GENERAL CHEMISTRY WORKSHOP ATTENDANCE AND IMPROVED STUDENT PERFORMANCE

Many universities offer supplemental courses or workshops to help students improve their performance in general chemistry. Cornell University, the University of California, Davis (U. C. Davis), U. C. Irvine, and U. C. Riverside all have staff specialists that lead such workshops (1-3). Turner (1) found statistically significant differences between the final grades of students who regularly attended such workshops and the final grades of non-attending students. Deal (2), however, reported the higher grades found in his study not to be statistically significant. Clearly, variables other than workshop attendance may account for a student's final grade in a course, among them the student's mathematical ability and motivation.

As a test of the effectiveness of a second quarter general chemistry (Chemistry 1B) workshop offered at U. C. Davis, I collected data to determine which variables significantly influenced the students' final Chemistry 1B course grades. The variables assessed were the students' Mathematics Scholastic Aptitude Test (MSAT) scores, the students' Verbal Scholastic Aptitude Test (VSAT) scores, the number of workshop sessions the students attended, the students' perceptions of the average number of hours per week spent reading the chemistry text and notes, and the students' perceptions of the average number of hours per week spent solving chemistry problems. The averages of hours spent reading and solving problems served as a measure of motivation.

General Chemistry 1B

Approximately one half of the undergraduate students at U. C. Davis take General Chemistry 1A and 1B. General Chemistry 1B is a five-unit course given both in the winter and spring quarters. The prerequisite for Chemistry 1B is a passing grade in Chemistry 1A. Three different Chemistry 1 B sections with an enrollment of approximately 430 students each are offered in the winter quarter. These sections consist of three one hour lectures and a one hour discussion session per week. The discussion sessions are led by teaching assistants. In addition, Chemistry 1B has three hours of laboratory per week.

In 1989, Chemistry 1B topics included: liquids and solids, colligative properties of solutions, chemical equilibria, acid/base chemistry, solubility and precipitation reactions, oxidation-reduction reactions, entropy and Gibbs energy, electrochemistry, and qualitative analysis.

The Workshop

The workshop, titled General Assistance in Chemistry 1B, is offered by the Learning Skills Center and is open to all Chemistry 1B students. It is free, carries no credit, and attendance is optional. There are two sections of the same workshop and each meets twice a week for the entire quarter. The pace of the topics covered in the workshop sessions generally parallels the pace of course topics presented in lecture. At the beginning of the quarter the professor announces the workshop during lecture. The workshop is led by the same staff specialist for the entire quarter.

The first Chemistry 1B workshop session reviews basic study skills. Thereafter, the workshops are more course-specific. In a typical workshop session, the specialist reviews new concepts, answers questions on course homework, and gives test problems used in previous Chemistry 1B courses. The specialist discusses solutions after the students have had time to attempt the problems themselves. Handouts, given throughout the quarter, summarize important concepts, give algorithms, or offer hints and mnemonic devices. Two days before each examination, the specialist gives an evening practice test session for one hour; solutions are discussed in the following hour. The specialist also schedules regular office hours, both by appointment and on a drop-in basis, for most days of each week.

In the winter quarter of 1989, the combined average attendance of both workshop sessions was about 75 students. This is about 6% of those who received a final grade for Chemistry 1B. The attendance in evening practice test sessions was as high as 110 students, or about 9% of those who received a final grade in the course.

Workshop students represent a broad range of abilities. A high percentage of participating students earn A or B grades in the course, but students who have earned lower grades have never complained about these students attending. In their evaluations of the workshops, most students of all ability levels express appreciation for the reviews of concepts and especially for practice problems. Many students feel they have a better understanding after practicing on old test problems. As pointed out by Turner (1), the presence of students who earn A and B grades in the workshop removes any perception of the workshop as remedial and helps integrate all workshop students into the chemistry course.

The Statistical Analysis

I used a stepwise multiple regression analysis to determine if students' final Chemistry 1B course grades, the dependent variable, were significantly influenced by the following independent variables:

1) The students' MSAT scores.
2) The students' VSAT scores.
3) The number of workshop sessions the students attended.
4) The students' perceptions of the average number of hours per week spent reading the chemistry text and notes.
5) The students' perceptions of the average number of hours per week spent solving chemistry problems.

Careful telephone interviews using a written protocol yielded the average number of hours per week students spent in reading and solving problems.
The analysis eliminates students for whom any data was missing. It includes only students who attended either one of two different Chemistry 1B sections taught by the same professor. Also, the analysis includes only students who had actually attended at least one workshop session, and this before the first midterm. Thus, the study does include data for students who decided the workshop was not for them. It excludes, however, data for students who might have benefited from the workshop but attended no sessions.

Results and Discussion

A regression model was built using a stepwise multiple regression analysis of the variables for 86 students. The regression equation is:

Final Grade = 0.473 + 0.004 MSAT - 0.002 VSAT + 0.068 Workshop Hrs.

The predicted Final Grade is a numerical value from 0.00 (representing an F grade) to 4.00 (representing an A grade). The independent variables MSAT, VSAT, and Workshop Hrs are preceded by their slope values. The intercept is 0.473. The three digit grade derived from the equation is a point estimate and should not be viewed as indicating three significant figures. The true grade must be within the confidence limits as described below.

Residuals analysis indicated no problems with the model. The F test value for the regression equation is 16.68 and is significant at the 0.01% level. Thus, the equation gives a reasonable model to predict the final grade in the course.

The variables which fit the equation explain about 38% of entire variability of the final grades, as shown by the coefficient of multiple determination, R², which is 0.379. Although there is about 62% of variability left unexplained, the partial F test values (Table 1) take this variability into consideration and demonstrate that the variable of workshop hours is a significant and useful predictor of final course grade.

TABLE 1
PARTIAL F TEST AND SIGNIFICANCE LEVEL
VALUES FOR VARIABLES

Another way of viewing the unexplained variability is to look at confidence intervals. The 95% confidence intervals for the plot of the regression line using the averages for MSAT and VSAT scores are shown in Figure 1: Final Grade versus Workshop Hours

Stepwise regression shows that motivation as measured by student perceptions of average weekly hours spent reading and solving problems is not a significant predictor of final course grade.

All of the above results are virtually identical to the results obtained when the total points earned in the course by each workshop student are used as the dependent variable rather than the student's final letter grade (as represented by a numerical value from 0.00 to 4.00).

Example of using the model: The 0.068 slope value for workshop hours predicts that a class of students who attended 10 workshop sessions would increase their grade by 0.68 grade points [(10)(0.068) = 0.68] or about 0.7 of a letter grade value. The 0.004 slope value for the MSAT score predicts that a class of students with an average score 100 points higher than another class of students would receive a letter grade 0.4 grade points higher than that of the lower MSAT score students [(100)(0.004) = 0.4]. The -0.002 slope value for the VSAT score predicts that a class of students with an average score 100 points higher than another class of students would receive a letter grade 0.2 grade points lower than that of the lower VSAT score students [(100)(-0.002) = -0.2].

The surprising negative correlation of higher VSAT scores with a slightly lower grade may be explained by a relatively high incidence of workshop students for whom English is a second language. In this analysis, 29% of the students were not United States citizens and another 41% of the students were non-Caucasian. The majority of the workshop students were Asian. The average values for SAT scores, when grouped by citizenship and ethnicity, reveal that the majority of foreign and non-Caucasian students had MSAT scores 100 points or more aboves their VSAT scores (See Tables 2 and 3). It therefore seems likely that first language differences would explain the slightly negative correlation between VSAT scores and final course grade.

In Table 2, the higher final grades of foreign students, 92% of whom were Asian, when compared to the grades of U. S. non-Caucasian students, 77% of whom were Asian, may be the result of the foreign students solving more problems and their possessing cultural values which emphasize a greater dedication to learning (4). The lower MSAT scores of the U. S. Caucasian students make it more difficult to compare them to the two previous groups.

TABLE 2
AVERAGES OF VARIABLES FOR STUDENTS

TABLE 3
PERCENTAGE OF STUDENT GROUPS WITH MSAT SCORES
100 POINTS OR MORE ABOVE VSAT SCORES

Conclusions

The regression model presented here, after accounting for varying mathematical and verbal abilities, demonstrates that increased workshop attendance is associated with higher final course grades for workshop students. This grade improvement is independent of varying degrees of motivation as measured in this analysis.

These results are consistent with those of Turner (1) and St. John (3). This study, however, is the first analysis to:

1) select only for students who had attended at least one workshop session;
2) analyze those students' mathematical and verbal abilities as measured by their SAT scores; and,
3) attempt to use a measure of motivation.

It is difficult to say why Deal's study (2) did not yield similar results. Deal reported that grade improvement by workshop students was statistically insignificant. However, he had a small sample size and did not discuss the nature of the offered workshop. Nevertheless, he acknowledged that grade improvement might be significant if those workshops were more extensive. Unfortunately, he did not discuss the number of workshop sessions attended by each student.

Although student perceptions of the average number of hours per week spent solving problems and reading seems a reasonable measure of motivation, such perceptions were not significant predictors of final course grade. Even if the reported hours were more rigorously monitored, there may not necessarily be a linear relationship consistent with the variables which fit the equation. Furthermore, it may be impossible to separate out the motivation variable. In the final analysis, workshop attendance itself may be a measure of motivation. This would not preclude the actual benefits of attendance, as so often expressed by students in their evaluations of the workshops at U. C. Davis.

Acknowledgments

I wish to thank the following people of the University of California, Davis: Timothy Donnelly of the Department of Chemistry for providing the total points earned by each student; Peter Rock of the Department of Chemistry, Jerry Knutson of the Department of Agricultural Engineering and Jim Dykes of the Department of Statistics for critical discussions; Midge Clinton of the Learning Skills Center for helping collect and process the data; Virginia Martucci of the Learning Skills Center for helpful discussions; and Gary Perkins of the EOP/SAA Information Office for arranging for student advising counselors to help with telephone surveys.

Literature Cited

1. Turner, K. E. J. Chem. Educ. 1990,.67, 954-957.
2. Deal, W. J. J. Coll. Sci. Teach. 1984, 13, 154-156.
3. St. John, E. unpublished paper. Academic Support Services, U. C. ,Irvine. 1989,1-14.
4. Caplan, N.; Choy, M.H.; Whitmore, J. K. Sci. Amer. 1992, 266(2), 36-42.