Hunter, Bradley

Document Sample
Hunter, Bradley Powered By Docstoc
					                                Hunter, Bradley
    Determinants of Weekly Review Session Attendance and the Effect of Review
               Sessions on Student Performance in Economics 110
                        Faculty Mentor: Eric Eide, Economics Department


Every semester, each of the fifteen teaching assistants for Economics 110 classes hold two
weekly review sessions, during which they solve practice test questions, discuss the implications
of lecture material, and help students solidify difficult concepts. These review sessions, while
provided at a relatively low cost by the Economics department (approximately $4500), also
constitute for an estimated 100 student-hours per week in just one of BYU’s six sections of Econ

Both teaching assistants and professors believe that regular attendance at these weekly review
sessions will improve a student’s understanding of Economics and his/her ultimate performance
in the class; however, the extent to which regular review attendance improves test scores is still
unclear. To the knowledge of the author, no previous attempts have been made to quantify the
effect of review sessions on Econ 110 performance.

If review sessions are indeed effective, as we expect them to be, then they are vastly
underutilized (less than 25%) by the student body. One way to extend the reach of these review
sessions is to identify the key determinants of review session attendance. Doing so enables Econ
110 professors and teaching assistants to identify which sub-groups are most likely to
underutilize the review sessions. This allows for the development of new recruiting strategies in
order to help more students succeed in their elementary study of Economics.

The two questions this research seeks to answer are “Who’s coming to the reviews, and who
isn’t coming?” and “How much better off is a student for coming to the weekly reviews?” In
order to answer these question, data was gathered on the students who choose to come and on
those who choose not to come to weekly review sessions and regressions were run on their
review session attendance as well as their grade in the class as a function of gender, GPA,
standardized test scores, age, year in school, and major to identify the most significant predictors
of both attendance and non-attendance of review sessions.


Attendance was taken by teaching assistants at bi-weekly review sessions for Professor Mark
Showalter’s Economics 110 class. The attendance data was linked with descriptive variables for
the students (age, sex, major, GPA, ACT score, year in school). Student data was only used if
the student signed a consent form which was provided to students during one of their Economics
110 classes. Data analysis was performed using STATA 9.0 (College Station, TX).

A total of 58 students filled out the survey. An initial regression of the variables against the
students’ grades in the class showed that each successive review raises a student’s grade in the
class by .09 grade points (.33 grade points is the equivalent of moving from a B to a B+, or
moving up one letter grade). The p-value was .01, r2 = 0.51. A subsequent regression included a
dummy variable for ever attending a review session. This second model resulted in a coefficient
of 0.57 grade points difference in students who attended, even once, compared with those who
never attended a review session (p value = .037, r2 = 0.51). The coefficient for additional
attendance was .03 and was not statistically significant. ACT score was correlated with better
performance in the class (0.17 grade point increase per one point increase in ACT score). For
attendance, class standing was a predictor of attendance. Seniors were mostly likely to attend,
followed by juniors, sophomores, and then freshmen. High performers were more likely to
attend, while ACT score was negatively correlated with attendance. All of these results were
statistically significant.


This study likely uncovers more about the character or work ethic of those attending the review
sessions than it does about the inherent effect of the review sessions themselves. The fact that a
dummy variable for attending a review session even once erased the benefit of attending an
additional review session leads to the conclusion that students who attend these review sessions
are substituting self-study for these review sessions. They likely would perform about the same
in the class if they did not attend and simply studied on their own. Because a student was more
likely to attend the review session if they had been at BYU for a longer period of time, it is likely
that over time, students learn that attending teaching assistant-led review sessions is a more
efficient use of time than self-study alone. For this reason, it appears that review sessions are a
good idea as they save study time for the students which they are able to then use in more
productive ways (studying for other classes, working, etc). If the economics department is
holding review sessions for this purpose, than they should be continued. If they are held to
increase the grades of the students, then they should not be held as they have little effect on a
student’s eventual grade in the class.