Effect of Project Get Going on Computer Skills of Incoming Graduate Students

Project Get Going was developed to help incoming graduate students enhance their computer skills. The project used a multimedia package that incorporated music and culturally appropriate vignettes and exercises in the hope that the incoming graduate students would have their computer skills enhanced.

The sample was divided by teams. Each team attended daily hour-long, lab-based instructions using the Project Get Going multimedia package for 25 days before they started their graduate studies.

All students were given pre and posttest to measure their level of comprehension and mastery of computer concepts and skills prior to, and upon completion of Project Get Going.

Analysis of the findings of the pre and posttest scores will be used to determine the validity of Project Get Going and help determine if the project is succeeding in enhancing graduate students’ computer skills (Raw data is presented in Appendix A). It intends to answer questions such as: Did participation in Project Get Going increase students’ comprehension and mastery of computer concepts and skills? Is there a significant relationship between pre and posttest scores for the entire group? For males? For females? For each of the teams? Is there a significant difference between pre and posttest scores for the entire group? For males? For females? For each of the team? Did the males and females perform differently?

Descriptive Analysis

A sample of 21 students was chosen to test the validity of this software multimedia package. As seen in Table 1, a total of 21 incoming graduate students were divided into three teams (red, blue, and green) with each team having the same number of participants (seven). Of the 21 students, 47.6% were female and 52.4% were male. The Red Team had five females (71% of the team) and two males (29% of them). The Blue Team had two females (29% of the team) and five males (71% of the team) and the Green Team had four males (57% of the team) and three females (43% of the team).

The Pretest. The maximum score attained on the pretest was 23, with a minimum score of nine. The range of the scores was 14 while the mode and median was 14. The mean for the pretest for the sample population was 14.52. The variance was 14.06 and the standard deviation was 3.75.

Males scored overall lower on the pretest than females. One of the two lowest scores attained on the test was by a male. Males also did not achieve as high scores as females on the pretest. The maximum score for males was 17. The overall high score was 23. There were three modes for the pretest for males making it multimode. The median for the males was 14 while the mean was 13.82. The variance was 7.36 and the standard deviation 2.71.

Females scored higher on the pretest than males. Though one female scored one of the two lowest scores, females had a better mean and median. Of the top five scores of the sample group, females scored the highest, the second highest, third and fourth highest as well. The females also had multiple modes of 14 and 18. The variance was 21.79 and the standard deviation 4.67.

When looking at the results in terms of teams we also find very interesting information. The Red Team had scores ranging from 10 to 21, and was amodal. The median score for the Red Team was 15 while the mean was 15.29 while the standard deviation was 3.73 and the variance 13.90.

The Blue Team had scores ranging from nine to twenty-three (the highest score attained in the pretest). It is also amodal, but the median was 14 and the mean 14.71 with a variance of 20.90 and a standard deviation of 4.57.

The Green Team is also amodal, with members attaining scores from 9 to 18. The median for the group was 14, while the mean was 13.57 with a variance of 10.29 and a standard deviation of 3.21.

Of the three teams, the Red Team had the highest mean and the second lowest standard deviation, making most of the teams’ scores to fall near their mean. The Blue Team had a second highest mean but had the largest standard deviation, which meant most of the teams’ scores were close to the mean.

Since overall females attained higher scores on the pretest than males, it would be logical to explore if the team with the most females would have the higher mean. In this case, the Red Team had the most females, so the evidence does support this theory. On the other side, the team with the most males did not have the lowest mean.

The Posttest. The maximum score attained on the posttest was 29, with a minimum score of 13 being acquired by only one individual. The range of the scores was 16 while it was multimode. The median score 20. The mean for the posttest for the students involved in the sample population of Project Get Going was 19.95 with a variance of 22.95. The standard deviation was 4.79.

The males’ scores improved on the posttest (a 4.55 gain in the mean). The mean, the median and the mode all increased from the pretest, the mode was 20, the median being 19 and the mean being 18.36. The variance was 10.65 and a standard deviation of 3.26. Both the minimum and maximum scores increased for the males, yet the range between the scores also increased, causing a larger standard deviation.

Scores for females on the posttest also improved over the pretest (a 6.41 gain in the mean) on the posttest increased to 22.5. It is multimode. The mean for the females on the posttest was 21.70 with a 32.68 variance and a 5.72 standard deviation. The minimum and maximum score also improved on the posttest, but the standard deviation also increased as well. Of the four posttest scores for the whole sample, females had the highest and the second highest.

Comparing the posttest for males to the posttest scores for females, males on average scored lower and with a standard deviation of 3.26, the male scores were closer together. Females on average scored higher, but due to a larger standard deviation, had scores spread over a larger range.

Analyzing the results of the posttest for the Red Team, members of that team had scores ranging from 17 to 29. The mode was 20. The median score for the Red Team was 22 while the mean was 22.71 and a standard deviation of 4.39.

The Blue Team had scores ranging from 13 to 29 (the highest score attained on the post-test). The mode and the median for the posttest were 16. The mean for the Blue Team on the posttest was 18.57 with a variance of 27.62 and a standard deviation of 5.26.

The Green Team had a posttest mode of 15, with members attaining scores from 14 to 24. The median for the group was 19, while the mean was 18.57. The variance was 16.29 and a standard deviation of 4.04.

Of the three teams the Red Team had the highest mean and the second highest standard deviation. The Blue Team had the same mean as the Green Team, but the Blue Team had the highest standard deviation, meaning that the scores were spread wider than the scores of the Green Team.

Looking at the range we can attest that 50% of the females scored under 14 points in the pretests while on the posttest none did. Showing this a great improvement in the scores for females. Males also improved, but none achieved 25 to 29 points, while females did. There was improvement in all teams, but no one in the Green Team scored from 25 to 29.
Correlation Analysis

In order to carry out correlation analysis a Pearson r test was utilized, (see Table 2) comparing the means of the pretest and the posttest for the whole group, females, males and teams. Both Pearson r and the scattergrams (Appendix B) show positive correlation between pre and posttest. The scattergrams are all linear. Very high correlation was found between pre and posttest in the group as a whole (.825), females (.903), Red Team (.862) and Blue Team (.910). High substantial correlation was found in males (.607) and the Green Team (.756). This means that the pretest scores predicted posttest scores. Participants that attained high scores on the pretest attained high scores in the posttest.

An outlier was identified among males. This participant attained the same scores on both pre and posttests; showing no improvement. This score lowered the correlation. Computation of correlation was carried out without the outlier and scores for correlation went from .607 to .71 and significance went from .048 to a .020 (a stronger statistical significance). Removing the outlier also affected the posttest mean for males, which went from 18.36 to 18.60. Once the outlier is removed, the correlation is indeed higher.

All correlations demonstrated to be statistically significant at a .05 alpha level. Comparing males and females, statistical significance of the correlation is stronger for females than males. Looking at the teams we could identify that the Blue Team showed the strongest statistical significance, the Red Team the second strongest and the Green Team with the lowest significance level.

Computation yielded a correlation of .825 for the sample between the pre and posttest. The coefficient of determination is the square of the correlation making the percent of Explained Variance 68%. This means that 68% of the variation in performance on the posttest (y axis) is associated with changes in the pretest ; 32% of the variation is due to other factors. Most correlation computation showed high level of explained variance, yet males showed a 37% of variance, which might indicate that 63% of the explained variance is due to external factors not related to the tests. It is also important to note that the outlier identified was a male. His score might be affecting the results. Considerations will be discussed in the section titled Overall Interpretation and Recommendations. Another factor that should be considered is the sample size. One outlier is making a big difference.

Leave a Reply