Fig 1.
The Malate dehydrogenase CUREs Community (MCC) model.
Faculty can implement a malate dehydrogenase focused CURE for a short module within a laboratory course (mCURE) or a full semester (cCURE). All CUREs start with the hypothesis module. There are three experimental clusters: Protein Conformation, Cellular Biochemistry, and Mechanism.
Table 1.
Student demographics.
Table 2.
IRB agreement numbers and authorizing institutions.
Fig 2.
Students experience more CURE elements in cCUREs than mCUREs or control laboratories.
Students reported their experiences with three elements of a CURE at the end of the semester. Collaboration was scored using 1 (Never), 2 (One or two times), 3 (Monthly), and 4 (Weekly). Iteration and discovery/relevance elements were scored 1 (Strongly Disagree) to 7 (Strongly Agree). An ANOVA indicates that iteration and discovery/relevance are significantly different (cCURE > mCURE > control). For collaboration, cCURE is significantly higher than mCURE and control. See S1 Table for complete results. ** p < 0.0001, * p < 0.05.
Fig 3.
Faculty descriptions of their course elements are consistent with CURE and control classes.
The ten elements of a CURE are listed with the percentage of faculty that indicated each element was included in their course. Due to the small sample size, statistical analysis was not done. Descriptively, most cCURE faculty included all 10 elements, most mCURE faculty included 8 of the 10 elements, and most control faculty included 4 of the 10 elements.
Fig 4.
Faculty and student perception of learning activities in courses is consistent with expected CURE outcomes.
Students and faculty scored the same 25 learning activities in their classes. On the pretest, students scored each activity on their level of experience with that activity prior to the course (1 = no experience or feel inexperienced to 5 = extensive experience or mastered this element). On the posttest, students scored each activity on how much learning they gained from the activity in their class (1 = no gain or very small gain to 5 = very large gain). After the course, faculty scored how much emphasis they placed on the learning activity on a scale of 0 (not applicable) to 3 (major). An ANCOVA controlling for pretest scores and incorporating a Bonferroni correction for 25 tests (p < 0.002) indicated ten learning activities that were significantly different between conditions as reported by students. These ten learning activities are from three groups: 1) knowledge of experimental outcomes (4A – 4C), student involvement in research projects (4D – 4H), and presenting results (4I & 4J). Panel 4A shows detailed axes labels that are used in Panels 4B – 4J. Color coding is as follows: gray to black hue is the control group, peach to dark orange is the mCURE group, and light blue to dark blue is the cCURE group. The lightest hue is the student pre-assessment. The medium hue is the student post-assessment. The darkest hue is the faculty post-assessment. The specific topics in each category are: Knowledge of experimental outcomes—(A) lab/project where student knows outcome (mCURE & control > cCURE), (B) lab/project where instructor knows outcome (control > cCURE & mCURE), (C) lab/project where no one knows outcome (cCURE > mCURE > control); Student involvement in research projects—(D) student input into research (cCURE > mCURE & control, (E) responsible for part of the project (cCURE > control), (F) project entirely of student design (cCURE > mCURE & control), (G) read primary literature (cCURE > mCURE & control), (H) write research proposal (cCURE > mCURE > control); and Presenting results—(I) present poster (cCURE > mCURE & control), (J) present orally (cCURE > mCURE & control). There were no significant differences between conditions for the statements in the data handling and analysis group or course structure group. For all the statistics see S2 Table.
Fig 5.
cCURE students report greater course satisfaction, while URM students report greater course satisfaction than White/Asian students in all conditions.
Students reported their experiences with the course at the end of the semester. Statements were scored 1 (Strongly Disagree) to 7 (Strongly Agree). ANOVAs, with a Bonferroni correction for the four items tested (p < 0.0125), found the statement “This course was a good way of learning about the process of scientific research” rated significantly higher in the cCURE than the mCURE and control laboratories. See S10 Table for results. * p < 0.05.
Fig 6.
CURE students self-reported learning gains are not significantly different from control courses.
Students self-reported learning gains on 21 items at the end of the semester from 1 = “No gain or very small gain” to 5 = “Very large gain.” A national dataset of CURE responses [64] is included. ANOVAs found no significant differences in student self-reported learning between the three conditions. Thirteen items in the national CURE dataset (yellow dots) fall within the range of the control, mCURE, and cCURE courses in the MCC dataset. See S4 Table for complete results.
Fig 7.
cCURE students score significantly higher on experimental design regardless of URM status.
The experimental design assessment test (EDAT) [60] was given at the beginning and end of each semester. Responses were scored from 0 to 10 and an ANCOVA controlling for pretest scores was used to compare conditions. (A) Students in the cCUREs scored significantly higher than students in control courses, while students in the mCURE were not significantly different from either the control or cCURE students. (B) There was no significant difference in the performance of students by URM status within or between conditions. Statistical significance is represented as standard error (S.E.). See S11 Table for full results.
Fig 8.
Students in mCUREs have the lowest positive attitudes towards science while there are no differences between conditions for negative attitudes.
Students responded to (A) five statements that reflect a positive attitude and (B) six items that reflect a negative attitude towards science. Items were scored on a scale of 1 (Strongly Disagree) to 7 (Strongly Agree) on the pretest and 1 (Strongly Disagree) to 5 (Strongly Agree) on the posttest. An average composite score was calculated for each scale. An ANCOVA to compare between conditions while controlling for pretest scores was run. Students in the mCURE had significantly lower positive attitudes towards science than the cCURE and control groups. There were no differences between groups on negative attitudes. Statistical significance of student response data is represented as standard error (S.E.). See S6 Table for complete results.
Fig 9.
Students in cCUREs report the highest levels of STEM support, career interest, and importance.
On the posttest, students responded to four items each about STEM support, STEM career interest, and importance and a scale of 1 (Strongly Disagree) to 5 (Strongly Agree). An average of each scale was calculated, and the three conditions were compared with an ANOVA. Tukey post-hoc tests found cCURE students reported significantly more STEM support, career interest, and importance than mCURE and control students. See S8 Table for complete results. *p < 0.05.
Fig 10.
More students in the cCURE report plans to conduct undergraduate research while more URM students in the mCURE report plans to conduct research.
At the end of the semester, students reported on their plans to conduct research in the future and conditions were compared using χ2 analysis. Significantly more cCURE students planned to conduct research in the future than in the mCURE or control students. Student plans to conduct research by URM status and condition found more URM students in the mCURE planned to conduct research than White/Asian students in the mCURE. There was no significant difference by URM status in the cCURE or control conditions. See S12 Table for complete results.
Fig 11.
The CURE courses included more time in the course devoted to elements of a CURE. Student improvement in experimental design was higher in cCUREs than controls, while mCURE students were not significantly different than either. STEM interest and plans to conduct future research were higher in cCUREs compared to mCUREs and controls.