluate experimental methods [ 7 – 11 ], form hypotheses and make predictions [ 12 , 13 ], evaluate data [ 2 , 12 – 14 ], or draw conclusions based on a scenario or figure [ 2 , 12 – 14 ]. Many of these assessments are open-response, so they can be difficult to score, and several are not freely available.
In addition, there is an ongoing debate regarding whether critical thinking is a domain-general or context-specific skill. That is, can someone transfer their critical thinking skills from one domain or context to another (domain-general) or do their critical thinking skills only apply in their domain or context of expertise (context-specific)? Research on the effectiveness of teaching critical thinking has found mixed results, primarily due to a lack of consensus definition of and assessment tools for critical thinking [ 15 , 16 ]. Some argue that critical thinking is domain-general—or what Ennis refers to as the “general approach”—because it is an overlapping skill that people use in various aspects of their lives [ 17 ]. In contrast, others argue that critical thinking must be elicited in a context-specific domain, as prior knowledge is needed to make informed decisions in one’s discipline [ 18 , 19 ]. Current assessments include domain-general components [ 2 , 7 , 8 , 14 , 20 , 21 ], asking students to evaluate, for instance, experiments on the effectiveness of dietary supplements in athletes [ 20 ] and context-specific components, such as to measure students’ abilities to think critically in domains such as neuroscience [ 9 ] and biology [ 10 ].
Others maintain the view that critical thinking is a context-specific skill for the purpose of undergraduate education, but argue that it should be content accessible [ 22 – 24 ], as “thought processes are intertwined with what is being thought about” [ 23 ]. From this viewpoint, the context of the assessment would need to be embedded in a relatively accessible context to assess critical thinking independent of students’ content knowledge. Thus, to effectively elicit critical thinking among students, instructors should use assessments that present students with accessible domain-specific information needed to think deeply about the questions being asked [ 24 , 25 ].
Within the context of STEM, current critical thinking assessments primarily ask students to evaluate a single experimental scenario (e.g., [ 10 , 20 ]), though compare-and-contrast questions about more than one scenario can be a powerful way to elicit critical thinking [ 26 , 27 ]. Generally included in the “Analysis” level of Bloom’s taxonomy [ 28 – 30 ], compare-and-contrast questions encourage students to recognize, distinguish between, and relate features between scenarios and discern relevant patterns or trends, rather than compile lists of important features [ 26 ]. For example, a compare-and-contrast assessment may ask students to compare the hypotheses and research methods used in two different experimental scenarios, instead of having them evaluate the research methods of a single experiment. Alternatively, students may inherently recall and use experimental scenarios based on their prior experiences and knowledge as they evaluate an individual scenario. In addition, evaluating a single experimental scenario individually may act as metacognitive scaffolding [ 31 , 32 ]—a process which “guides students by asking questions about the task or suggesting relevant domain-independent strategies [ 32 ]—to support students in their compare-and-contrast thinking.
Our primary objective of this study was to better understand what features of assessment questions elicit student critical thinking using two existing instruments in STEM: the Biology Lab Inventory of Critical Thinking in Ecology (Eco-BLIC) and Physics Lab Inventory of Critical Thinking (PLIC). We focused on biology and physics since critical thinking assessments were already available for these disciplines. Specifically, we investigated (a) how students critically evaluate aspects of research studies in biology and physics when they are individually evaluating one study at a time or comparing and contrasting two studies and (b) whether individual evaluation questions are needed to encourage students to engage in critical thinking when comparing and contrasting.
Providing undergraduates with ample opportunities to practice critical thinking skills in the classroom is necessary for evidence-based critical thinking in their future careers and everyday life. While most critical thinking instruments in biology and physics contexts have undergone some form of validation to ensure they are accurately measuring the intended construct, to our knowledge none have explored how different question types influence students’ critical thinking. This research offers new insight on the types of questions that elicit critical thinking, which can further be applied by educators and researchers across disciplines to measure cognitive student outcomes and incorporate more effective critical thinking opportunities in the classroom.
The procedures for this study were approved by the Institutional Review Board of Cornell University (Eco-BLIC: #1904008779; PLIC: #1608006532). Informed consent was obtained by all participating students via online consent forms at the beginning of the study, and students did not receive compensation for participating in this study unless their instructor offered credit for completing the assessment.
We administered the Eco-BLIC to undergraduate students across 26 courses at 11 institutions (six doctoral-granting, three Master’s-granting, and two Baccalaureate-granting) in Fall 2020 and Spring 2021 and received 1612 usable responses. Additionally, we administered the PLIC to undergraduate students across 21 courses at 11 institutions (six doctoral-granting, one Master’s-granting, three four-year colleges, and one 2-year college) in Fall 2020 and Spring 2021 and received 1839 usable responses. We recruited participants via convenience sampling by emailing instructors of primarily introductory ecology-focused courses or introductory physics courses who expressed potential interest in implementing our instrument in their course(s). Both instruments were administered online via Qualtrics and students were allowed to complete the assessments outside of class. The demographic distribution of the response data is presented in Table 1 , all of which were self-reported by students. The values presented in this table represent all responses we received.
https://doi.org/10.1371/journal.pone.0273337.t001
Question types..
Though the content and concepts featured in the Eco-BLIC and PLIC are distinct, both instruments share a similar structure and set of question types. The Eco-BLIC—which was developed using a structure similar to that of the PLIC [ 1 ]—includes two predator-prey scenarios based on relationships between (a) smallmouth bass and mayflies and (b) great-horned owls and house mice. Within each scenario, students are presented with a field-based study and a laboratory-based study focused on a common research question about feeding behaviors of smallmouth bass or house mice, respectively. The prompts for these two Eco-BLIC scenarios are available in S1 and S2 Appendices. The PLIC focuses on two research groups conducting different experiments to test the relationship between oscillation periods of masses hanging on springs [ 1 ]; the prompts for this scenario can be found in S3 Appendix . The descriptive prompts in both the Eco-BLIC and PLIC also include a figure presenting data collected by each research group, from which students are expected to draw conclusions. The research scenarios (e.g., field-based group and lab-based group on the Eco-BLIC) are written so that each group has both strengths and weaknesses in their experimental designs.
After reading the prompt for the first experimental group (Group 1) in each instrument, students are asked to identify possible claims from Group 1’s data (data evaluation questions). Students next evaluate the strengths and weaknesses of various study features for Group 1 (individual evaluation questions). Examples of these individual evaluation questions are in Table 2 . They then suggest next steps the group should pursue (next steps items). Students are then asked to read about the prompt describing the second experimental group’s study (Group 2) and again answer questions about the possible claims, strengths and weaknesses, and next steps of Group 2’s study (data evaluation questions, individual evaluation questions, and next steps items). Once students have independently evaluated Groups 1 and 2, they answer a series of questions to compare the study approaches of Group 1 versus Group 2 (group comparison items). In this study, we focus our analysis on the individual evaluation questions and group comparison items.
https://doi.org/10.1371/journal.pone.0273337.t002
To determine whether the individual evaluation questions impacted the assessment of students’ critical thinking, students were randomly assigned to take one of two versions of the assessment via Qualtrics branch logic: 1) a version that included the individual evaluation and group comparison items or 2) a version with only the group comparison items, with the individual evaluation questions removed. We calculated the median time it took students to answer each of these versions for both the Eco-BLIC and PLIC.
We also conducted one-on-one think-aloud interviews with students to elicit feedback on the assessment questions (Eco-BLIC n = 21; PLIC n = 4). Students were recruited via convenience sampling at our home institution and were primarily majoring in biology or physics. All interviews were audio-recorded and screen captured via Zoom and lasted approximately 30–60 minutes. We asked participants to discuss their reasoning for answering each question as they progressed through the instrument. We did not analyze these interviews in detail, but rather used them to extract relevant examples of critical thinking that helped to explain our quantitative findings. Multiple think-aloud interviews were conducted with students using previous versions of the PLIC [ 1 ], though these data are not discussed here.
Our analyses focused on (1) investigating the alignment between students’ responses to the individual evaluation questions and the group comparison items and (2) comparing student responses between the two instrument versions. If individual evaluation and group comparison items elicit critical thinking in the same way, we would expect to see the same frequency of responses for each question type, as per Fig 1 . For example, if students evaluated one study feature of Group 1 as a strength and the same study feature for Group 2 as a strength, we would expect that students would respond that both groups were highly effective for this study feature on the group comparison item (i.e., data represented by the purple circle in the top right quadrant of Fig 1 ). Alternatively, if students evaluated one study feature of Group 1 as a strength and the same study feature for Group 2 as a weakness, we would expect that students would indicate that Group 1 was more effective than Group 2 on the group comparison item (i.e., data represented by the green circle in the lower right quadrant of Fig 1 ).
The x- and y-axes represent rankings on the individual evaluation questions for Groups 1 and 2 (or field and lab groups), respectively. The colors in the legend at the top of the figure denote responses to the group comparison items. In this idealized example, all pie charts are the same size to indicate that the student answers are equally proportioned across all answer combinations.
https://doi.org/10.1371/journal.pone.0273337.g001
We ran descriptive statistics to summarize student responses to questions and examine distributions and frequencies of the data on the Eco-BLIC and PLIC. We also conducted chi-square goodness-of-fit tests to analyze differences in student responses between versions within the relevant questions from the same instrument. In all of these tests, we used a Bonferroni correction to lower the chances of receiving a false positive and account for multiple comparisons. We generated figures—primarily multi-pie chart graphs and heat maps—to visualize differences between individual evaluation and group comparison items and between versions of each instrument with and without individual evaluation questions, respectively. All aforementioned data analyses and figures were conducted or generated in the R statistical computing environment (v. 4.1.1) and Microsoft Excel.
We asked students to evaluate different experimental set-ups on the Eco-BLIC and PLIC two ways. Students first evaluated the strengths and weaknesses of study features for each scenario individually (individual evaluation questions, Table 2 ) and, subsequently, answered a series of questions to compare and contrast the study approaches of both research groups side-by-side (group comparison items, Table 2 ). Through analyzing the individual evaluation questions, we found that students generally ranked experimental features (i.e., those related to study set-up, data collection and summary methods, and analysis and outcomes) of the independent research groups as strengths ( Fig 2 ), evidenced by the mean scores greater than 2 on a scale from 1 (weakness) to 4 (strength).
Each box represents the interquartile range (IQR). Lines within each box represent the median. Circles represent outliers of mean scores for each question.
https://doi.org/10.1371/journal.pone.0273337.g002
Our results indicate that when students consider Group 1 or Group 2 individually, they mark most study features as strengths (consistent with the means in Fig 2 ), shown by the large circles in the upper right quadrant across the three experimental scenarios ( Fig 3 ). However, the proportion of colors on each pie chart shows that students select a range of responses when comparing the two groups [e.g., Group 1 being more effective (green), Group 2 being more effective (blue), both groups being effective (purple), and neither group being effective (orange)]. We infer that students were more discerning (i.e., more selective) when they were asked to compare the two groups across the various study features ( Fig 3 ). In short, students think about the groups differently if they are rating either Group 1 or Group 2 in the individual evaluation questions versus directly comparing Group 1 to Group 2.
The x- and y-axes represent students’ rankings on the individual evaluation questions for Groups 1 and 2 on each assessment, respectively, where 1 indicates weakness and 4 indicates strength. The overall size of each pie chart represents the proportion of students who responded with each pair of ratings. The colors in the pie charts denote the proportion of students’ responses who chose each option on the group comparison items. (A) Eco-BLIC bass-mayfly scenario (B) Eco-BLIC owl-mouse scenario (C) PLIC oscillation periods of masses hanging on springs scenario.
https://doi.org/10.1371/journal.pone.0273337.g003
These results are further supported by student responses from the think-aloud interviews. For example, one interview participant responding to the bass-mayfly scenario of the Eco-BLIC explained that accounting for bias/error in both the field and lab groups in this scenario was a strength (i.e., 4). This participant mentioned that Group 1, who performed the experiment in the field, “[had] outliers, so they must have done pretty well,” and that Group 2, who collected organisms in the field but studied them in lab, “did a good job of accounting for bias.” However, when asked to compare between the groups, this student argued that Group 2 was more effective at accounting for bias/error, noting that “they controlled for more variables.”
Another individual who was evaluating “repeated trials for each mass” in the PLIC expressed a similar pattern. In response to ranking this feature of Group 1 as a strength, they explained: “Given their uncertainties and how small they are, [the group] seems like they’ve covered their bases pretty well.” Similarly, they evaluated this feature of Group 2 as a strength as well, simply noting: “Same as the last [group], I think it’s a strength.” However, when asked to compare between Groups 1 and 2, this individual argued that Group 1 was more effective because they conducted more trials.
Given that students were more discerning when they directly compared two groups for both biology and physics experimental scenarios, we next sought to determine if the individual evaluation questions for Group 1 or Group 2 were necessary to elicit or helpful to support student critical thinking about the investigations. To test this, students were randomly assigned to one of two versions of the instrument. Students in one version saw individual evaluation questions about Group 1 and Group 2 and then saw group comparison items for Group 1 versus Group 2. Students in the second version only saw the group comparison items. We found that students assigned to both versions responded similarly to the group comparison questions, indicating that the individual evaluation questions did not promote additional critical thinking. We visually represent these similarities across versions with and without the individual evaluation questions in Fig 4 as heat maps.
The x-axis denotes students’ responses on the group comparison items (i.e., whether they ranked Group 1 as more effective, Group 2 as more effective, both groups as highly effective, or neither group as effective/both groups were minimally effective). The y-axis lists each of the study features that students compared between the field and lab groups. White and lighter shades of red indicate a lower percentage of student responses, while brighter red indicates a higher percentage of student responses. (A) Eco-BLIC bass-mayfly scenario. (B) Eco-BLIC owl-mouse scenario. (C) PLIC oscillation periods of masses hanging on springs scenario.
https://doi.org/10.1371/journal.pone.0273337.g004
We ran chi-square goodness-of-fit tests on the answers between student responses on both instrument versions and there were no significant differences on the Eco-BLIC bass-mayfly scenario ( Fig 4A ; based on an adjusted p -value of 0.006) or owl-mouse questions ( Fig 4B ; based on an adjusted p-value of 0.004). There were only three significant differences (out of 53 items) in how students responded to questions on both versions of the PLIC ( Fig 4C ; based on an adjusted p -value of 0.0005). The items that students responded to differently ( p <0.0005) across both versions were items where the two groups were identical in their design; namely, the equipment used (i.e., stopwatches), the variables measured (i.e., time and mass), and the number of bounces of the spring per trial (i.e., five bounces). We calculated Cramer’s C (Vc; [ 33 ]), a measure commonly applied to Chi-square goodness of fit models to understand the magnitude of significant results. We found that the effect sizes for these three items were small (Vc = 0.11, Vc = 0.10, Vc = 0.06, respectively).
The trend that students answer the Group 1 versus Group 2 comparison questions similarly, regardless of whether they responded to the individual evaluation questions, is further supported by student responses from the think-aloud interviews. For example, one participant who did not see the individual evaluation questions for the owl-mouse scenario of the Eco-BLIC independently explained that sampling mice from other fields was a strength for both the lab and field groups. They explained that for the lab group, “I think that [the mice] coming from multiple nearby fields is good…I was curious if [mouse] behavior was universal.” For the field group, they reasoned, “I also noticed it was just from a single nearby field…I thought that was good for control.” However, this individual ultimately reasoned that the field group was “more effective for sampling methods…it’s better to have them from a single field because you know they were exposed to similar environments.” Thus, even without individual evaluation questions available, students can still make individual evaluations when comparing and contrasting between groups.
We also determined that removing the individual evaluation questions decreased the duration of time students needed to complete the Eco-BLIC and PLIC. On the Eco-BLIC, the median time to completion for the version with individual evaluation and group comparison questions was approximately 30 minutes, while the version with only the group comparisons had a median time to completion of 18 minutes. On the PLIC, the median time to completion for the version with individual evaluation questions and group comparison questions was approximately 17 minutes, while the version with only the group comparisons had a median time to completion of 15 minutes.
To determine how to elicit critical thinking in a streamlined manner using introductory biology and physics material, we investigated (a) how students critically evaluate aspects of experimental investigations in biology and physics when they are individually evaluating one study at a time versus comparing and contrasting two and (b) whether individual evaluation questions are needed to encourage students to engage in critical thinking when comparing and contrasting.
We found that students were more discerning when comparing between the two groups in the Eco-BLIC and PLIC rather than when evaluating each group individually. While students tended to independently evaluate study features of each group as strengths ( Fig 2 ), there was greater variation in their responses to which group was more effective when directly comparing between the two groups ( Fig 3 ). Literature evaluating the role of contrasting cases provides plausible explanations for our results. In that work, contrasting between two cases supports students in identifying deep features of the cases, compared with evaluating one case after the other [ 34 – 37 ]. When presented with a single example, students may deem certain study features as unimportant or irrelevant, but comparing study features side-by-side allows students to recognize the distinct features of each case [ 38 ]. We infer, therefore, that students were better able to recognize the strengths and weaknesses of the two groups in each of the assessment scenarios when evaluating the groups side by side, rather than in isolation [ 39 , 40 ]. This result is somewhat surprising, however, as students could have used their knowledge of experimental designs as a contrasting case when evaluating each group. Future work, therefore, should evaluate whether experts use their vast knowledge base of experimental studies as discerning contrasts when evaluating each group individually. This work would help determine whether our results here suggest that students do not have a sufficient experiment-base to use as contrasts or if the students just do not use their experiment-base when evaluating the individual groups. Regardless, our study suggests that critical thinking assessments should ask students to compare and contrast experimental scenarios, rather than just evaluate individual cases.
We found that individual evaluation questions were unnecessary for eliciting or supporting students’ critical thinking on the two assessments. Students responded to the group comparison items similarly whether or not they had received the individual evaluation questions. The exception to this pattern was that students responded differently to three group comparison items on the PLIC when individual evaluation questions were provided. These three questions constituted a small portion of the PLIC and showed a small effect size. Furthermore, removing the individual evaluation questions decreased the median time for students to complete the Eco-BLIC and PLIC. It is plausible that spending more time thinking about the experimental methods while responding to the individual evaluation questions would then prepare students to be better discerners on the group comparison questions. However, the overall trend is that individual evaluation questions do not have a strong impact on how students evaluate experimental scenarios, nor do they set students up to be better critical thinkers later. This finding aligns with prior research suggesting that students tend to disregard details when they evaluate a single case, rather than comparing and contrasting multiple cases [ 38 ], further supporting our findings about the effectiveness of the group comparison questions.
Individual evaluation questions were not effective for students to engage in critical thinking nor to prepare them for subsequent questions that elicit their critical thinking. Thus, researchers and instructors could make critical thinking assessments more effective and less time-consuming by encouraging comparisons between cases. Additionally, the study raises a question about whether instruction should incorporate more experimental case studies throughout their courses and assessments so that students have a richer experiment-base to use as contrasts when evaluating individual experimental scenarios. To help students discern information about experimental design, we suggest that instructors consider providing them with multiple experimental studies (i.e., cases) and asking them to compare and contrast between these studies.
When designing critical thinking assessments, questions should ask students to make meaningful comparisons that require them to consider the important features of the scenarios. One challenge of relying on compare-and-contrast questions in the Eco-BLIC and PLIC to elicit students’ critical thinking is ensuring that students are comparing similar yet distinct study features across experimental scenarios, and that these comparisons are meaningful [ 38 ]. For example, though sample size is different between experimental scenarios in our instruments, it is a significant feature that has implications for other aspects of the research like statistical analyses and behaviors of the animals. Therefore, one limitation of our study could be that we exclusively focused on experimental method evaluation questions (i.e., what to trust), and we are unsure if the same principles hold for other dimensions of critical thinking (i.e., what to do). Future research should explore whether questions that are not in a compare-and-contrast format also effectively elicit critical thinking, and if so, to what degree.
As our question schema in the Eco-BLIC and PLIC were designed for introductory biology and physics content, it is unknown how effective this question schema would be for upper-division biology and physics undergraduates who we would expect to have more content knowledge and prior experiences for making comparisons in their respective disciplines [ 18 , 41 ]. For example, are compare-and-contrast questions still needed to elicit critical thinking among upper-division students, or would critical thinking in this population be more effectively assessed by incorporating more sophisticated data analyses in the research scenarios? Also, if students with more expert-like thinking have a richer set of experimental scenarios to inherently use as contrasts when comparing, we might expect their responses on the individual evaluation questions and group comparisons to better align. To further examine how accessible and context-specific the Eco-BLIC and PLIC are, novel scenarios could be developed that incorporate topics and concepts more commonly addressed in upper-division courses. Additionally, if instructors offer students more experience comparing and contrasting experimental scenarios in the classroom, would students be more discerning on the individual evaluation questions?
While a single consensus definition of critical thinking does not currently exist [ 15 ], continuing to explore critical thinking in other STEM disciplines beyond biology and physics may offer more insight into the context-specific nature of critical thinking [ 22 , 23 ]. Future studies should investigate critical thinking patterns in other STEM disciplines (e.g., mathematics, engineering, chemistry) through designing assessments that encourage students to evaluate aspects of at least two experimental studies. As undergraduates are often enrolled in multiple courses simultaneously and thus have domain-specific knowledge in STEM, would we observe similar patterns in critical thinking across additional STEM disciplines?
Lastly, we want to emphasize that we cannot infer every aspect of critical thinking from students’ responses on the Eco-BLIC and PLIC. However, we suggest that student responses on the think-aloud interviews provide additional qualitative insight into how and why students were making comparisons in each scenario and their overall critical thinking processes.
Overall, we found that comparing and contrasting two different experiments is an effective and efficient way to elicit context-specific critical thinking in introductory biology and physics undergraduates using the Eco-BLIC and the PLIC. Students are more discerning (i.e., critical) and engage more deeply with the scenarios when making comparisons between two groups. Further, students do not evaluate features of experimental studies differently when individual evaluation questions are provided or removed. These novel findings hold true across both introductory biology and physics, based on student responses on the Eco-BLIC and PLIC, respectively—though there is much more to explore regarding critical thinking processes of students across other STEM disciplines and in more advanced stages of their education. Undergraduate students in STEM need to be able to critically think for career advancement, and the Eco-BLIC and PLIC are two means of measuring students’ critical thinking in biology and physics experimental contexts via comparing and contrasting. This research offers new insight on the types of questions that elicit critical thinking, which can further be applied by educators and researchers across disciplines to teach and measure cognitive student outcomes. Specifically, we recommend instructors incorporate more compare-and-contrast questions related to experimental design in their courses to efficiently elicit undergraduates’ critical thinking.
S1 appendix. eco-blic bass-mayfly scenario prompt..
https://doi.org/10.1371/journal.pone.0273337.s001
https://doi.org/10.1371/journal.pone.0273337.s002
https://doi.org/10.1371/journal.pone.0273337.s003
We thank the members of the Cornell Discipline-based Education Research group for their feedback on this article, as well as our advisory board (Jenny Knight, Meghan Duffy, Luanna Prevost, and James Hewlett) and the AAALab for their ideas and suggestions. We also greatly appreciate the instructors who shared the Eco-BLIC and PLIC in their classes and the students who participated in this study.
Our systems are now restored following recent technical disruption, and we’re working hard to catch up on publishing. We apologise for the inconvenience caused. Find out more: https://www.cambridge.org/universitypress/about-us/news-and-blogs/cambridge-university-press-publishing-update-following-technical-disruption
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings .
Critical Issues
Published online by Cambridge University Press: 05 June 2012
[T]he application of the experimental method to the problem of mind is the great outstanding event in the study of the mind, an event to which no other is comparable.
The author of this quote is Edwin G. Boring (1886–1968), one of the great psychologists of the 20th century and author of A History of Experimental Psychology (1929; the quote comes from p. 659). Contemporary psychologists take “the psychology experiment” as a given, but it is actually a relatively recent cultural invention. Although fascination with human behavior is doubtless as old as the emergence of Homo sapiens , the application of experimental methods to the study of the human mind and behavior is only 150 or so years old. Scientific methods, with heavy reliance on experimental technique, arose in Western civilization during the time of the Renaissance, when great insights and modes of thoughts from the ancient Greek, Roman, and Arab civilizations were rediscovered. The 17th century witnessed the great discoveries of Kepler, Galileo, and Newton in the physical world. Interest in chemistry and biology arose after the early development of physics. Experimental physiology arose as a discipline in the late 1700s and early 1800s. Still, despite great advances in these fields and despite the fact that scientists of the day usually conducted research in many different fields, no one at that time performed experiments studying humans or their mental life. The first physiologists and anatomists mostly contented themselves with the study of corpses.
Save book to kindle.
To save this book to your Kindle, first ensure [email protected] is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle .
Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Find out more about the Kindle Personal Document Service .
To save content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about saving content to Dropbox .
To save content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about saving content to Google Drive .
An official website of the United States government
The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.
The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.
Email citation, add to collections.
Your saved search, create a file for external citation management software, your rss feed.
Affiliations.
Background: Traditionally, doctoral student education in the biomedical sciences relies on didactic coursework to build a foundation of scientific knowledge and an apprenticeship model of training in the laboratory of an established investigator. Recent recommendations for revision of graduate training include the utilization of graduate student competencies to assess progress and the introduction of novel curricula focused on development of skills, rather than accumulation of facts. Evidence demonstrates that active learning approaches are effective. Several facets of active learning are components of problem-based learning (PBL), which is a teaching modality where student learning is self-directed toward solving problems in a relevant context. These concepts were combined and incorporated in creating a new introductory graduate course designed to develop scientific skills (student competencies) in matriculating doctoral students using a PBL format.
Methods: Evaluation of course effectiveness was measured using the principals of the Kirkpatrick Four Level Model of Evaluation. At the end of each course offering, students completed evaluation surveys on the course and instructors to assess their perceptions of training effectiveness. Pre- and post-tests assessing students' proficiency in experimental design were used to measure student learning.
Results: The analysis of the outcomes of the course suggests the training is effective in improving experimental design. The course was well received by the students as measured by student evaluations (Kirkpatrick Model Level 1). Improved scores on post-tests indicate that the students learned from the experience (Kirkpatrick Model Level 2). A template is provided for the implementation of similar courses at other institutions.
Conclusions: This problem-based learning course appears effective in training newly matriculated graduate students in the required skills for designing experiments to test specific hypotheses, enhancing student preparation prior to initiation of their dissertation research.
Keywords: Critical thinking; Doctoral Student; Experimental design; Graduate; Problem-based learning; Training.
© 2023. BioMed Central Ltd., part of Springer Nature.
PubMed Disclaimer
The authors declare no competing interests.
Grants and funding.
Full text sources.
NCBI Literature Resources
MeSH PMC Bookshelf Disclaimer
The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.
An official website of the United States government
The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.
The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.
The PMC website is updating on October 15, 2024. Learn More or Try it out now .
Michael d. schaller.
Department of Biochemistry and Molecular Medicine, West Virginia University School of Medicine, Robert C. Byrd Health Sciences Center 64 Medical Center Drive, P.O. Box 9142, Morgantown, WV 26506 USA
Michael r. gunther, scott a. weed, associated data.
All data generated in this study are included in this published article and its supplementary information files.
Traditionally, doctoral student education in the biomedical sciences relies on didactic coursework to build a foundation of scientific knowledge and an apprenticeship model of training in the laboratory of an established investigator. Recent recommendations for revision of graduate training include the utilization of graduate student competencies to assess progress and the introduction of novel curricula focused on development of skills, rather than accumulation of facts. Evidence demonstrates that active learning approaches are effective. Several facets of active learning are components of problem-based learning (PBL), which is a teaching modality where student learning is self-directed toward solving problems in a relevant context. These concepts were combined and incorporated in creating a new introductory graduate course designed to develop scientific skills (student competencies) in matriculating doctoral students using a PBL format.
Evaluation of course effectiveness was measured using the principals of the Kirkpatrick Four Level Model of Evaluation. At the end of each course offering, students completed evaluation surveys on the course and instructors to assess their perceptions of training effectiveness. Pre- and post-tests assessing students’ proficiency in experimental design were used to measure student learning.
The analysis of the outcomes of the course suggests the training is effective in improving experimental design. The course was well received by the students as measured by student evaluations (Kirkpatrick Model Level 1). Improved scores on post-tests indicate that the students learned from the experience (Kirkpatrick Model Level 2). A template is provided for the implementation of similar courses at other institutions.
This problem-based learning course appears effective in training newly matriculated graduate students in the required skills for designing experiments to test specific hypotheses, enhancing student preparation prior to initiation of their dissertation research.
The online version contains supplementary material available at 10.1186/s12909-023-04569-7.
For over a decade there have been calls to reform biomedical graduate education. There are two main problems that led to these recommendations and therefore two different prescriptions to solve these problems. The first major issue is the pursuit of non-traditional (non-academic) careers by doctorates and concerns of adequate training [ 1 , 2 ]. The underlying factors affecting career outcomes are the number of PhDs produced relative to the number of available academic positions [ 1 , 3 – 5 ], and the changing career interests of doctoral students [ 6 – 9 ]. One aspect in the proposed reformation to address this problem is incorporation of broader professional skills training and creating awareness of a greater diversity of careers into the graduate curriculum [ 1 , 4 , 5 ]. The second issue relates to the curricula content and whether content knowledge or critical scientific skills should be the core of the curriculum [ 10 , 11 ]. The proposed reformation to address this issue is creation of curricula focusing upon scientific skills, e.g. reasoning, experimental design and communication, while simultaneously reducing components of the curricula that build a foundational knowledge base [ 12 , 13 ]. Components of these two approaches are not mutually exclusive, where incorporation of select specialized expertise in each area has the potential to concurrently address both issues. Here we describe the development, implementation and evaluation of a new problem-based learning (PBL)-based graduate course that provides an initial experience in introducing the scientific career-relevant core competencies of critical thinking and experimental design to incoming biomedical doctoral students. The purpose of this course is to address these issues by creating a vehicle to develop professional skills (communication) and critical scientific skills (critical thinking and experimental design) for first year graduate students.
One approach that prioritizes the aggregate scientific skill set required for adept biomedical doctorates is the development of core competencies for doctoral students [ 5 , 14 , 15 ], akin to set milestones that must be met by medical residents and fellows [ 16 ]. Key features of these competencies include general and field-specific scientific knowledge, critical thinking, experimental design, evaluation of outcomes, scientific rigor, ability to work in teams, responsible conduct of research, and effective communication [ 5 , 14 , 15 ]. Such competencies provide clear benchmarks to evaluate the progress of doctoral students’ development into an independent scientific professional and preparedness for the next career stage. Historically, graduate programs relied on traditional content-based courses and supervised apprenticeship in the mentor’s laboratory to develop such competencies. An alternative to this approach is to modify the graduate student curriculum to provide a foundation for these competencies early in the curriculum in a more structured way. This would provide a base upon which additional coursework and supervised dissertation research could build to develop competencies in doctoral students.
Analyses of how doctoral students learn scientific skills suggest a threshold model, where different skillsets are mastered (a threshold reached), before subsequent skillsets can be mastered [ 17 , 18 ]. Skills like using the primary literature, experimental design and placing studies in context are earlier thresholds than identifying alternatives, limitations and data analysis [ 18 ]. Timmerman et al. recommend revision of graduate curricula to sequentially build toward these thresholds using evidence-based approaches [ 18 ]. Several recent curricular modifications are aligned with these recommendations. One program, as cited above, offers courses to develop critical scientific skills early in the curriculum with content knowledge provided in later courses [ 12 , 13 ]. A second program has built training in experimental design into the coursework in the first semester of the curriculum. Improvements in students experimental design skills and an increase in self-efficacy in experimental design occurred over the course of the semester [ 19 ]. Other programs have introduced exercises into courses and workshops to develop experimental design skills using active learning. One program developed interactive sessions on experimental design, where students give chalk talks about an experimental plan to address a problem related to course content and respond to challenges from their peers [ 20 ]. Another program has developed a workshop drawing upon principles from design thinking to build problem solving skills and creativity, and primarily uses active learning and experiential learning approaches [ 21 ]. While these programs are well received by students, the outcomes of training have not been reported. Similar undergraduate curricula that utilize literature review with an emphasis on scientific thought and methods report increased performance in critical thinking, scientific reasoning and experimental design [ 22 , 23 ].
It is notable that the changes these examples incorporate into the curriculum are accompanied with a shift from didactic teaching to active learning. Many studies have demonstrated that active learning is more effective than a conventional didactic curriculum in STEM education [ 24 ]. Problem-based learning (PBL) is one active learning platform that the relatively new graduate program at the Van Andel Institute Graduate School utilizes for delivery of the formal curriculum [ 25 ]. First developed for medical students [ 26 ], the PBL learning approach has been adopted in other educational settings, including K-12 and undergraduate education [ 27 , 28 ]. A basic tenet of PBL is that student learning is self-directed [ 26 ]. Students are tasked to solve an assigned problem and are required to find the information necessary for the solution (self-directed). In practice, learning occurs in small groups where a faculty facilitator helps guide the students in identifying gaps in knowledge that require additional study [ 29 ]. As such, an ideal PBL course is “well organized” but “poorly structured”. The lack of a traditional restrictive structure allows students to pursue and evaluate different solutions to the problem.
The premise for PBL is that actively engaging in problem solving enhances learning in several ways [ 29 , 30 ]. First, activation of prior knowledge, as occurs in group discussions, aids in learning by providing a framework to incorporate new knowledge. Second, deep processing of material while learning, e.g. by answering questions or using the knowledge, enhances the ability to later recall key concepts. Third, learning in context, e.g. learning the scientific basis for clinical problems in the context of clinical cases, enables and improves recall. These are all effective strategies to enhance learning [ 31 ]. PBL opponents argue that acquisition of knowledge is more effective in a traditional didactic curriculum. Further, development of critical thinking skills requires the requisite foundational knowledge to develop realistic solutions to problems [ 32 ].
A comprehensive review of PBL outcomes from K-12 through medical school indicated that PBL students perform better in the application of knowledge and reasoning, but not in other areas like basic knowledge [ 33 ]. Other recent meta-analyses support the conclusion that PBL, project-based learning and other small group teaching modalities are effective in education from primary school to university, including undergraduate courses in engineering and technology, and pharmacology courses for professional students in health sciences [ 34 – 39 ]. While the majority of the studies reported in these meta-analyses demonstrate that PBL results in better academic performance, there are contrasting studies that demonstrate that PBL is ineffective. This prompts additional investigation to determine the salient factors that distinguish the two outcomes to establish best practices for better results using the PBL platform. Although few studies report the outcomes of PBL based approaches in graduate education, this platform may be beneficial in training biomedical science doctoral students for developing and enhancing critical thinking and practical problem-solving skills.
At our institution, biomedical doctoral students enter an umbrella program and take a core curriculum in the first semester prior to matriculating into one of seven biomedical sciences doctoral programs across a wide range of scientific disciplines in the second semester. Such program diversity created difficulty in achieving consensus on the necessary scientific foundational knowledge for a core curriculum. Common ground was achieved during a recent curriculum revision through the development of required core competencies for all students, regardless of field of study. These competencies and milestones for biomedical science students at other institutions [ 5 , 14 , 15 ], along with nontraditional approaches to graduate education [ 12 , 25 ], were used as guidelines for curriculum modification.
A course was created to develop competencies required by all biomedical sciences doctoral students regardless of their program of interest [ 14 ]. As an introductory graduate level course, this met the needs of all our seven diverse biomedical sciences doctoral programs where our first-year doctoral students matriculate. A PBL platform was chosen for the course to engage the students in an active learning environment [ 25 ]. The process of problem solving in small teams provided the students with experience in establishing working relationships and how to operate in teams. The students gained experience in researching material from the literature to establish scientific background, find current and appropriate experimental approaches and examples of how results are analyzed. This small group approach allowed each team to develop different hypotheses, experimental plans and analyses based upon the overall interests of the group. The course was designed following discussions with faculty experienced in medical and pharmacy school PBL, and considering course design principles from the literature [ 27 , 40 ]. The broad learning goals are similar to the overall objectives in another doctoral program using PBL as the primary course format [ 25 ], and are aligned with recommended core competencies for PhD scientists [ 14 ]. These goals are to:
Students were organized into groups of four or five based on their scientific background. Student expertise in each group was deliberately mixed to provide different viewpoints during discussion. A single faculty facilitator was assigned to each student group, which met formally in 13 separate sessions (Appendix II). In preparation for each session, the students independently researched topics using the literature (related to goal 6) and met informally without facilitator oversight to coordinate their findings and organize the discussion for each class session. During the formal one-hour session, one student served as the group leader to manage the discussion. The faculty facilitator guided the discussion to ensure coverage of necessary topics and helped the students identify learning issues, i.e. areas that required additional development, for the students to research and address for the subsequent session. At the end of each session, teams previewed the leading questions for the following class and organized their approach to address these questions prior to the next session. The whole process provided experiences related to goal 5.
As the course was developed during the COVID-19 pandemic, topics related to SARS-CoV2 and COVID-19 were selected as currently relevant problems in society. Session 1 prepared the students to work in teams by discussing about how to work in teams and manage conflict (related to goal 5). In session 2, the students met in their assigned groups to get to know each other, discuss problem-based learning and establish ground rules for the group. Sessions 3 and 4 laid the course background by focusing on the SARS-CoV2 virus and COVID-19-associated pathologies (related to goal 1). The subsequent nine sessions were organized into three separate but interrelated three-session blocks: one on COVID-19 and blood clotting, one on COVID-19 and loss of taste, and one on SARS-CoV2 and therapeutics. The first session in each of these blocks was devoted to covering background information (blood clotting, neurosensation and drug application)(related to goal 1). The second session of each block discussed hypothesis development (mechanisms that SARS-CoV2 infection might utilize to alter blood clotting, the sense of taste, and identification of therapeutic targets to attenuate SARS-CoV2 infection)(related to goal 3). In the second sessions the students also began to design experiments to test the hypothesis. The final session of each block fleshed out the details of the experimental design (related to goals 2 and 3).
The process was iterative, where the students had three opportunities to discuss hypothesis development, experimental design and analysis during sessions with their facilitators. Written and oral presentation assignments (Appendix V) provided additional opportunities to articulate a hypothesis, describe experimental approaches to test the hypotheses, propose analysis of experimental results and develop communication skills (related to goal 4).
Rigor and reproducibility was incorporated into the course. This was an important component given the emphasis recently placed on rigor and reproducibility by federal agencies. As the students built the experimental design to address the hypothesis, recurring questions were posed to encourage them to consider rigor. Examples include: “ Are the methods and experimental approaches rigorous? How could they be made more rigorous? ” “ Discuss how your controls validate the outcome of the experiment. What additional controls could increase confidence in your result? ” The facilitators were instructed to direct discussion to topics related to the rigor of the experimental design. The students were asked about numbers of replicates, number of animals, additional methods that could be applied to support the experiment, and other measurements to address the hypothesis in a complementary fashion. In the second iteration of the course, we introduced an exercise on rigor and reproducibility for the students using the NIH Rigor and Reproducibility Training Modules (see Appendix III). In this exercise, the students read a short introduction to rigor and reproducibility and viewed a number of short video modules to introduce lessons on rigor. The students were also provided the link to the National Institute of General Medical Sciences clearinghouse of training modules on rigor and reproducibility as reference for experimental design in their future (see Appendix III).
The first delivery of the course was during the COVID-19 pandemic and sessions were conducted on the Zoom platform. The thirteen PBL sessions, and two additional sessions dedicated to oral presentations, were spaced over the course of the first semester of the biomedical sciences doctoral curriculum. The second iteration of the course followed the restructuring of the graduate first year curriculum and the thirteen PBL sessions, plus one additional session devoted to oral presentations, were held during the first three and a half weeks of the first-year curriculum. During this period in the semester, this was the only course commitment for the graduate students. Due to this compressed format, only one written assignment and a single oral presentation were assigned. As the small group format worked well via Zoom in the first iteration of the course, the small groups continued to meet using this virtual platform.
IRB Approval. The West Virginia University Institutional Review Board approved the study (WVU IRB Protocol#: 2008081739). Informed consent was provided by the participants in writing and all information was collected anonymously.
Surveys. Evaluation of training effectiveness was measured in two ways corresponding to the first two levels of the Kirkpatrick Model of Evaluation [ 41 ]. First, students completed a questionnaire upon completion of the course to capture their perceptions of training (Appendix VII). Students were asked their level of agreement/disagreement on a Likert scale with 10 statements about the course and 7 statements about their facilitator. Second, students took a pre- and post-test to measure differences in their ability to design experiments before and after training (Appendix VIII). The pre- and post-tests were identical, asking the students to design an experiment to test a specific hypothesis, include controls, plan analyses, and state possible results and interpretation. Five questions were provided for the pre- and post-test, where each question posed a hypothesis from a different biomedical discipline, e.g. cancer biology or neuroscience. Students were asked to choose one of the five questions to answer.
Peer-to-peer evaluations were collected to provide feedback on professionalism and teamwork. This survey utilized a Goldilocks scale ranging from 1 to 7, with 4 being the desired score. An example peer question asked about accountability, where responses included not accountable, e.g. always late (score = 1), accountable, e.g. punctual, well prepared, follows up (score = 4) and controlling, e.g. finds fault in others (score = 7). Each student provided a peer-to-peer evaluation for each student in their group. (see Appendix VII). In the second course iteration, Goldilocks surveys were collected three times over the three-week course period due to the compressed time frame. This was necessary to provide rapid feedback to the students about their performance during the course in order to provide opportunities to address and rectify any deficits before making final performance assessments.
Evaluating Pre- and Post-Tests. All pre- and post-test answers were evaluated by three graders in a blind fashion, where the graders were unaware if an answer came from a pre- or post-test. Prior to grading, each grader made up individual answer keys based upon the question(s) on the tests. The graders then met to compare and deliberate these preliminary keys, incorporating changes and edits to produce a single combined key to use for rating answers. While the students were asked to answer one question, some students chose to answer several questions. Superfluous answers were used as a training dataset for the graders. The graders independently scored each answer, then met to review the results and discuss modification of the grading key. The established final grading key, with a perfect score of 16, was utilized by the graders in independently evaluating the complete experimental dataset consisting of all pre- and post-test answers (Appendix IX). To assess the ability of student cohorts to design experiments before and after the course, three of the authors graded all of the pre- and post-test answers. Grading was performed in a blind fashion and the scores of the three raters were averaged for each answer.
Statistical analysis. To measure the interrater reliability of the graders, the intraclass correlation coefficient (ICC) was calculated. A two-way mixed effects model was utilized to evaluate consistency between multiple raters/measurements. The ICC for grading the training dataset was 0.82, indicating a good inter-rater agreement. The ICC for grading the experimental dataset was also 0.82. For comparison of pre-test vs. post-test performance, the scores of the three raters were averaged for each answer. Since answers were anonymous, the analyses compared responses between individuals. Most, but not all scores, exhibited a Gaussian distribution and therefore a nonparametric statistic, a one-tailed Mann Whitney U test, was used for comparison. The pre-test and post-test scores for 2020 and 2021 could not be compared due to the different format used for the course in each year.
Thirty students participated in the course in the first offering, while 27 students were enrolled in the second year. The students took pre- and post-tests to measure their ability to design an experiment before and after training (Appendix VIII). As the course progressed, students were surveyed on their views of the professionalism of other students in their group (Appendix VII). At the end of the course, students were asked to respond to surveys evaluating the course and their facilitator (see Appendix VII).
Student reception of the course (Kirkpatrick Level 1) . In the first year, 23 students responded to the course evaluation (77% response rate) and 26 students submitted facilitator evaluations (87% response rate), whereas in the second year there were 25 responses to the course evaluation (93% response rate) and 26 for facilitators (96% response rate). Likert scores for the 2020 and 2021 course evaluations are presented in Fig. 1 . The median score for each question was 4 on a scale of 5 in 2020. In 2021, the median scores for the questions about active learning and hypothesis testing were 5 and the median score of the other questions was 4. The students appreciated the efforts of the facilitators in the course, based upon their evaluations of the facilitators. The median score for every facilitator across all survey questions is shown in Fig. 2 . The median score for a single question in 2020 and 2021 was 4.5 and the median score for all other questions was 5. The results of the peer-to-peer evaluations are illustrated in Fig. 3 . The average score for each student were plotted, with scores further from the desired score of 4 indicating perceived behaviors that were not ideal. The wide range of scores in the 2020 survey were noted. The students completed three peer-to-peer surveys during the 2021 course. The range of scores in the 2021 peer-to-peer evaluation was narrower than the range in the 2020 survey. The range of scores was expected to narrow from the first (initial) to third (final) survey as students learned and implemented improvements in their professional conduct based upon peer feedback. The narrow range of scores in the initial survey left little room for improvement.
Results of Course Evaluations by Students. Student evaluations of the course were collected at the end of each offering. The evaluation surveys are in Appendix VII. Violin plots showing the distribution and median score for each question in the 2020 survey (A) and the 2021 survey (B) are shown. The survey used a Likert scale (1 – low to 5 – high)
Results of Facilitator Evaluations by Students. Student evaluations of the facilitators were collected at the end of each offering of the course. The evaluation surveys are in Appendix VII. Violin plots showing the distribution and median score for each question in the 2020 survey (A) and the 2021 survey (B) are shown. The survey used a Likert scale (1 – low to 5 – high)
Results of Student Peer-to-Peer Evaluations. Student peer-to-peer evaluations were collected at the end of the course in year 1 (A) , and at the beginning (B) , the middle (C) and the end (D) of the course in year 2. Each student evaluated the professionalism of each other student in their group using the evaluation survey shown in Appendix VII. The average score for each student is plotted as a data point. The survey used a Goldilocks scale (range of 1 to 7) where the desired professional behavior is reflected by a score of 4
Student learning (Kirkpatrick Level 2). Twenty-six students completed the pre-test in each year and consented to participate in this study (87% response in the first year and 96% response in the second year). Eighteen students completed the post-test at the end of the first year (60%) and 26 students completed the test at the end of the second year (96%). Question selection (excluding students that misunderstood the assignment and answered all questions) is shown in Table 1 . The most frequently selected questions were Question 1 (45 times) and Question 2 (23 times). Interestingly, the results in Table 1 also indicate that students did not necessarily choose the same question to answer on the pre-test and post-test.
Student Choice of Experimental Question to Answer (Only those who made a choice)
2020 | 2021 | |||
---|---|---|---|---|
Pre-test | Post-test | Pre-test | Post-test | |
10 | 8 | 13 | 14 | |
3 | 5 | 7 | 8 | |
2 | 2 | 2 | 3 | |
4 | 1 | 4 | 0 | |
2 | 0 | 0 | 1 |
Average scores on pre-tests and post-tests were compared using a one-tailed Mann Whitney U test. Since the format of the course was different in the two iterations, comparison of test results between the two years could not be made. The average scores of the pre- and post-test in 2020 were not statistically different (p = 0.0673), although the post-test scores trended higher. In contrast, the difference between the pre- and post-test in 2021 did reach statistical significance (p = 0.0329). The results collectively indicate an overall improvement in student ability in experimental design (Fig. 4 ).
Pre- and Post-Test Scores. At the beginning and end of each offering, the students completed a test to measure their ability to design an experiment (see Appendix VIII for the details of the exam). Three faculty graded every answer to the pre- and post-test using a common grading rubric (see Appendix IX). The maximum possible score was 16. The average score for each individual answer on the pre-test and post-test is represented as a single data point. The bar indicates the mean score across all answers +/- SD. The average scores of the pre- and post-test scores were compared using a one-tailed Mann Whitney U test. For the 2020 data (A) , p = 0.0673, and for the 2021 data (B) , p = 0.0329
This course was created in response to biomedical workforce training reports recommending increased training in general professional skills and scientific skills, e.g. critical thinking and experimental design. The course utilizes a PBL format, which is not extensively utilized in graduate education, to incorporate active learning throughout the experience. It was well received by students and analysis suggests that major goals of the course were met. This provides a template for other administrators and educators seeking to modify curricula in response to calls to modify training programs for doctoral students.
Student evaluations indicated the course was effective at motivating active learning and that students became more active learners. The evaluation survey questions were directly related to three specific course goals: (1) Students reported developing skills in problem solving, hypothesis testing and experimental design. (2) The course helped develop oral presentation skills and written communication skills (in one iteration of the course) and (3) students developed collaboration and team skills. Thus, from the students’ perspective, these three course goals were met. Student perceptions of peer professionalism was measured using peer-to-peer surveys. The wide range of Goldilocks scores in the first student cohort was unexpected. In the second student cohort changes in professional behavior were measured over time and the score ranges were narrower. The reasons for the difference between cohorts is unclear. One possibility for this discrepancy is that the first iteration of the course extended over one semester and was during the first full semester of the pandemic, impacting professional behavior and perceptions of professionalism. The second cohort completed a professionalism survey three times during the course. The narrow range of scores from this cohort in the initial survey made detection of improved professionalism over the course difficult. Results do indicate that professionalism improved in terms of respect and compassion between the first and last surveys. Finally, the results of the pre-test and post-test analysis demonstrated a trend of improved performance on the post-test relative to the pre-test for students in each year of the course and a statistical difference between the pre- and post-test scores in the second year.
Areas for improvement. The course was initially offered as a one-credit course. Student comments on course evaluations and comments in debriefing sessions with facilitators at the end of the course concurred that the work load exceeded that of a one credit course. As a result, the year two version was offered as a two-credit course to better align course credits with workload.
There were student misperceptions about the goals of the course in the first year. Some students equated experimental design with research methods and expressed disappointment that this was not a methods course. While learning appropriate methods is a goal of the course, the main emphasis is developing hypotheses and designing experiments to test the hypotheses. As such, the choice of methods was driven by the hypotheses and experimental design. This misperception was addressed in the second year by clearly elaborating on the course goals in an introductory class session.
The original course offering contained limited statistical exercises to simulate experimental planning and data analysis, e.g. students were required to conduct a power analysis. Between the first and second years of the course, the entire first semester biomedical sciences curriculum was overhauled with several new course offerings. This new curriculum contained an independent biostatistics workshop that students completed prior to the beginning of this course. Additional statistics exercises were incorporated into the PBL course to provide the students with more experience in the analysis of experimental results. Student evaluations indicated that the introduction of these additional exercises was not effective. Improved coordination between the biostatistics workshop and the PBL course is required to align expectations, better equipping students for the statistical analysis of experimental results encountered later in this course.
An important aspect that was evident from student surveys, facilitator discussions and debrief sessions was that improved coordination between the individual facilitators of the different groups is required to reduce intergroup variability. Due to class size, the students were divided into six groups, with each facilitator assigned to the same group for the duration of the course to maintain continuity. The facilitators met independent of the students throughout the course to discuss upcoming sessions and to share their experiences with their respective groups. This allowed the different facilitators to compare approaches and discuss emerging or perceived concerns/issues. In the second year, one facilitator rotated between different groups during each session to observe how the different student groups functioned. Such a real time faculty peer-evaluation process has the potential to reduce variability between groups, but was challenging to implement within the short three-week time period. Comprehensive training where all facilitators become well versed in PBL strategies and adhere to an established set of guidelines/script for each session is one mechanism that may reduce variability across different facilitator-group pairings.
Limitations. The current study has a number of limitations. The sample size for each class was small, with 30 students enrolled in the first year of the course and 27 students enrolled in the second. The response rates for the pre-tests were high (> 87%) but the response rate for the post-test varied between the first year (60%) and second year (96%) of the course. The higher response rate in the second year might be due to fewer end of semester surveys since this was the only course that the students took in that time period. Additionally, the post-test in the second year was conducted at a scheduled time, rather than on the student’s own time as was the case in year one. Due to restructuring of the graduate curriculum and the pandemic, the two iterations of the course were formatted differently. This precluded pooling the data from the two offerings and makes comparison between the outcomes difficult.
Presentation of the course was similar, but not identical, to all of the students. Six different PBL groups were required to accommodate the number of matriculating students in each year. Despite efforts to provide a consistent experience, there was variability between the different facilitators in running their respective groups. Further, the development of each session in each group was different, since discussion was driven by the students and their collective interests. These variables could be responsible for increasing the spread of scores on the post-tests and decreasing the value of the course for a subset of students.
The pre- and post-tests were conducted anonymously to encourage student participation. This prevented correlating the differential between pre- and post-test scores for each student and in comparing learning between different groups. The pre-test and post-test were identical, and provided the students with five options to design experiments (with identical instructions) in response to a different biomedical science problem. An alternative approach could have used isomorphic questions for the pre- and post-tests. It is clear that some students answered the same question on the pre- and post-test, and may benefit from answering the same question twice (albeit after taking the course). Some students clearly answered different questions on the pre- and post-test and the outcomes might be skewed if the two questions challenged the student differently.
While the course analysis captured the first two levels of the Kirkpatrick model of evaluation (reaction and learning), it did not attempt to measure the third level (behavior) or fourth level (results) [ 41 ]. Future studies are required to measure the third level. This could be achieved by asking students to elaborate on their experimental design used in recent experiments in their dissertation laboratory following completion of the course, or by evaluating the experimental design students incorporate into their dissertation proposals. The fourth Kirkpatrick level could potentially be assessed by surveying preceptors about their students’ abilities in experimental design in a longitudinal manner at semi- or annual committee meetings and accompanying written progress reports. The advantage of focusing on the first two Kirkpatrick levels of evaluation is that the measured outcomes can be confidently attributed to the course. Third and fourth level evaluations are more complicated, since they necessarily take place at some point after completion of the course. Thus, the third and fourth level outcomes can result from additional factors outside of the course (e.g. other coursework, working in the lab, attendance in student-based research forum, meeting with mentors, etc.). Another limiting factor is the use of a single test to measure student learning. Additional alternative approaches to measure learning might better capture differences between the pre- and post-test scores.
Implementation. This curriculum is readily scalable and can be modified for graduate programs of any size, with the caveat that larger programs will require more facilitators. At Van Andel, the doctoral cohorts are three to five new students per year and all are accommodated in one PBL group [ 25 ]. At our institution, we have scaled up to a moderate sized doctoral program with 25 to 30 matriculating students per year, dividing the students into six PBL groups (4–5 students each). Medical School classes frequently exceed 100 students (our program has 115–120 new students each fall) and typically have between five and eight students per group. Our graduate course has groups at the lower end of this range. This course could be scaled up by increasing the number of students in the group or by increasing the number of groups.
Consistency between groups is important so each group of students has a similar experience and reaps the full benefit of this experience. Regular meetings between the course coordinator and facilitators to discuss the content of upcoming sessions and define rubrics to guide student feedback and evaluation were mechanisms used to standardize between the different groups in this course (Appendix VI). In hindsight, the course would benefit from more rigorous facilitator training prior to participation in the course. While a number of our facilitators were veterans of a medical school PBL course, the necessary skillset required to effectively manage a graduate level PBL course that is centered on developing critical thinking and experimental design are different. Such training requires an extensive time commitment by the course coordinators and participating facilitators.
The most difficult task in developing this course involved the course conception and development of the problem-based assignments. Designing a COVID-19 based PBL course in 2020 required de novo development of all course material. This entailed collecting and compiling information about the virus and the disease to provide quick reference for facilitators to guide discussion in their groups, all in the face of constantly shifting scientific and medical knowledge, along with the complete lack of traditional peer-based academic social engagement due to the pandemic. In development of this course, three different COVID-based problems were identified, with appropriate general background material for each problem requiring extensive research and development. Background material on cell and animal models, general strategies for experimental manipulation and methods to measure specific outcomes were collected in each case. Student copies for each session were designed to contain a series of questions as a guide to identifying important background concepts. Facilitator copies for each session were prepared with the goal of efficiently and effectively guiding each class meeting. These guidelines contained ideas for discussion points, areas of elaboration and a truncated key of necessary information to guide the group (Appendix IV). Several PBL repositories exist (e.g. https://itue.udel.edu/pbl/problems/ , https://www.nsta.org/case-studies ) and MedEdPORTAL ( https://www.mededportal.org/ ) publishes medical-specific cases. These provide valuable resources for case-based ideas, but few are specifically geared for research-focused biomedical graduate students. As such, modification of cases germane to first year biomedical graduate students with a research-centered focus is required prior to implementation. Finally, appropriate support materials for surveys and evaluation rubrics requires additional development and refinement of current or existing templates to permit improved evaluation of learning outcomes (Appendix VI).
Development of an effective PBL course takes considerable time and effort to conceive and construct. Successful implementation requires the requisite higher administrative support to identify and devote the necessary and appropriate faculty needed for course creation, the assignment of skilled faculty to serve as facilitators and staff support to coordinate the logistics for the course. It is critical that there is strong faculty commitment amongst the facilitators to devote the time and energy necessary to prepare and to successfully facilitate a group of students. Strong institutional support is linked to facilitator satisfaction and commitment to the PBL-based programs [ 42 ]. Institutional support can be demonstrated in multiple ways. The time commitment for course developers, coordinators and facilitators should be accurately reflected in teaching assignments. Performance in these roles in PBL should factor into decisions about support for professional development, e.g. travel awards, and merit based pay increases. Further, efforts in developing, implementing and executing a successful PBL course should be recognized as important activities during annual faculty evaluations by departmental chairs and promotion and tenure committees.
Key Takeaways. The creation and implementation of this course was intellectually stimulating and facilitators found their interactions with students gratifying. From student survey responses and test results the course was at least modestly successful at achieving its goals. Based upon our experience, important issues to consider when deciding to implement such a curriculum include: (1) support of the administration for developing the curriculum, (2) facilitator buy-in to the approach, (3) continuity (not uniformity) between PBL groups, (4) other components of the curriculum and how they might be leveraged to enhance the effectiveness of PBL and (5) effort required to develop and deliver the course, which must be recognized by the administration.
Future Directions. Novel curriculum development is an often overlooked but important component to contemporary graduate student education in the biomedical sciences. It is critical that modifications incorporated in graduate education are evidence based. We report the implementation of a novel PBL course for training in the scientific skill sets required for developing and testing hypotheses, and demonstrate its effectiveness. Additional measures to assess the course goals in improving critical thinking, experimental design and self-efficacy in experimental design will be implemented using validated tests [ 22 , 43 – 45 ]. Further studies are also required to determine the long-term impact of this training on student performance in the laboratory and progression towards degree. It will be interesting to determine if similar curriculum changes to emphasize development of skills will shorten the time to degree, a frequent recommendation for training the modern biomedical workforce [ 1 , 46 – 48 ].
Incorporation of courses emphasizing development of skills can be done in conjunction with traditional didactic instruction to build the necessary knowledge base for modern biomedical research. Our PBL course was stand-alone, requiring the students to research background material prior to hypothesis development and experimental design. Coordination between the two modalities would obviate the need for background research in the PBL component, reinforce the basic knowledge presented didactically through application, and prepare students for higher order thinking about the application of the concepts learned in the traditional classroom. Maintaining a balance between problem-based and traditional instruction may also be key in improving faculty engagement into such new and future initiatives. Continued investments in the creation and improvement of innovative components of graduate curricula centered around developing scientific skills of doctoral students can be intellectually stimulating for faculty and provide a better training environment for students. The effort may be rewarded by streamlining training and strengthening the biomedical workforce of the future.
Below is the link to the electronic supplementary material.
Thanks to Mary Wimmer and Drew Shiemke for many discussions over the years about PBL in the medical curriculum and examples of case studies. We thank Steve Treisenberg for initial suggestions and discussions regarding PBL effectiveness in the Van Andel Institute. Thanks to Paul and Julie Lockman for discussions about PBL in School of Pharmacy curricula and examples of case studies. Special thanks to the facilitators of the groups, Stan Hileman, Hunter Zhang, Paul Chantler, Yehenew Agazie, Saravan Kolandaivelu, Hangang Yu, Tim Eubank, William Walker, and Amanda Gatesman-Ammer. Without their considerable efforts the course could never have been successfully implemented. Thanks to the Department of Biochemistry and Molecular Medicine for supporting the development of this project. MS is the director of the Cell & Molecular Biology and Biomedical Engineering Training Program (T32 GM133369).
PBL | Problem-based learning |
STEM | Science, technology, engineering, and math |
K-12 | kindergarten through grade 12 |
ICC | Intraclass coefficient> |
SARS-CoV2 | severe acute respiratory syndrome coronavirus 2 |
COVID-19 | Coronavirus disease 19 |
SW and MS developed the concept for the course. MS was responsible for creation and development of all of the content, for the implementation of the course, the design of the study and creating the first draft of the manuscript. MG, MRG and SW graded the pre- and post-test answers in a blind fashion. MS, MG, MRG and SW analyzed the data and edited the manuscript.
There was no funding available for this work.
Declarations.
The authors declare no competing interests.
The West Virginia University Institutional Review Board approved the study (WVU IRB Protocol#: 2008081739). Informed consent was provided in writing and all information was collected anonymously. All methods were carried out in accordance with relevant guidelines and regulations.
Not applicable.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Scientific Reports volume 14 , Article number: 21575 ( 2024 ) Cite this article
Metrics details
Intrinsically safe solenoids drive solenoid valves in coal mining equipment. The low power consumption of these solenoids limits the response time of the solenoid valves. Additionally, the low viscosity and high susceptibility to dust contamination of the emulsion fluid often lead to leakage and sticking of hydraulic valves. This study proposes a low-power-driven, large-flux, fast-response three-stage valve structure with an internal displacement feedback device to address these issues. The critical parameters of the valve were optimized using a novel multi-objective optimization algorithm. A prototype was manufactured based on the obtained parameters and subjected to simulation and experimental verification. The results demonstrate that the valve has an opening time of 21 ms, a closing time of 12 ms, and a maximum flow rate of approximately 225 L/min. The driving power of this structure is less than 1.2 W. By utilizing this valve for hydraulic cylinder control, a positioning accuracy of ± 0.15 mm was achieved. The comparative test results show that the proposed structural control error fluctuation is smaller than that of the 3/4 proportional valve.
High-pressure water-based hydraulic valves are widely used in coal mining equipment. However, the water-based transmission medium in the hydraulic system of comprehensive mining faces is highly susceptible to dust contamination, leading to common issues such as valve components jamming and leaking during operation 1 , 2 . Therefore, enhancing the pollution resistance of valves and reducing internal leakage have become urgent issues to address 3 . The failure of hydraulic valves can result in numerous malfunctions of controlled equipment, directly threatening the safety of coal miners, damaging mechanical equipment in the work area, increasing maintenance workload, and severely affecting production efficiency on the work surface 4 .
The working medium of the hydraulic valves in the fully mechanized mining face is a high water-based fluid with 5% oil and 95% water, forming an oil-in-water (O/W) emulsion, which has low viscosity and operates at high pressures, typically around 32 MPa 5 . Under the exact clearance and pressure drop, the leakage loss of high-pressure water-based hydraulic valves of the same specifications is tens of times that of oil pressure control valves, requiring a tight matching in their structure 6 . Additionally, internal leakage must be almost zero when the spool is closed, necessitating the avoidance of slide-valve structures 7 . To meet these performance indicators, it is necessary to consider the sole physicochemical properties of water-based medium comprehensively, improve the structure and design parameters of hydraulic valves, or develop entirely new designs.
In addition to the issues of clutch and leakage, the driving solenoid power of control valves in coal mining equipment should be less than 1.2 W 8 . To meet the system's mass flow rate requirements, existing research often adopts a two-stage or three-stage valve structure 9 , 10 . In addition to the high flow rate, the hydraulic system in the fully mechanized mining face imposes stricter demands on response time than conventional hydraulic systems. However, due to factors such as large flux, spool mass, and large stroke, a conflict exists between large flux and fast response, making the design of such control valves challenging 11 .
High-speed on–off valves (HSV), belonging to a type of digital hydraulic component, are always in either fully open or fully closed states. They can convert discrete control signals into discrete flow rates, offering advantages such as fast response, simple structure, high reliability, and insensitivity to hydraulic fluid contamination 12 . However, due to their inability to provide significant flow rates during rapid responses, multiple actuators are often employed for small-scale equipment or pilot valves 13 . To address this limitation, without altering the structure and type of high-speed valves, multiple high-speed valves are typically connected in parallel to form a Digital Flow Control Unit (DFCU) to expand flow rate gain and resolve the contradiction mentioned above 14 .
Previous research on high-flow and fast-response valves has primarily focused on proportional/servo or HSV 15 , 16 , 17 . In recent years, there has also been significant progress in pilot valves for HSV 18 , 19 . However, it is almost inevitable to have spool structures when controlling flow direction in hydraulic systems using proportional/servo valves 20 . Some HSVs, driven by armature coils with permanent magnet structures in on–off valves, can generate significant flow rates, but they operate at low pressures and have internal piston-type structures 21 , 22 . On the other hand, standard high-flow cartridge valves cannot meet the demand due to their lack of fast response 23 .
Multi-objective optimization algorithms are commonly employed in the optimization design of valve structures 24 , 25 , 26 . The proposed three-stage valve (TSV) exhibits strong coupling relationships among its stages, making its critical design parameters difficult to determine. With multiple design parameters and mutual constraints, obtaining the optimal solution by optimizing a single parameter is highly challenging. The optimization design problem of the TSV is a multi-objective optimization problem, with advanced solutions achieved through optimization algorithms. Among them, metaheuristic methods find extensive engineering applications. Metaheuristic algorithms refer to an algorithmic framework independent of specific problems, where heuristic methods are inspired by natural phenomena, biological behaviors, or even mathematical principles 27 . Metaheuristic methods possess advantages such as randomness, ease of implementation, and consideration of black-box scenarios, making them adept at tackling complex engineering problems 28 .
Among various metaheuristic algorithms, the Electric Eel Foraging Algorithm (EEFO) has recently gained significant attention in the literature. This algorithm mathematically models four key foraging behaviors of electric eel groups: “Interacting, Resting, Hunting, and Migrating.” It provides exploration and exploitation during the optimization process. An energy factor is developed to manage the transition from global to local search and balance exploration and exploitation in the search space. Test results have shown that EEFO exhibits excellent performance in exploitation, balancing exploration and exploitation, and avoiding local optima 29 .
Structural adjustments are made to the poppet valve to avoid using a slide-valve structure. In this study, the displacement-flow feedback principle is utilized, where the piston in the lift valve structure is replaced with a pressure control chamber. Four displacement feedback orifices are installed on the valve spool to control the valve opening based on the pressure control chamber's pressure.
The displacement-flow feedback principle, also known as the “Valvistor” valve control principle, was proposed by ANDERSSON from Linköping University in Sweden in the 1980s. It has advantages such as a simple structure and gratifying dynamic and static characteristics and has been widely applied 30 , 31 . By replacing the slide structure in the poppet valve with a pressure control chamber, a TSV structure, as shown in Fig. 1 , is proposed in this research. The three-stage poppet valve comprises a pilot HSV, a secondary stage, and a main valve. The two pilot valves are a bidirectional poppet valve and an HSV. The modified poppet valve still has a slide-valve structure, which means there may be leakage. However, the sealing of this valve does not rely on the piston structure. The leakage flow from the piston structure towards the C port is reliably blocked by the poppet valve in the previous stage, together with the flow passing through the feedback orifice 32 . Therefore, the novel TSV structure does not have leakage issues. Since there is no requirement for the sealing performance of the piston structure and a displacement feedback orifice is present on the mating surface, the spool will not experience clutch phenomena.
Working principle of the three-stage valve and structure of the spool.
The flow controlled by the secondary stage should be sufficiently large to achieve a fast step response at the main stage. Therefore, the size and flow capacity of the secondary valve are always more extensive than those of conventional high-flow directional valves and HSVs. As a result, the HSV's electromagnetic driving force is insufficient to drive the secondary valve directly without a pilot valve. Adding a pilot stage to the secondary stage drives the secondary stage by pressure instead of magnetic force. Since the steady-state hydrodynamic force acting on the hydraulic lift valve is much weaker than the pressure, hydrodynamic forces' influence on the spool motion is neglected 28 . At the same time, the pilot valve's size and the electromagnetic valve's driving force are reduced, resulting in a shorter response time for the pilot stage. Due to these advantages, the valve adopts a three-stage structure. The secondary stage and the main valve are poppet valves with displacement feedback structures that connect ports A and C . Due to the larger area of the top end of the poppet valve compared to the bottom end, the poppet valve can reliably close under static pressure. At the same time, the internal leakage between the spool and the sleeve can be shielded by the pilot stage. The oil supply port P is connected to high-pressure oil, and the oil return port T is connected to the load.
The working principle of the TSV is as follows. When a control signal is input, the pilot HSV opens, causing a pressure decrease at port C 2 . As a result, the secondary stage opens, leading to a pressure decrease at port C 1, and the main valve opens. The closing process is the same as the opening process, and after the pilot HSV is closed, the secondary and main valves close sequentially. This valve has a proportional throttling function in the A-B direction and a check valve cutoff function in the other direction.
The hydraulic system discussed in this article is depicted in Fig. 2 . As shown, the position of the inertial load is driven by an asymmetric single-rod hydraulic cylinder, actuated by four TSVs (Throttle Servo Valves). The flow input of the TSVs is adjusted via the control signal u . When u > 0, valves (a) and (d) are open, and valves (b) and (c) are closed. High-pressure fluid from the oil pump flows through valve (a) into the rodless chamber of the hydraulic cylinder, causing it to extend under the pressure differential. Conversely, when u < 0, valves (b) and (c) are open, and valves (a) and (d) are closed. High-pressure fluid from the pump flows through valve (c) into the rod chamber of the hydraulic cylinder, causing it to retract due to the pressure differential. P s = 7 MPa, P t = 0 MPa, and each TSV’s flow rate should exceed 100 L/min ( Q 01 > 100 L/min, Q 02 > 100 L/min). The response time should be minimized while meeting the required flow rates.
Hydraulic system.
Before analyzing the system's dynamic characteristics, it is necessary to determine the displacements of the spools in each stage of the TSV at the steady state. This will provide information on the flow rates at each stage and the relationship between the total flow rate and critical design parameters.
The flow through the pilot HSV can be represented as:
The flow through the secondary stage can be represented as:
The flow through the feedback throttling grooves of the secondary stage can be represented as:
The flow through the main stage can be represented as:
The flow through the feedback throttling grooves of the main valve can be represented as:
The total flow equation through the TSV can be represented as:
When the valve is in a steady state, the flow through the feedback throttling grooves of the main stage is equal to the sum of the flows through the secondary stage and the pilot HSV:
The flow through the secondary feedback throttle slot is equal to the flow through the pilot HSV:
In the stable open state of the valve, neglecting the influence of hydrodynamic forces, the force balance equations for the main and secondary spools can be derived:
To obtain the displacement of the main stage spool when the pilot HSV is opened, the following assumptions are made: the flow coefficient of each orifice is the same, the flow inside the pressure control chamber has no effect on the pressure distribution, and the pressure acting on the surfaces of the spool is uniformly distributed, and the force exerted by the spring on each spool is neglected. Firstly, when the spool is at maximum lift, the net force acting on the spool should be zero in the axial direction, and the pressures inside each pressure control chamber are as follows:
Substituting Eqs. ( 1 ), and ( 3 ) into ( 8 ), the displacement of the secondary stage spool is:
Substituting Eqs. ( 1 ), ( 2 ), and ( 5 ) into ( 7 ), the displacement of the main valve spool is:
Equations ( 11 ), ( 12 ), ( 13 ), and ( 14 ) can be used to solve for the openings of the main and secondary stages. An iterative method can be employed here, starting with a reasonable initial value for Eqs. ( 9 ) and ( 10 ), then substituting the calculated values of P c 1 and P c 2 into Eqs. ( 13 ) and ( 14 ), and obtaining x 1 and x 2 . This process is repeated until the interpolated result of two iterations is small enough.
If the influence of the spring force is neglected, by observing Eqs. ( 13 ) and ( 14 ), it can be concluded that W t 2 affects the opening of the secondary stage spool; increasing x 02 or W t 2 will decrease the value of x 2 . And x 2 affects the opening of the main stage; increasing x 01 , decreasing x 2 , or increasing W t 1 will decrease the value of x 2 . Therefore, the design parameters that impact the system flow most are W t 1 , W t 2 , x 01 , and x 02 . To improve the flow capacity of the valve, further optimization design should be carried out.
Even without considering the effects of the circuit, the HSV system is still a third-order system. Since no adjustments are made to the structure of the HSV to simplify calculations, it is assumed that the movement of the secondary stage and the main valve starts after the HSV is fully opened. At this point, the flow through the HSV can be simplified as a function of the control chamber pressure P c2 . The analysis of the dynamic characteristics of the TSV involves the damping analysis of the orifices in the hydraulic control circuit and the force analysis of the poppet valves at each stage, as shown in Fig. 3 . It can be observed that the shared control chamber pressure P ci serves as the bridge and link between the stages. There is a strong coupling relationship between each spool, and the interaction between the stages achieves the implementation of the displacement feedback principle and the dynamic and static performance of the hydraulic valve. The dynamic equations for the poppet valve components can be derived based on Newton’s second law:
Force analysis of the moving parts.
When considering dynamics, the instantaneous flow through each orifice no longer has a simple mathematical relationship, and real-time calculations need to be performed based on the dynamic pressures before and after the orifices. The pressure in the pressure control chambers also needs to be integrated using the flow continuity equation. According to the chamber-node method, each control chamber is treated as a flow-node. The dynamic equations for the pressure in each pressure control chamber can be written as follows:
For the general flow Eq. ( 1 ) of an orifice, if the flow area A has an analytical relationship with the displacement x of a particular moving component, the flow equation can be linearized near the operating point using a first-order Taylor expansion:
Here, Q ω , A ω , P ω in , P ω out , and x ω represent the flow, flow area, inlet pressure, outlet pressure, and displacement of the orifice at the operating point, while P in , P out , and x represent the incremental changes in inlet pressure, outlet pressure, and displacement near the operating point. k p and k x are the incremental gain factors for the flow-pressure and flow-displacement relationships of the orifice at the operating point. The above equation can be rewritten to express the incremental flow:
Here, δQ represents the incremental flow of the orifice near the operating point. By neglecting the influence of the displacement of the moving component on the dynamic chamber volume, the dynamic Eqs. ( 16 ) and ( 17 ) for the pressure control chambers can be linearized to obtain their linearized forms:
The coefficients k p 1 , k p 2 , k p 3 , k p 4 , k x 1 , and k x 2 can be calculated as follows:
The input variables P s and P t are proposed, while the incremental displacements x 1 and x 2 of the spools in the poppet valves are selected as the respective output variables. The incremental displacements x 1 and x 2 between the second-stage and main spools, their first-order derivatives concerning time \(\dot{x}_{1}\) and \(\dot{x}_{2}\) , and the pressures P c 1 and P c 2 in the control chambers form six state variables. By rearranging Eqs. ( 15 )–( 28 ), the state-space representation of the linearized feedback principle of displacement flow within the TSV near the operating point can be obtained.
where the state variables X are structured as follows:
The input variable U and the output variable Y are structured as follows:
The components of the state matrix A , input matrix B , and output matrix C for the linearized state-space model of the expanded TSV are as follows:
Using the system's state-space model, the unit step response can be obtained, which allows for the analysis of the system’s time response. The step response curve of the system is shown in Fig. 3 , where the TSV's response time is considered the adjustment time. Adjusting the input parameters can achieve different system response times. The system's stability can be checked during the computation based on the system’s poles and zeros distribution.
As shown in Table 1 , there are six key design parameters determining the TSV system, including W t1 , W t2 , D 1 , D 2 , α 1 , and α 2 , representing the width of the main valve spool throttle slot, the width of the secondary valve spool throttle slot, the diameter of the main spool, the diameter of the secondary school, the area ratio of the main spool, and the area ratio of the secondary spool, respectively. The initial values and optimization ranges of these parameters are given in Table 1 , where the initial values are based on the flow requirements of the valve to meet system stability. The optimization ranges comprehensively consider factors such as the volume of the valve, output flow rate, machining difficulty, etc.
The orifice area of the throttle slot in the TSV should exceed that of the second-stage valve. Hence, W t1 is set with a minimum value of 2 mm. Increasing W t1 reduces the opening of the third-stage spool, so a more prominent upper bound is set to allow the optimization algorithm to compute over a broader range. Although choosing a smaller value for W t2 can achieve a more significant flow amplification coefficient, the throttle slot area should exceed that of the pilot HSV. Considering machining precision and potential liquid contamination during TSV operation, W t2 should not be set too small. Hence, its minimum is set at 1mm. Furthermore, a conservative upper limit is defined to prevent system instability. The initial values of D 1 , D 2 , α 1 , and α 2 are selected based on the flow amplification formula ( 14 ), aiming for roughly equal openings of the two relief valves to ensure simultaneous closure of their spools while maintaining system stability.
The preliminary design of the system parameters results in the step response curve of the system shown in Fig. 4 . After the control signal input, the system exhibits an overshoot of 73.6% and reaches a steady-state value of 0.557 after three oscillations. The required settling time is 18.17 ms.
System step response.
Optimization refers to finding the optimal solution or acceptable approximations among numerous solutions for a given problem under certain conditions. Optimization can significantly improve problem-solving efficiency, reduce computational requirements, and save financial resources. The optimization design problem of the novel TSV is a MOO problem. In general, the sub-objectives in an MOO problem are conflicting, and improving one sub-objective may lead to a decrease in the performance of another or several other sub-objectives. It is not possible to simultaneously achieve the optimal values for multiple sub-objectives. Still, coordination and compromise among them are required to achieve the best possible values for each sub-objective 33 . The essential difference between MOO and single-objective optimization problems is that the solution is not unique but consists of a set of Pareto optimal solutions composed of numerous non-dominated solutions. The decision vectors in the solution set are called non-dominated solutions. The corresponding objective functions to the non-dominated vectors in the Pareto optimal set are represented graphically as the Pareto frontier.
The main parameters that affect the performance of the TSV are W t1 , W t2 , A c1 , A c2 , ɑ 1 , and ɑ 2 . When optimizing these parameters, they are interrelated and have trade-offs: increasing the area ratio ɑ 1 , ɑ 2 can enhance the valve spool lift and flux, but it also increases the response time. Increasing the width of the feedback grooves W t1 and W t2 can improve the response time of the poppet valves but may reduce the valve lift and result in reduced flow. The fluctuation of each parameter makes it challenging to obtain an optimal solution by optimizing individual parameters.
This study's critical performance indicators for the designed TSV include response time, flow capacity, volume, weight, and manufacturing complexity. Response time denotes how quickly the valve reacts to input signals; a shorter response time enables precise control over fluid or pressure changes, thereby enhancing system controllability. Flow capacity refers to the amount of fluid the valve can process; higher flow output reduces actuator response times and improves system efficiency. Minimizing the valve’s size and weight facilitates easier installation and transportation. However, these parameters are not primary optimization targets due to the inherent conflict between volume/weight and flow capacity in three-way valves. Manufacturing complexity primarily pertains to technical challenges and cost factors during production and assembly, influencing structural dimensions to avoid excessively narrow channels and intricate grooves. Therefore, this study focuses on optimizing two objectives: response time and flow capacity.
This study utilizes an optimizer called EEFO, proposed in reference 29 , to simulate the foraging behavior of electric eels in a socially intelligent manner. EEFO incorporates four foraging behaviors: interaction, rest, hunting, and migration. It simulates interaction behavior for better exploration and rest, hunting, and migration behaviors for better utilization. The energy factor used in EEFO improves the balance between exploration and exploitation. The algorithm demonstrates excellent performance in development and exploration, balancing growth and exploration, and avoiding local optima. Further details of the EEFO algorithm can be found in reference 29 .
The computational flowchart for MOO is shown in Fig. 5 . Before performing the objective optimization, the system is preliminarily designed, and the following parameters are determined: the lengths of the second-stage valve and the main valve, the stiffness of the reset spring K 1 and K 2 , the pre-opening x 01 and x 02 , and all parameters related to the HSV. The parameters to be optimized are W t1 , W t2 , D 1 , D 2 , ɑ 1 , and ɑ 2 . Based on the dynamic analysis of the system in “ Dynamic modeling ” section, the TSV system needs to ensure stability while achieving the most muscular flow capacity in the shortest response time. The optimization design problem of the TSV can be described as follows: under given constraints, select appropriate design variables x to optimize the objective function f ( x ) to its optimal value. The mathematical model is as follows:
Design flowchart in the present study.
Here, x = ( x 1 , x 2 , …, x 6 ) represents the design variables, f ( x ) represents the fitness function, and f ( s ) means the stability constraint condition, which should be equal to logic 1 when the system is stable. Considering the difficulty of machining, X min and X max are the lower and upper bounds of the design variables, respectively. The optimization has two fitness functions, representing the valve response time and flow rate. The response time can be calculated based on the adjusting time of the system’s step response. Among them, U ( s ) is the Laplace transform of the input signal, and H ( s ) is the system's transfer function. The system's flow rate can be calculated using Eq. ( 35 ).
The initial particle swarm is randomly generated and introduced into the main program loop based on the abovementioned constraints and optimization objectives. The particle swarm continues to search for the ideal values. The fitness values for the TSV's output flow rate and response time are calculated. The program can determine the maximum and minimum values of the fitness function and generate a series of Pareto solutions while comparing the distances between the solutions to ensure a certain distance is maintained between each solution and the others. Compared to the previous generations, the current generation's extreme values and positions will be updated. Then, the positions and velocities of the particles are updated, and the next generation of the particle swarm is generated. The loop continues until the specified number of generations. It is worth noting that the solver settings in this research are slightly different from the simulation model analysis in reference 29 . The solver in this research can simultaneously solve multiple objective functions and generate a Pareto solution set rather than solving the optimal value of a single objective function.
The parameter ranges are set as follows (mm): W t1 ∈ [2,10], W t2 ∈ [1,3], D 1 ∈ [20,100], D 2 ∈ [10,30], ɑ 1 ∈ [0.1,1] and ɑ 2 ∈ [0.1,1]. The number of iterations for EEFO is set to 500, and the population size is set to 500.
As shown in Fig. 6 b, as the iterative calculations progress, the overall trend of the number of obtained solutions shows an increase. From the trend of the solution quantity, it can be observed that at the beginning of the calculation, particles can quickly approach the targets and obtain a significant number of solutions within a relatively short number of iterations. As the iterations proceed, the particles gradually converge near the optimal solution until they finally converge to a specific region, and the number of solutions in the set no longer increases. This indicates that the population and the number of iterations achieve convergence in the calculation. At the same time, a decrease in the number of solutions in the set can be observed. This is because the solutions generated in this iteration dominate one or more solutions in the existing solution set, requiring the removal of dominated solutions. Therefore, a decrease in the total number of solutions in the iteration calculation is expected.
( a ) Solutions distribution; ( b ) changes in the number of solutions.
After the iterative calculations, all non-dominated solutions generated are stored, and completely identical non-dominated solutions are removed. This process results in the Pareto frontier (Fig. 6 a), plotted as a two-dimensional scatter plot. From the plot, it can be seen that the obtained solution set does not have any dominance relationships among them and is uniformly distributed along the Pareto frontier. The flow rate is positively correlated with the response time, meaning that as the flow rate increases, the system’s response time also increases. This is because obtaining a larger flow rate requires an increase in the valve spool lift and diameter, which leads to a longer valve action time and increased inertia.
Directly obtaining usable design parameters from the solutions obtained through the optimization algorithm is still impossible. Here, an Analytical Hierarchy Process (AHP) method is introduced. AHP is a simple method for making decisions on complex and ambiguous problems, especially those that are difficult to analyze quantitatively. It was proposed by Professor T. L. Saaty, an American operations researcher, in the early 1970s as a convenient, flexible, and practical multi-criteria decision-making method 34 . When determining the weights of factors that influence a specific criterion, these weights are often difficult to quantify. When there are many factors, decision-makers may inconsistently provide data that does not reflect their perceived importance due to incomplete consideration. The pairwise comparison method can establish pairwise comparison judgment matrices for the factors. The comparison judgment matrix for the TSV is shown in Table 3 . In the analysis, two additional design parameters, D 1 and D 2 , are included and given relatively low weights. We hope the TSV can have a smaller volume and weight while meeting the optimized performance criteria.
The values in Table 2 indicate the relative importance of the abscissa compared to the ordinate. An immense numerical value implies a greater significance of the parameter corresponding to the abscissa than the parameter corresponding to the ordinate. Therefore, the values on the diagonal are all 1. In this study, the valve's response time ( T ) is considered the most important performance indicator. Thus, the importance of T is three times that of the flow rate ( Q v ), five times that of the diameter of the main valve spool ( D 1 ), and seven times that of the diameter of the secondary valve spool ( D 2 ). Flow rate ( Q v ) is the second most important performance indicator. The influence of the diameter of the secondary valve spool ( D 2 ) on the valve volume is smaller than that of the main valve spool ( D 1 ). Hence, it is given a weight lower than that of the main valve spool.
First, the obtained design parameters are normalized, and then the processed data is multiplied by the weights obtained from the AHP method and summed. The results are shown in Fig. 7 . Taking into account the machining accuracy and the influence of the AHP analysis results, the final values of the design parameters are obtained as follows.
AHP overall weight percent for solutions on the Pareto front.
The optimized values of each design parameter are shown in Table 3 .
Response time
A prototype of the valve was manufactured to validate the performance of the proposed TSV, and an experimental setup was constructed based on the calculations from the EEFO algorithm, as shown in Fig. 8 .
( a ) Secondary stage and main spool, ( b ) TSV structure.
To test the response time of the HSV, the experimental setup shown in Fig. 9 was utilized for conducting response time tests. A step signal with a duration of 200 ms and a voltage of DC12V was applied as the input to the HSV. The change in current was observed to determine the switching time. The experimental results are shown in Fig. 9 . When the step control signal was applied, the electromagnetic coil exhibited inductance, impeding the rise of the current in the coil. The input current was insufficient to overcome the spring’s preloading force, resulting in an increase in current without any movement of the spool. This corresponds to the period from 50 ms to point T A in Fig. 9 , during which the HSV had no flow output. In the second phase, the spool began to release, and the reverse cutting of the magnetic flux lines induced a counter-electromotive force, causing the driving current to decrease. This corresponds to the segment from T A to T B in Fig. 9 . However, the armature quickly stopped moving, and the driving current continued to rise to its maximum value, corresponding to the portion from point T B to just before the current started to decrease. Similarly, the closing process of the solenoid valve can be divided into two parts: after the control voltage is reduced to zero, due to the combined effect of coil inductance, eddy currents, and residual magnetism in the armature, the armature is still attracted for a certain period, corresponding to the segment from 150 ms to T A′ in Fig. 9 . When the current decreases to a level where the armature cannot remain attracted, the armature is pushed out by the resetting spring until it reaches the maximum air gap, corresponding to the segment from T A′ to T B′ in the graph. The opening time of the ball valve is approximately 8 ms, while the closing time is around 9 ms.
Experiment result of dynamic performance.
To test the flow capacity of the TSV, a test setup was designed, as shown in Fig. 10 . The output pressure of the hydraulic pump was set to 31.5MPa, and after passing through the TSV, the fluid was connected to a flow meter. The flow meter provided flow signals to a data acquisition device, and fluid flow through the flow transmitter entered the relief valve, which simulated the load. Subsequently, the fluid flowed into the tank. The TSV was controlled by a step signal generated by a signal generator, and the oscilloscope collected the current output from the signal generator. A step signal with a duration of 1s and a voltage of DC12V was applied to the TSV system, and flow data was collected. The selection of experimental components is shown in Table 4 .
Simulation model and experimental setup for dynamic performance.
Due to the unmeasurable valve spool position, as shown in Fig. 11 , an AMESim2021.1 dynamic simulation model was developed, and the total flow output of the TSV is shown in Fig. 11 . In the simulation, the maximum flow rate was 235 L/min when the pressure drop was 5 MPa. Experimental results indicated an actual output flow rate of approximately 225 L/min, slightly lower than the simulation results. One possible reason for the reduced output flow rate is that the clearance between the poppet valve and the valve sleeve in the manufactured prototype is larger than the value set in the simulation, resulting in an increased flow to the pressure control chamber, causing a decrease in the spool lift. The spool position curve shows that the maximum lift of the main valve is approximately 0.55 mm, the opening of the secondary stage is around 0.21 mm, and the output of the HSV is in terms of flow, with a maximum flow rate of approximately 3.8 L/min. The trend of flow changes is similar to the motion trend of the solenoid coil. When the main valve opens, the startup delay and opening time are 6 ms and 15 ms, respectively. This means that the main valve can fully open within approximately 21 ms. The response time of the main valve is consistent with the response time obtained from the flow changes, indicating that the rapid response of the main valve is crucial for the total flow rate.
Experiment and simulation result of dynamic performance. ( a ) Flow rate; ( b ) spool displacement and flow rate of the HSV.
Coal mining machinery requires control valves continuously controlling pressure, flow, and direction. By adjusting the duty cycle of the PWM signal, proportional output of the HSV can be achieved. The proportional control of the TSV can be realized by utilizing the proportional characteristics of the HSV. By inputting a 2 kHz PWM signal to the system and adjusting the duty cycle, the flow output of the HSV can be controlled, thereby controlling the lift of the main spool. Due to the smaller dead zone of the HSV compared to the secondary and main stages, the separate control of the TSV can be achieved by utilizing the dead zone characteristics of each stage.
Figure 12 a illustrates the comparison between simulated and experimental flow outputs of the TSV as the system input signal duty cycle increases. Both exhibit similar trends: a significant increase in flow begins around a duty cycle of approximately 0.62 in simulation, whereas in experimentation, this increase starts around a duty cycle of 0.7. Beyond a duty cycle of 0.75, the flow outputs from simulation and experimentation align closely, consistent with the results shown in Fig. 10 .
Experiment and simulation result of proportional characteristic, ( a ) total flow rate; ( b ) spool displacement.
As shown in Fig. 12 b, by observing the displacement of the valve spool in the simulation, it is found that when the PWM ratio is less than 0.56, only the HSV can output flow, with a maximum flow rate of approximately 3 L/min. When the duty cycle increases, the secondary stage starts production flow, with a maximum flow rate of roughly 16 L/min. When the duty cycle reaches 0.62, all stages start to output flow, and the system’s output flow rate increases almost linearly with the increase in the duty cycle.
By comparing Fig. 12 a and b, it can be inferred that the difference in output flow is likely due to factors that interfere with the opening of the main valve spool in experiments, which start around a duty cycle of 0.7. These factors may include friction between the valve spool and sleeve, fluctuations in supply hydraulic pressure, and variations in component machining precision.
The performance of the directional valve ultimately hinges on the precision of the positioning of the controlled hydraulic cylinder. To validate its control accuracy, an experimental setup was constructed, as depicted in Fig. 13 a. Four sets of TSVs were employed to govern the extension and retraction of the hydraulic cylinder, with a displacement sensor installed inside the cylinder to record displacement data. Figure 13 b illustrates the arrangement of a hydraulic cylinder positioning test bench controlled by a 3/4 proportional valve, utilized for comparative purposes with the performance of the TSVs.
Experimental setup for cylinder control performance.
Open-loop control experiments were conducted on the position control system to verify the output characteristics of the TSV. As shown in Fig. 14 , the speed and position control of the hydraulic cylinder extension were achieved by adjusting the duty cycle of the input PWM signal. The overall adjustable range of the system’s speed exhibits three regions. When the duty cycle is less than 0.56, the extension speed is proportional to the signal ratio. The growth rate increases from a ratio of 0.56 to 0.62, reaching the maximum speed at a ratio of 0.62. The speed does not increase further with an increase in the duty cycle. The displacement curve of the hydraulic cylinder shows that the TSV allows for a rapid approach to the target and more precise displacement control by reducing the duty cycle when the target is close.
Experiment result of cylinder control.
The experimental results of the hydraulic cylinder displacement tracking controlled by the TSV are shown in Fig. 15 . A trapezoidal displacement signal was input, and the hydraulic cylinder could extend and retract following the displacement signal. Based on the system’s tracking error curve, the positioning accuracy of the hydraulic cylinder controlled by the TSV is ± 0.15 mm. In contrast, the displacement tracking accuracy is ± 3 mm, indicating that the system's positioning accuracy is higher than the tracking accuracy. The positioning error curve exhibits continuous fluctuations, and the higher tracking speed of the displacement causes the fluctuation of the error value during displacement tracking. The TSV can only control the flow by continuously switching, resulting in repeated increases and decreases in the error. The error value still fluctuates during positioning but with a smaller amplitude. This may be due to minor leakage in the hydraulic cylinder or deviations introduced by the displacement sensor during signal transmission, affecting the valve opening.
The experiment result of the following control.
To assess the control performance of the TSV, a PI control was applied using a sinusoidal position signal with control parameters k p = 8 and k i = 1. As a performance benchmark, a 3/4 proportional valve control system (depicted in Fig. 16 b) was employed, with the TSV system containing an emulsion and the proportional valve control system using hydraulic oil.
Experiment result of cylinder PI control.
The experimental results are illustrated in Fig. 16 , where (a) shows the actuator displacement of the proportional valve control system, (b) depicts the actuator displacement error of the proportional system, (c) displays the actuator displacement of the TSV control system, and (d) presents the actuator displacement error of the TSV control system. Due to the PI controller utilized, both systems exhibit some lag error in position tracking. The error profiles of the two control systems are similar, with the TSV system showing smoother error curves and a maximum error magnitude smaller than that of the proportional valve control system.
The entire TSV system is governed by the pilot HSV. Power control experiments were independently conducted on the pilot HSV, and the experimental findings are illustrated in Fig. 17 . Panel (a) displays the current within the HSV coil, with the control signal applied at 1 s. A dual-duty cycle control method was utilized to minimize control power consumption: the duty cycle was set to 1 at T = 1 s and reduced to 0.7 after 0.1 s. Panel (b) demonstrates the flow rate of the pilot HSV, achieving an output flow rate of approximately 2.8 L/min. The PWM signal operates at 12 V, indicating that the TSV can be driven by power less than 1.2W.
System control power test.
A novel low-power-driven three-stage poppet valve with an internal feedback structure is proposed in this study based on the displacement feedback principle. The structure utilizes a high-speed on/off valve as the pilot stage, enabling rapid response in high-pressure and water-based environments. It effectively avoids valve leakage and sensitivity to medium contamination, making it a potential replacement for traditional proportional directional valves in coal mining equipment.
Static analysis of the TSV structure was conducted to determine the lift of each stage under steady-state conditions. The state-space equations between the spools and the pressure control chamber system were established, and the step response was analyzed. A MOO design of critical parameters for the secondary and main stages was performed using the electric eel foraging algorithm. The optimization results showed that the solutions obtained by the optimization algorithm could approximate the Pareto front. The design parameters were selected using the analytic hierarchy process.
An experimental setup for the flow characteristics of the TSV was constructed. A prototype was manufactured based on the design parameters and subjected to experimental verification. The results showed that the output flow of the novel TSV was slightly smaller than the theoretical calculation, and the response time was somewhat longer than the theoretical calculation. Proportional control of the TSV was achieved by adjusting the ratio of the PWM control signal. Hydraulic cylinder control experiments were conducted, and a positioning accuracy of ± 0.15 mm was achieved. The TSV system shows smoother error curves and a maximum error magnitude smaller than that of the proportional valve control system.
All data generated or analysed during this study are included in this published article.
Control chamber section area ( A ci = π( R u ) 2 )
Throttle grooves flow area ( A ti = W ti ( x i + x 0i ))
Flow area at the operating point
Flow coefficient
Throttle grooves flow coefficient
Damping rate
Medium bulk modulus
Friction between the spool and valve sleeve
1 For the main valve, 2 for the secondary stage, 3 for the HSV
Part “ i ” flow pressure gain
Part “ i ” flow displacement gain
Main valve moving parts mass
Return spring stiffness
Supply pressure
Load pressure
Pressure control chamber pressure
Inlet pressure at the operating point
Outlet pressure at operating point
Inlet pressure increment at the operating point
Outlet pressure increment at the operating point
Flow rate increment at the operating point
Spool displacement
Throttle grooves pre-opening
Displacement increment at the operating point
Area ratio of upper and lower end faces of the spool ( α i = R L / R u )
Medium density
Trechera, P. et al. Comprehensive evaluation of potential coal mine dust emissions in an open-pit coal mine in Northwest China. Int. J. Coal Geol. 235 , 103677 (2021).
Article Google Scholar
Dai, J., Tang, J., Huang, S. & Wang, Y. Signal-based intelligent hydraulic fault diagnosis methods: Review and prospects. Chin. J. Mech. Eng. 32 (1), 75 (2019).
Ng, F., Harding, J. A. & Glass, J. Improving hydraulic excavator performance through in line hydraulic oil contamination monitoring. Mech. Syst. Signal Process. 83 , 176–193 (2017).
Article ADS Google Scholar
Shi, J. C., Ren, Y., Tang, H. S. & Xiang, J. W. Hydraulic directional valve fault diagnosis using a weighted adaptive fusion of multi-dimensional features of a multi-sensor. J. Zhejiang Univ.-Sci. A 23 (4), 257–271 (2022).
Kiama, N. & Ponchio, C. Photoelectrocatalytic reactor improvement towards oil-in-water emulsion degradation. J. Environ. Manag. 279 , 111568 (2021).
Li, D., Ma, X., Wang, S., Lu, Y. & Liu, Y. Failure analysis on the loose closure of the slipper ball-socket pair in a water hydraulic axial piston pump. Eng. Fail. Anal. 155 , 107718 (2024).
Fischer, F., Bady, D. & Schmitz, K. Leakage of metallic ball seat valves with anisotropic surfaces. Chem. Eng. Technol. 46 (1), 102–109 (2023).
Jing, L., Xi, Z., & Zhi, L. The experimental study of mine intrinsically safe electromagnet. In 2015 Joint International Mechanical, Electronic and Information Technology Conference (JIMET-15) , 967–970 (Atlantis Press, 2015).
Zhong, Q. et al. Dynamic performance and control accuracy of a novel proportional valve with a switching technology-controlled pilot stage. J. Zhejiang Univ.-Sci. A 23 (4), 272–285 (2022).
Qian, J. Y., Gao, Z. X., Wang, J. K. & Jin, Z. J. Experimental and numerical analysis of spring stiffness on flow and valve spool movement in pilot control globe valve. Int. J. Hydrogen Energy 42 (27), 17192–17201 (2017).
Liao, Y., Lian, Z., Feng, J., Yuan, H. & Zhao, R. Effects of multiple factors on water hammer induced by a large flow directional valve. Strojniski Vestnik/J. Mech. Eng. 64 (5), 329–338 (2018).
Google Scholar
Wang, H., Chen, Z., Huang, J., Quan, L. & Zhao, B. Development of high-speed on–off valves and their applications. Chin. J. Mech. Eng. 35 (1), 67 (2022).
Qiu, H. & Su, Q. Simulation research of hydraulic stepper drive technology based on high speed on/off valves and miniature plunger cylinders. Micromachines 12 (4), 438 (2021).
Article PubMed PubMed Central Google Scholar
Elsaed, E., Abdelaziz, M. & Mahmoud, N. A. Using a neural network to minimize pressure spikes for binary-coded digital flow control units. Int. J. Fluid Power 20 , 323–352 (2019).
Yuan, X. et al. Dynamic modeling method for an electro-hydraulic proportional valve coupled mechanical–electrical-electromagnetic-fluid subsystems. J. Magn. Magn. Mater. 587 , 171312 (2023).
Zhichuang, C. H. E. N., Shenghong, G. E., Jiang, Y., Cheng, W. & Yuchuan, Z. H. U. Refined modeling and experimental verification of a torque motor for an electro-hydraulic servo valve. Chin. J. Aeronaut. 36 , 302–317 (2023).
Wu, S., Zhao, X., Li, C., Jiao, Z. & Qu, F. Multiobjective optimization of a hollow plunger type solenoid for high speed on/off valve. IEEE Trans. Ind. Electron. 65 (4), 3115–3124 (2017).
Tamburrano, P. et al. Full simulation of a piezoelectric double nozzle flapper pilot valve coupled with a main stage spool valve. Energy Procedia 148 , 487–494 (2018).
Ling, M., He, X., Wu, M. & Cao, L. Dynamic design of a novel high-speed piezoelectric flow control valve based on compliant mechanism. IEEE/ASME Trans. Mechatron. 27 (6), 4942–4950 (2022).
Zhao, R., Liao, Y., Lian, Z., Li, R. & Guo, Y. Research on the performance of a novel electro-hydraulic proportional directional valve with position-feedback groove. Proc. Inst. Mech. Eng. Part E J. Process Mech. Eng. 235 (6), 1930–1944 (2021).
Madsen, E. L., Jørgensen, J. M., Nørgård, C., & Bech, M. M. Design optimization of moving magnet actuated valves for digital displacement machines. In Fluid Power Systems Technology , Vol. 58332, V001T01A026 (American Society of Mechanical Engineers, 2017).
Noergaard, C., Bech, M. M., Roemer, D. B., & Pedersen, H. C. Optimization of moving coil actuators for digital displacement machines. In Proceedings of the 8th Workshop on Digital Fluid Power (DFP16) , 39–54 (Tampere University of Technology, 2016).
Xu, B., Ding, R., Zhang, J. & Su, Q. Modeling and dynamic characteristics analysis on a three-stage fast-response and large-flow directional valve. Energy Convers. Manag. 79 , 187–199 (2014).
Li-mei, Z. H. A. O., Huai-chao, W. U., Lei, Z. H. A. O., Yun-xiang, L. O. N. G., Guo-qiao, L. I., & Shi-hao, T. A. N. G. Optimization of the high-speed on-off valve of an automatic transmission. In IOP Conference Series: Materials Science and Engineering , Vol. 339, No. 1, 012035 (IOP Publishing, 2018).
Qingtong, L., Fanglong, Y., Songlin, N., Ruidong, H. & Hui, J. Multi-objective optimization of high-speed on-off valve based on surrogate model for water hydraulic manipulators. Fusion Eng. Des. 173 , 112949 (2021).
Liu, P., Fan, L., Xu, D., Ma, X., & Song, E. Multi-objective optimization of high-speed solenoid valve based on response surface and genetic algorithm (No. 2015-01-1350). SAE Technical Paper (2015).
António, C. C. Memeplex-based memetic algorithm for the multi-objective optimal design of composite structures. Compos. Struct. 329 , 117789 (2023).
Zhong, C., Li, G. & Meng, Z. Beluga whale optimization: A novel nature-inspired metaheuristic algorithm. Knowl.-Based Syst. 251 , 109215 (2022).
Zhao, W. et al. Electric eel foraging optimization: A new bio-inspired optimizer for engineering applications. Expert Syst. Appl. 238 , 122200 (2024).
Andersson, B. R. On the Valvistor, a proportionally controlled seat valve. (Linkoping Studies in Science and Technology, Dissertations, 1984).
Wang, H., Wang, X., Huang, J. & Quan, L. Flow control for a two-stage proportional valve with hydraulic position feedback. Chin. J. Mech. Eng. 33 (1), 1–13 (2020).
Park, S. H. Development of a proportional poppet-type water hydraulic valve. Proc. Inst. Mech. Eng. C J. Mech. Eng. Sci. 223 (9), 2099–2107 (2009).
Tavakkoli-Moghaddam, R. et al. Multi-objective boxing match algorithm for multi-objective optimization problems. Expert Syst. Appl. 239 , 122394 (2024).
Saaty, T. L. Decision making with the analytic hierarchy process. Int. J. Serv. Sci. 1 (1), 83–98 (2008).
Download references
Authors and affiliations.
School of Mechanical and Electrical Engineering, China University of Mining & Technology, Beijing, 100083, China
Aixiang Ma, Heruizhi Xiao, Yue Hao, Xihao Yan & Sihai Zhao
You can also search for this author in PubMed Google Scholar
Aixiang Ma and Heruizhi Xiao authored the main manuscript text, Yue Hao prepared Figs. 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , Xihao Yan revised the manuscript format, and Sihai Zhao performed content review. All authors have reviewed the manuscript.
Correspondence to Sihai Zhao .
Competing interests.
The authors declare no competing interests.
Publisher's note.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .
Reprints and permissions
Cite this article.
Ma, A., Xiao, H., Hao, Y. et al. Multi-objective optimization design of low-power-driven, large-flux, and fast-response three-stage valve. Sci Rep 14 , 21575 (2024). https://doi.org/10.1038/s41598-024-70353-2
Download citation
Received : 06 May 2024
Accepted : 14 August 2024
Published : 16 September 2024
DOI : https://doi.org/10.1038/s41598-024-70353-2
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
Contaminated soil can reduce the stability of structures and infrastructure, endangering their structural integrity. Hence, this study tries to determine how oil pollution influences the torsion behavior of model steel piles at varied soil densities. This study is critical for determining piles' structural integrity and stability in oil-contaminated situations. A mixture of heavy motor oil and clean sand samples was prepared in proportions ranging from 0 to 8% of the dry weight of the soil. In this study, the relative densities (Dr), pile slenderness ratio (Lp/Dp), oil concentration (O.C%), and contaminated sand layer thickness (LC) all varied. The study also includes an examination of piles of combined load (vertical and torsional). Results revealed that the pre-applied torsion force reduced the pile's vertical bearing capabilities. Furthermore, at Dr = 30%, we determined that the maximum vertical load under amalgamated load at constant torsional load T = (1/3Tu, 2/3Tu, and Tu) in cases of (Lc/Lp) = 1 and (Lp/Dp) = 13.3 is 1.67, 3.4, and 5% less than piles under pure vertical load, respectively. This highlights the importance of considering torsional forces in pile design to guarantee precise load-bearing capabilities. Engineers should carefully assess both vertical and torsional loads to optimize the performance and stability of piles in various conditions.
This is a preview of subscription content, log in via an institution to check access.
Subscribe and save.
Price includes VAT (Russian Federation)
Instant access to the full article PDF.
Rent this article via DeepDyve
Institutional subscriptions
Not applicable.
Abdelhalim R et al (2020) Experimental and numerical studies of later ally loaded piles located near oil-contaminated sand slope. Eng Sci Technol Int J. https://doi.org/10.1016/j.jestch.2020.03.001
Article Google Scholar
ASTM (2010) ASTM D2487: standard practice for classification of soils for engineering purposes (Unified Soil Classification System). ASTM International, West Conshohocken
Google Scholar
Basack S, Nimbalkar S (2017) Numerical solution of single pile subjected to torsional cyclic load. Int J Geomech 17(8):04017016. https://doi.org/10.1061/(ASCE)GM.1943-5622.0000905
Bizaliele MM (1992) Torsional cyclic loading response of a single pile in sand. Dissertation, presented to Schriftenreihe des Instituts fur Grundbau, Ruhr-Univ., Bochum, Germany
Bolton MD et al (1999) Centrifuge cone penetration tests in sand. Geotechnique 49(4):543–552. https://doi.org/10.1680/geot.1999.49.4.543
Craig WH, Sabagh SK (1994) Stress-level effects in model tests on piles. Can Geotech J 31(1):28–41. https://doi.org/10.1139/t94-004
Dhatrak A, Sabale S, Thakare S (2017) Performance of single pile in oil contaminated sand under axial loading. Int J Eng Sci Res Technol. https://doi.org/10.5281/zenodo.817966
Dutt RN (1976) Torsional response of piles in sand. Ph.D. thesis, University of Houston, Houston, TX, USA, 7 115–123
Dutt RN, O’Neill MW (1983) Torsional behaviour of model piles in sand. Geotech Pract Offshore Eng 315:334
Franke E, Muth G (1985) Scale effect in 1g model tests on horizontally loaded piles. In: Proceedings of the 11th international conference on soil mechanics and foundation engineering, San Francisco, CA, USA, vol 2, pp 1011–1014
Garnier J (2001) Physical models in geotechnics: state of the art and recent advances, 1st Coulomb Lecture, Paris
Garnier J et al (2007) Catalogue of scaling laws and similitude questions in geotechnical centrifuge modeling. Int J Phys Model Geotech 7(3):1–23. https://doi.org/10.1680/ijpmg.2007.070301
Georgiadis M, Saflekou S (1990) Piles under axial and torsional loads. Comput Geotech 9(4):291–305. https://doi.org/10.1016/0266352X(90)90043-U
Guo P, Zou X (2018) Bearing capacity of a single pile in sand under combined vertical force-horizontal force-torque load. Chin J Rock Mech Eng 31(11):2593–2600 ( in Chinese )
Khamehchiyan M, Charkhabi AH, Tajik M (2007) Effects of crude oil contamination on geotechnical properties of clayey and sandy soils. Eng Geol 89:220–229
Laue J, Sonntag T (1998) Pile subjected to torsion. In: Proceedings of centrifuge ’98, Balkema, Rotterdam, the Netherlands, pp 187–192
Leblanc C, Houlsby GT, Byrne BW (2010) Response of stiff piles in sand to long-term cyclic lateral loading. Geotechnique 60(2):79–90. https://doi.org/10.1680/geot.7.00196
Li Q, Stuedlein AW (2018) Simulation of torsionally-loaded deep foundations considering state-dependent load transfer. J Geotech Geoenviron Eng 144(8):0401805. https://doi.org/10.1061/(ASCE)GT.1943-5606.0001930
Li Q, Stuedlein AW, Barbosa AR (2017) Torsional load transferred of drilled shaft foundations. J Geotech Geoenviron Eng 143(8):725–735. https://doi.org/10.1061/(ASCE)GT.1943-5606.0001701
McVay MC, Hu Z (2003) Determine optimum depths of drilled shafts subject to combined torsion and lateral loads from centrifuge test ing (using KB polymer slurry)-supplemental research. Report no. BC-545, WO #4. Florida Department of Transportation, Tal lahassee, FL
McVay MC, Herrera R, Hu Z (2003) Determine optimum depths of drilled shafts subject to combined torsion and lateral loads using centrifuge testing. Rep. contract no. BC-354, RPWO #9
McVay MC, Bloomquist D, Thiyyakkandi S (2014) Field testing of jet grouted piles and drilled shafts. Report BDK75-977-41, 91977. Florida Department of Transportation, Tallahassee
Mohammadi A et al (2018) Axial compressive bearing capacity of piles in oil-contaminated sandy soil using FCV. Mar Georesour Geotechnol. https://doi.org/10.1080/1064119X.2017.1414904
Nabil FI (2001) Axial load tests on bored piles and pile groups in cemented sands. J Geotech Geoenviron Eng 127(9):766–773
Nasr AMA (2009) Experimental and theoretical studies for the behavior of strip fppting on oil-contaminated sand. J Geotech Geoenviron Eng ASCE 135(12):1814–1822
Nasr A (2013) Uplift behavior of vertical piles embedded in oil contaminated sand. J Geotech Geoenviron Eng. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000739
Nasr A (2014) Experimental and theoretical studies of laterally loaded finned piles in sand. Can Geotech J 51:381–393
Nasr AAM, Azzam W (2017) Behaviour of eccentrically loaded strip footings resting on sand. Int J Phys Model Geotech. https://doi.org/10.1680/jphmg.16.00008
Nasr A, Krishna Rao S (2016) Behaviour of laterally loaded pile groups embedded in oil-contaminated sand. Géotechnique 66(1):58–70. https://doi.org/10.1680/jgeot.15.P.076
Nasr et al (2023) Exemplary studies on the torsional response of single pile in contaminated sand. J Eng Res 7(6): Article 3. https://digitalcommons.aaru.edu.jo/erjeng/vol7/iss6/3
Poulos HG (1975) Torsional response of piles. J Geotech Eng Div 101(10):1019–1035. https://doi.org/10.1061/AJGEB6.0000203
Poulos HG, Davis EH (1980) Pile foundation analysis and design. In: Rainbow bridge book. University of Sydney
Ramadan et al (2023) Model study of the geotechnical behavior of a single pile under torsional load in contaminated sand. Arab J Geosci 16:674. https://doi.org/10.1007/s12517-023-11793-4
Randolph MF (1983) Design consideration for offshore piles. Geotech Pract Offshore Eng. https://doi.org/10.1016/0148-9062(85)92244-2
Ratnaweera P, Meegoda JN (2006) Shear strength and stress strain behavior of contaminated soils. Geotech Test J 29(2):1–8
Reddy KM, Ayothiraman R (2015) Experimental studies on behaviour of single pile under combined uplift and lateral loading. J Geotech Geoenviron Eng 141(7):04015030
Sakr M, Nazir AK, Azzam WR, Sallam A (2016) Behavior of grouted single screw piles under inclined tensile loads in sand. Electron J Geotech Eng 21:571–591
Sakr M, Nazir AK, Azzam WR, Sallam A (2020) Model study of single pile with wings under uplift loads. Appl Ocean Res 100(102187):1–17
Stoll UW (1972) Torque shear test of cylindrical friction piles. J Civ Eng 42(4):63–65
Tawfiq K (2000) Drilled shafts under torsional loading conditions. Report no. B-9191, Florida Department of Transportation, Tallahassee
Thiyyakkandi S, McVay MC, Lai P, Herrera R (2016) Full-scale coupled torsion and lateral response of mast arm drilled shaft foundations Can. Geotech J 53(12):1928–1938. https://doi.org/10.1139/cgj-2016-0241
Wang WS, Jiang J, Fu ZC, Ou DX, Tang J (2020) Analysis of the bearing characteristics of single piles under vertical and torsional combined loads. J Hindawi. https://doi.org/10.1155/2021/8896673
Zhang LM, Kong LG (2006) Centrifuge modeling of torsional response of piles in sand. Can Geotech J 43(5):500–515. https://doi.org/10.1139/t06-020
Zou XI, Zhou CH, Wang YI (2021) Experimental studies on the behaviour of single pile under combined vertical-torsional loads in layered soil. J Appl Ocean Res. https://doi.org/10.1016/j.apor.2020.102457
Download references
The authors did not receive support from any organization for the submitted work. No funding was received to assist with the preparation of this manuscript. No funding was received for conducting this study and No funds, grants, or other support was received.
Authors and affiliations.
Structural Engineering Department, Faculty of Engineering, Tanta University, Tanta, Egypt
Nada Osama Ramadan
Faculty of Engineering, Tanta University, Tanta, Egypt
Ahmed Mohamed Nasr & Waseim Ragab Azzam
You can also search for this author in PubMed Google Scholar
Correspondence to Nada Osama Ramadan .
Competing interests.
The authors have no competing interests in advertising related to the content of this article.
This research did not contain any studies involving animal or human participants, nor did it take place on any private or protected areas. No specific permissions were required for corresponding locations. All authors agree to Participate.
Publisher's note.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
Reprints and permissions
Ramadan, N.O., Nasr, A.M. & Azzam, W.R. Experimental Study on the Behavior of Single Piles Under Combined Torsional and Vertical Loads in Contaminated Sand. Geotech Geol Eng (2024). https://doi.org/10.1007/s10706-024-02921-2
Download citation
Received : 29 May 2024
Accepted : 07 August 2024
Published : 16 September 2024
DOI : https://doi.org/10.1007/s10706-024-02921-2
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
IMAGES
VIDEO
COMMENTS
Table of contents. Step 1: Define your variables. Step 2: Write your hypothesis. Step 3: Design your experimental treatments. Step 4: Assign your subjects to treatment groups. Step 5: Measure your dependent variable. Other interesting articles. Frequently asked questions about experiments.
Additional measures to assess the course goals in improving critical thinking, experimental design and self-efficacy in experimental design will be implemented using validated tests [22, 43,44,45]. Further studies are also required to determine the long-term impact of this training on student performance in the laboratory and progression ...
Creative thinking shares several attributes with experimental design thinking which is similar with scientific process skills comprised by comparing, classifying, inferring and predicting, hypothesizing, defining, and controlling variables etc. (Dikici et al., 2018). Experimental design thinking and creative thinking may be related.
Critical thinking has been a longstanding goal of education, while design thinking has gradually emerged as a popular method for supporting entrepreneurship, innovation, and problem solving in modern business. While some scholars have posited that design thinking may support critical thinking, empirical research examining the relationship between these two modes of thinking is lacking because ...
Summary. Experiments allow researchers to determine cause and effect relations between variables. As such, they are a critical component in the advance of scientific psychology. In this chapter, we discuss the theory behind the design of good experiments, and provide a sample study for evaluation. We outline three important types of replication ...
1.01 Identify and create questions and hypotheses that can be answered through scientific investigations. 1.02 Develop appropriate experimental procedures for: Given questions. Student generated questions. 1.04 Analyze variables in scientific investigations: Identify dependent and independent. Use of a control.
Experimental design is a fundamental skill, essential for achieving success in science (Coil et al. 2010) and gaining fluency in scientific literacy and critical thinking in general (Brewer and Smith 2011).However, explicit instruction and practice in experimental design is often lacking in introductory biology lecture courses because of perceived time pressures, large class sizes, and the ...
Experimental design thinking and creative thinking have been increasingly recognized as the crucial basic thinking to promote scientific and technological innovation. ... whereas divergers are likely to be arts specialists (Hudson, 1966, 1968). Critical thinking is considered crucial factor to facilitating problem solving in design fields such ...
Critical-Thinking and Experimental Design Skills—Tools of Science. A significant number of students show interest in science in high school or before (often significantly before [Gopnik, 2012]), but do not pursue STEM studies at the tertiary level.
Prospective: we follow the individuals in the future to know who will develop the disease. Retrospective: we look to the past to know who developed the disease (e.g. using medical records) This design is the strongest among the observational studies. For example - to find out the relative risk of developing chronic obstructive pulmonary ...
Critical thinking is the process by which people make decisions about what to trust and what to do. Many undergraduate courses, such as those in biology and physics, include critical thinking as an important learning goal. ... To help students discern information about experimental design, we suggest that instructors consider providing them ...
Keywords: Graduate, Training, Critical Thinking, Experimental Design, Problem-Based Learning Introduction For over a decade there have been calls to reform biomedical graduate education. There are two main problems that led to these recommendations and therefore two different prescriptions to solve these problems.
Experiments in science are creative, iterative, & source critical thinking. We naturally experiment in art, science, and life. Here, we hone these skills through principles and practice. The principles are here, and the practice is in the form a lab manual entitled Designcraft for experiments.
Experimental group- the group in an experiment that receives the intervention; Posttest- a measurement taken after the intervention; Posttest-only control group design- a type of experimental design that uses random assignment, and an experimental and control group, but does not use a pretest; Pretest- a measurement taken prior to the intervention
Critical thinking is the process by which people make decisions about what to trust and what to do. Many undergraduate courses, such as those in biology and physics, include critical thinking as an important learning goal. ... To help students discern information about experimental design, we suggest that instructors consider providing them ...
Organizations and universities use Design Thinking (DT) to facilitate team innovation. However, few empirical DT studies have quantified it. Across two experiments, each based on semester-long DT projects to generate innovative solutions to sustainability problems, several different DT strategies were compared.
Second, our quasi-experimental design results provide evidence that the ICM has positively impacted students' critical thinking skills. Results have shown improvement in all critical thinking dimensions researched in this study: presenting evidence, explaining issues, articulating influence of context and assumptions, and systematically ...
Critical Thinking in Psychology - September 2006. The author of this quote is Edwin G. Boring (1886-1968), one of the great psychologists of the 20th century and author of A History of Experimental Psychology (1929; the quote comes from p. 659). Contemporary psychologists take "the psychology experiment" as a given, but it is actually a relatively recent cultural invention.
Fostering the development of students' critical thinking (CT) is regarded as an essential outcome of higher education. However, despite the large body of research on this topic, there has been little consensus on how educators best support the development of CT. ... This study employed a quasi-experimental design involving 147 first-year ...
Pre- and post-tests assessing students' proficiency in experimental design were used to measure student learning. Results: The analysis of the outcomes of the course suggests the training is effective in improving experimental design. The course was well received by the students as measured by student evaluations (Kirkpatrick Model Level 1).
A quasi-experimental design was adopted to explore the effects of the proposed learning approach on students' performance in art appreciation, digital painting creation, creative thinking tendency, and critical thinking awareness. A total of 48 students from two classes in a university in central Taiwan were recruited to participate in this ...
The aim of this study was to examine the effects of critical thinking (CT) teaching involving general, immersion, and mixed approaches on the CT skills and dispositions of high-school students. The study, which had three experimental groups (EG) and one control group, employed a pretest-posttest control-group quasi-experimental design.
Conclusion Teaching critical thinking about health is possible within the current Kenyan lower secondary school curriculum, but the learning resources will need to be designed for inclusion in and ...
Additional measures to assess the course goals in improving critical thinking, experimental design and self-efficacy in experimental design will be implemented using validated tests [22, 43-45]. Further studies are also required to determine the long-term impact of this training on student performance in the laboratory and progression towards ...
2.2. Evaluate and redesign project-based learning (PBL) course. We used the eight "critical questions" proposed by Ashford-Rowe et al. (Citation 2014) to assess the existing syllabus and redesign the authentic assessment.Although the course comprised most of the critical elements, several elements were added (Table 1): First, we provided more structure to the project by including a mid ...
A prototype was manufactured based on the obtained parameters and subjected to simulation and experimental verification. The results demonstrate that the valve has an opening time of 21 ms, a ...
The aim of this study was to examine the effects of critical thinking (CT) teaching involving general, immersion, and mixed approaches on the CT skills and dispositions of high-school students. The study, which had three experimental groups (EG) and one control group, employed a pretest-posttest control-group quasi-experimental design. CT teaching was initiated with a general approach in EG ...
Contaminated soil can reduce the stability of structures and infrastructure, endangering their structural integrity. Hence, this study tries to determine how oil pollution influences the torsion behavior of model steel piles at varied soil densities. This study is critical for determining piles' structural integrity and stability in oil-contaminated situations. A mixture of heavy motor oil and ...