Learning Assistant Research Project
- 1. The Impact of LAHHH on Student Learning in a General Physics Course Liam Contino; Christopher Dang; Amandip Singh; Drew Turner Department of Physics at Oregon State University Learning Assistant Homework Help Hours (LAHHH) are three-hour sessions held weekly by Physics Learning Assistants. LAHHH is led by Learning Assistants, undergraduate students who successfully completed the general physics series and who now offer academic support to students enrolled in the PH 20x series. This research project set out to investigate the effectiveness of LAHHH as a resource for students enrolled in PH 201 at Oregon State University. The main goal of the LAHHH includes offering a time and space for peer-to-peer learning, providing classmates an opportunity to work collaboratively on homework problems, and to study for exams with the help and guidance of Learning Assistants (LAs). The study focused on evaluating the impact of LAHHH on student performance in PH 201. The Learning Catalytics clicker system was used to deliver survey questions regarding LAHHH utilization. GPA data was collected to separate the students into four GPA quartiles to account for ability-based variance. Finally, midterm scores were gathered. Collected data was then utilized to map out the relationship between LAHHH attendance and exam scores. Normalized gains were considered in order to observe the relationship between changes in LAHHH attendance and changes in performance between exam 1 and exam 2. Analysis of the data yielded significant observations in the lowest GPA quartile, as well as improvements in exam scores of attendees from exam 1 to exam 2. Analysis of normalized gains also showed a positive relationship between changes in LAHHH attendance and increase in exam scores. Research Question and Hypothesis Question: How does LAHHH attendance affect student performance on exams? Hypothesis: Students that attend LAHHH will have higher midterm scores than students who do not. Methods of Inquiry Studies conducted at University of Colorado Boulder, California State University Chico, and Loyola Marymount University support the idea that learning assistants and interactive engagement have a positive impact on student performance. This would support our hypothesis since LAHHH has a high amount of LA availability and LA- student interaction. Students were surveyed the following questions: How many LAHHH sessions did you attend before Midterm 1? How many LAHHH sessions did you attend after Midterm 1? Synthesis & Next Steps Students who participate in more interactive engagement opportunities, like LAHHH, tend to have larger gains than those who do not. The data suggests that students who struggle in their studies are more likely to seek out academic resources, such as LAHHH, in order to better understand the material. Selection bias in this study Midterm averages of study participants is higher than the average of the class as a whole. Students who were not present for class on the day the survey was conducted are not included. In a future study there should be a sign-in sheet at LAHHH sessions. Students in this study had to think back 8 weeks to remember how many sessions they went to, may be inaccurate. Data was not collected from students who withdrew from the course before week 8. Acknowledgements We want to thank the Oregon State University Physics Department for funding the Learning Assistant program. Special thanks to Kenneth Walsh, the supervisor of LAHHH and professor for the General Physics course used in this study, and also to Dennis Bennett for his contributions to the study design. References 1.Coletta, Vincent P., and Jeffrey A. Phillips. "Interpreting FCI scores: Normalized gain, preinstruction scores, and scientific reasoning ability." American Journal of Physics 73.12 (2005): 1172-1182. 2.Otero, Valerie, Steven Pollock, and Noah Finkelstein. "A physics department’s role in preparing physics teachers: The Colorado learning assistant model." American Journal of Physics 78.11 (2010): 1218-1224. 3.Van Dusen, Ben, Laurie Langdon, and Valerie Otero. "Learning Assistant Supported Student Outcomes (LASSO) study initial findings." arXiv preprint arXiv:1509.05358 (2015). Figure 1: LAHHH attendance breakdown by quartiles: Q1 represents students who were in the top quartile in terms of cumulative GPA prior to entering fall term. Analysis of the charts in Figure 1 shows that LAHHH attendance was similar for students from each quartile, and that each GPA quartile was represented fairly evenly throughout the term. Figure 2: Mean + SEM for quartile 1-4 comparing average midterm 1 scores for students who attended LAHHH prior to midterm 1 and students who did not attend. Q1: nNo = 45, nyes = 26, t-test probability, two-tailed, t-value = 0.36, p-value = 0.720 Q2: nNo = 48, nYes = 24, t-test probability, two-tailed, t-value = 1.39, p-value = 0.171 Q3: nNo = 52, nYes = 19, t-test probability, two-tailed, t-value = 1.05, p-value = 0.301 Q4: nNo = 53, nYes = 18, t-test probability, two-tailed, t-value = 2.83, p-value = 0.007 Figure 3: Mean + SEM for quartile 1-4 comparing average midterm 2 scores for students who attended LAHHH between midterm 1 and 2 and students who did not attend. Q1: nNo = 41, nyes = 30, t-test probability, two-tailed, t-value = 0.05, p-value = 0.954 Q2: nNo = 43, nYes = 29, t-test probability, two-tailed, t value = 0.612, p-value = 0.542 Q3: nNo = 54, nYes = 17, t-test probability, two-tailed, t value = 1.02, p-value = 0.318 Q4: nNo = 47, nYes = 24, t-test probability, two-tailed, t value = 0.429, p-value = 0.670 Based on Figure 2 and Figure 3, student GPA can be used as a good indicator of exam scores. The most drastic change between midterm 1 and midterm 2 occurred in quartile 4, which was analyzed more in-depth in Figure 4. Data & Findings Figure 4: Mean + SEM for Q4 students who attended LAHHH before midterm 1 but did not attend after midterm 1. Midterm 1 (adj) score = Midterm 2avg = 1.03*Midterm 1avg score n=7, t-test probability, two-tailed, matched pairs t-value = 3.713, p-value = 0.0099. Investigating the performance of Q4 students further provided some additional insight into LAHHH effects on student performance. Based on Figure 4, Q4 students who previously attended LAHHH before midterm 1, but stopped attending LAHHH after midterm 1, showed about an 8-10% drop on average in midterm 2 scores, another statistically significant find. These results seem to indicate that continued attendance aided in providing stability to exam scores from midterm 1 to midterm 2, whereas decreases in attendance were strongly associated with a significant drop in exam score averages Figure 5: Average normalized gains for students based on change to LAHHH attendance after midterm 1. n4: 8, n3: 11, n2: 21, n1: 40, n0:142, n-1: 38, n-2: 23 Normalized Gains (Coletta & Phillips) is a useful analytical tool to see if increased frequency (in this case, LAHHH attendance) correlates with higher academic gains (in this case, between exams 1 and 2). The normalized gains analysis quantified the gains in exam scores made by students who attended LAHHH hours, regardless of which GPA quartile they fell into. The positive trend is suggestive that those who visited LAHHH more frequently in between the two midterms in the class yielded the highest gains in midterm score. Data & Findings, Cont. Section 1: First, attendance was analyzed to determine if there were enough stu- dents from each quartile being represented at LAHHH sessions. It was important to deter- mine that our analysis was not being too heavily influenced by one quartile Section 2: Given a similar presence at LAHHH from each quartile, this section set out to determine if breaking up students into quartiles based on GPA was a valid method to analyze student performance on exams. This was necessary in order to determine if our matched pairs analysis for Q4 students (Figure 4) would provide valid results. Section 3: Based on Figure 2 and Figure 3 we noticed a sharp decline in midterm scores for quartile 4 students who were attending LAHHH. It was noted that most of the quartile 4 students who had performed well on the first exam stopped coming, prior to the second exam. If LAHHH was having no influence on these students we would expect quartile 4 students average score on midterm 2 to increase; however, we did not see an increase in midterm 2 scores for quartile 4 students that did not attend LAHHH. Section 4: Based on the results from analyzing Figure 4, we sought to further quantify the correlation present between changes in LAHHH attendances after the first exam and midterm 2 scores. Figure 5 expresses the usage of normalized gains to achieve quantification of such correlations. Data & Findings, Cont.Abstract