Carnegie Mellon University

Eberly Center

Teaching Excellence & Educational Innovation

R&S Digest Header Image

Archive of Past Articles

Should students assessed as needing remedial mathematics take college-level quantitative courses instead? A randomized controlled trial.

Students needing to take remedial math classes often do not pass those classes and have high levels of college dropout. Consequently, some colleges have started to allow students to take college level classes without first taking remedial courses. The researchers randomly assigned students (n=717) in need of remedial classes to take either a traditional remedial algebra class, a remedial algebra class with workshops, or a college level statistics class with workshops. Students assigned to the college-level statistics class with workshops passed the class at a higher rate than students in remedial algebra with or without workshops (56% vs 39% and 45%). Students assigned to the college-level statistics class accumulated more college credits than those in the remedial class.

Logue, A. W., Watanabe-Rose, M., & Douglas, D. (2016). Educational Evaluation and Policy Analysis38(3), 578-598.
[link to article]

Examining the contributions of desirable difficulty and reminding to the spacing effect. 

Three experiments were conducted (N=134) to evaluate the desirable difficulties (DD) account of the mechanism underlying the benefits of spaced practice on learning. The DD account hypothesizes that retrieving information after a lag in time creates a slightly more difficult task that, in turn, leads to stronger encoding of the information.  Results from the three experiments, however, show that a longer lag time in between study events improves memory performance above and beyond the measured “desirable difficulty” induced by the lag itself, suggesting that the DD account of the spacing effect is either inaccurate or incomplete. That is, increasing the time lag between study events is effective, but this is not simply due to the increased difficulty of acquisition relative to that of a study event following a shorter (or zero) time lag.  The results of experiment 3 suggest that other variables, such as encoding variability, likely play a key role in explaining the spacing effect, although more research is clearly needed before coming to a firm conclusion.

Maddox, G. B., Pyc, M. A., Kauffman, Z. S., Gatewood, J. D., & Schonhoff, A. M. (2018). Memory & cognition, 1-13. , 1-19.
[link to article]

Brain drain: the mere presence of one’s own smartphone reduces available cognitive capacity

Undergraduate students in a laboratory setting were randomly divided into three conditions: those whose phones (turned off) remained 1) in a separate room outside the testing location, 2) in their bags near them, but out of sight, and 3) nearby and in sight. All students then completed a measure of working memory capacity and functional fluid intelligence. Students whose phones were most salient (condition 3) showed the lowest available cognitive capacity, while those whose phones remained in another room (condition 1) displayed the highest. All students self-reported that they did not believe the presence of their phone affected their performance. In a follow-up experiment, similar results held true whether the phones were on or off.

Ward, A. F., Duke, K., Gneezy, A., & Bos, M. W. (2017). Journal of the Association for Consumer Research2(2), 140-154.
[link to article]

A study exploring the impact of lecture capture availability and lecture capture usage on student attendance and attainment

Lecture capture has become more common and in some cases may even be retrofitted into classrooms. Despite perceived benefits of lecture capture for students (e.g. option to revisit lecture content), previous research has reported mixed findings about whether they are truly beneficial to student learning. In this study, researchers examined 2 cohorts of students in an introductory course: before lecture capture was available (n=161) and after lecture capture was implemented (n=160). Findings suggest that lecture capture availability had a negative impact on student attendance, that lecture capture usage associated with attendance was “negligible," and that, overall, lecture capture is an insufficient replacement for attendance for the vast majority students. 

Edwards, M. R., & Clinton, M. E. (2018). Higher Education, 1-19.
[link to article]

Self-Regulated Learning of Principle-Based Concepts: Do Students Prefer Worked Examples, Faded Examples, or Problem Solving?

Some research has shown the benefit of using worked examples for learners before engaging in independent problem-solving. This article explores what self-regulated learners choose to use when learning a new skill (in this case, solving probability problems). In Experiment 1 (N=102), participants chose between a worked example and independent problem solving over a series of 12 trials. Overall, participants favored independent problem solving unless their past trial was incorrectly answered. In Experiment 2 (N=182) and 3 (N=136), participants had a third option: partially worked examples. Participants chose this option more frequently than the fully worked examples. Researchers concluded that, although most learners were more likely to choose some form of worked examples after incorrect answers. they are less likely to choose as their first practice worked examples, despite it being better for their learning. When partially worked examples are available, however, they are more frequently chosen than fully worked examples when participants have that option. 

Foster, N. L., K. A. Rawson, J. Dunlosky. 2018. Learning and Instruction, vol. 55, pp. 124-138.
[link to article]

The Negative Consequences of Threat: A Functional Magnetic Resonance Imaging Investigation of the Neural Mechanisms Underlying Women's Underperformance in Math.

Brain imaging experiments were performed using FMRI on 28 female undergraduate students in a laboratory setting. All participants were asked to perform math tasks, however some were told that "research has shown gender differences in math ability and performance", presenting a stereotype threat condition. FMRI scans revealed that the women who did not hear this statement used parts of their brain associated with mathematical learning, while the women who heard the message reinforcing gender stereotypes about math did not employ these parts of the brain. Instead, the parts that were activated were associated with social and emotional processing. This study provides a complementary perspective to other stereotype threat literature about women's math performance.

Krendl, A.C.; Richeson, J.A.; Kelley, W.M.; Heatherton, T.F. (2008). Psychological Science, 19(2). 168-175.
[link to article]

Anatomy of STEM teaching in North American Universities

A large study of over 2000 classes by more than 500 faculty across North America was conducted to characterize the current state of STEM teaching using the COPUS protocol. Despite the large body of evidence to support the benefits of active learning, investigators found that didactic lecture is still the most common instructional style (74.9 +- 27.8% of the total time observed). Smaller classes and those in rooms with flexible seating were more likely than large, fixed-seat lectures to use interactive or student-centered approaches, although didactic lecture was sill the predominant teaching method. The authors recommend that institutions revise tenure, promotion, and merit-recognition policies to support evidence-based teaching practices, and state that research-based guidelines to measure effective teaching need to continue to be funded and developed.

Stains, M., et al. (2018). Science359 (6383), 1468-1470.
[link to article]

Realistic Details in Visualizations Require Color Cues to Foster Retention

Previous research has shown that students retained less information from realistic representations of biological structures. Instead, simplified drawings have been promoted to enhance learning. Skulmowski and Rey conducted a study to test whether color coding a realistic representation of a fictional bone would enhance the learning of the bone’s structure. Participants (N= 108) were randomly assigned and tested on one of 4 representations of a fictional bone: a simplified bone representation with labeled names of structure, a simplified bone representation with labeled names of structure and color coding of different areas of the bone, a realistic bone representation with labeled names of structure, a realistic bone representation with labeled names of structure and color coding of different areas of the bone. The researchers found that the realistic rendering of the bone structure enhanced by color coding led to better retention of information while students’ retention scores lowered when color coding was used on the simplistic bone representation. The authors concluded that realistic visualizations require appropriate visual aids in order to be effective but can be more effective for retention than simplified drawings. 

Skulmowski, A. and G. Rey (2018). Computers & Education vol 122, pp. 23-31
[link to article]

A Bayesian network meta-analysis to synthesize the influence of contexts of scaffolding use on cognitive outcomes in STEM education

This meta-analysis looks at N=56 studies on the impact of computer-based scaffolding on student learning. The researchers looked at within subjects gains in pre-post analysis of learning. The reviewed studies span from K-12 to graduate level students. The researchers found that computer scaffolding showed strong effect across student populations particularly for college and graduate students and STEM disciplines. 

Belland, Brian R., Andrew E. Walker, and Nam Ju Kim (2017). SReview of Educational Research, vol. 87, no. 6, pp. 1042-1081.
[link to article]

Stereotype threat effects on learning from a cognitively demanding mathematics lesson

A study of predominantly African-American 5th grade students (N=135) examined the effects of stereotype threat on initial student learning and knowledge formation using a math lesson.  Students assigned to the “threat” condition were asked to give their race at the beginning of the lesson - led to believe the study was about how students ‘like them’ learn best - and students in the control condition were asked to record the date instead of their race.  Results showed that students in the threat condition retained less information, enjoyed the lesson less, and reported lower levels of motivation than students in the control condition.  This effect was especially pronounced for students with higher levels of baseline executive functioning.  These findings suggest that stereotype threat may play a key role in the initial development of achievement gaps in education, and that its effects impact students on a deeper level beyond just test-taking performance.

Lyons, E. M., Simms, N., Begolli, K. N., & Richland, L. E. (2017). Cognitive science
[link to article]

Changes in syllabus tone affect warmth (but not competence) ratings of both male and female instructors

This study aimed to show correlations between the "friendliness" of a syllabus and student perceptions of the instructor's approachability and competence by gender. 150 undergraduate students were given either "friendly" or " not friendly" syllabi from either a female, male, or gender-unspecified instructor. The results showed that students perceived the "friendly" syllabus instructors as "more approachable, more caring, and more motivating, but not any more or less competent than those receiving the not friendly syllabus," regardless of gender. The article also includes a detailed list of example phrases from both the friendly and unfriendly syllabi, providing examples that could be used when teaching instructors about syllabus tone.

Denton, A.W.; Veloso, J. (2017). Social Psychology of Education (20). 1-15.
[link to article]

Student learning with permissive and restrictive cell phone policies: A classroom experiment

Two sections of a class taught by the same instructor had either a restrictive cell phone policy or a permissive cell phone policy (N=31). Students in the restrictive section were told they could be removed from class for use of their cell phones, and students in the permissive section were allowed to use cell phones during class. Researchers found no significant difference in the students' learning but found a statistically significant difference in the ratings of the instructor. The restrictive section gave higher ratings to the instructor than the permissive section. This research shows that student attitudes toward the instructor may not be negatively affected by a restrictive cell phone policy. 

Lancaster, Alexander (2018). International Journal for the Scholarship of Teaching and Learning, vol. 12, no. 1
[link to article]

Humor in the classroom: The effects of integrated humor on student learning 

Some previous research has suggested that humor could positively affect student motivation and attention. In this lab-based study, researchers asked whether integrated humor used to teach course concepts would affect students’ recall. In Study 1, 87 students were divided into two conditions. Students in condition 1 received a written lesson on self-efficacy that employed a humorous example. Students in condition 2 received the same lesson but a serious example. Students in condition 2 performed better on a multiple-choice recall test. Study 2 replicated the findings for study 1 using a new student population (more upper-level students), a new lesson (on cohesion in communication), a new learning environment (two different universities), and a new recall test (multiple choice + short answer). This study suggests that using integrated humor to directly teach core course concepts may negatively affect students’ learning of this content because humorous examples “potentially depress students’ ability to retain and transfer the ideas being presented.”

Bolkan, S., Griffin, D. J., & Goodboy, A. K. (2018). Communication Education, 1–21.
[link to article]

Reverse the routine: Problem solving before instruction improves conceptual knowledge in undergraduate physics

Exploratory learning in STEM courses at the university level is a relatively under-researched area. In three studies, researchers tested the effects of completing a collaborative (or solo) learning activity prior to being lectured about that topic (explore-first condition) versus completing the same activity after having the lecture (instruct-first condition). Across two semesters, students in three separate sections of a physics course (N = 362) were randomly assigned to receive either explore-first or instruct-first treatment in groups or individually. After completing both the activity and the lecture (in either order), all students completed a quiz to assess their learning as well as an interest/enjoyment questionnaire. Results showed that students in the explore-first condition tended to struggle more with the activity itself, but performed better on conceptual questions on the quiz than students in the instruct-first condition. There was no difference in performance on procedural quiz questions. Furthermore, students in the explore-first condition reported higher levels of interest and enjoyment. Results also showed that there was no added benefit of exploring in a group over exploring individually, suggesting that the benefits of exploratory learning are not simply driven by working collaboratively with others.  

Weaver, J. P., Chastain, R. J., DeCaro, D. A., & DeCaro, M. S. (2018). Contemporary Educational Psychology52, 36-47.
[link to article]

Peer Mentor Program for the General Chemistry Laboratory Designed To Improve Undergraduate STEM Retention. 

Several hundred undergraduate students enrolled in introductory chemistry labs at a large state university were assigned an upper-class peer mentor to each laboratory section. Instead of grading student work or focusing on course content, the mentor's responsibility was to check in with lab groups to talk about adjusting to college, offer advice on being a successful STEM student, and discuss other relevant topics such as undergraduate research opportunities or pursuing graduate school. After four years of the program, the peer-mentored cohort graduation rate was substantially higher (∼30%) than the non-peer laboratories group, and ∼22% higher than the general student population. 

Damkaci, F.; Braun, T.F.; Gublo, K. (2017). Journal of Chemical Education, ASAP. 
[link to article]

The Dark Side of Interpolated Testing: Frequent Switching Between Retrieval and Encoding Impairs New Learning

In a laboratory experiment using undergraduate participants (N=376), researchers sought to replicate and extend prior work that showed how retrieval practice can impair new learning (see Davis and Chan, 2015). Experimenters had participants switch between retrieval practice trials for a set of previously learned items and trials for the learning of new items at varying frequencies to determine the effects of making that switch has on learning the new content.  Results showed that as the frequency of switching between retrieval practice and new learning increased, the successful learning of the new information decreased. This research suggests that information retrieval and information encoding rely on separate cognitive processes that may compete for the same mental resources, and contributes to our understanding of how the use of interpolated testing could potentially hinder the learning of new information.

Davis, S. D., Chan, J. C., & Wilford, M. M. (2017). Journal of Applied Research in Memory and Cognition.
[link to article]