Gen Ed Project Analysis Spring '09 + Fall '09

Because there was so little data in the spring of '09, we combined it with the Fall '09 data to create a better baseline. This data set provides 2760 data points spread over the 19 GenEd objectives.

Multiple Views of Data

First, the data are presented, in Multiple Views of data, as average scores versus each GenEd objective, numbered 1-19. The various views contrast male/female, CTE/AS, ethnicity, campus, and GPA.

Finally, histograms of GPA and credit-hours earned are presented.

Time-series Evaluation of Scores

The following text refers to Gen Ed Project Results - Spring '09 + Fall '09:

The number of credit-hours earned for each of the students was associated with each score earned, then the student names were stripped out for anonymity by referring only to S-number in the data table. None of this identifiable information is present in any of the publically available data. Several statistics were developed for this data set.

For each of the 19 GenEd objectives, the (S-number,score) pairs were graphed as a single point. These scatter plots should show an upward sloping trend if students' level of mastery for that objective is increasing as they take more courses at the college. In some cases, this is clearly true, for example 1c, 2a and 2b. In the cases of 2b, 3a, 3b, 3d, 4b, 4c, and 4d, the combined data set shows substantially different results from the spring data set. This is mostly because of the small data size in the spring '09 data set.

A re-sampling (bootstrap) analysis with replacement was run 100,000 times on each GenEd objective to develop an overall probability of a positive slope. Bootstrap was used because the distribution is unknown, and probably non-normal.
Each resampling resutls in a scatter-graph of credit-hours-earned versus score, which is fitted to a line and the slope of that line is recorded. The number of positive slopes are counted and thus a probability of positive slope easily computed (simply the count divided by 100,000).