Not surprisingly, the questionnaire data showed that CAPE Summer students arrived at CAPE already having a very high interest in computer science. Therefore, it was not surprising that we did not find any increase in overall reported computer science interest after CAPE. (See Table 1) Given that students' reports of computer science interest were already high upon arrival at CAPE, there was little room for them to increase further. Thus, while we cannot say that CAPE increases interest in computer science in students, we can say that CAPE attracts students who are very excited to further their computer science knowledge.

Notwithstanding changes before and after CAPE Summer in the overall construct, however, we did see some significant changes in responses to specific items. For example, students were significantly more likely to agree that they "Plan to do computer activities for fun" after CAPE Summer than they had agreed before CAPE Summer that they "Do computer activities for fun." Similarly, after CAPE Summer, students were more likely to agree that they "plan to spend free time doing programming or making cool things with the computer" than they had agreed before CAPE that they "spend free time programming or making cool things with the computer." (See Table 1a) While the difference in these items might seem subtle, it is an important shift. Students were much more likely after CAPE to have enthusiasm (ratings of high agreement) for doing computer science for fun. They had much higher levels of agreement that they plan to do fun computer science activities after CAPE Summer than their agreement levels indicated they did so before coming to CAPE Summer.

The CAPE Summer experience appeared to have a statistically significant (what is considered a statistically "large effect") on students' self-efficacy. Students reported much higher levels of self-efficacy for computer science at the end of CAPE compared to the beginning of CAPE. CAPE students felt more confident in their ability to use the computer, work with programming, and get good grades in high school computer science courses after the CAPE Summer experience.

There was a slight but significant increase (a statistically "small effect") in CAPE Summer students' perceptions of computer science as an appealing profession from before to after the CAPE Summer experience. This scale includes items that examine how interesting students think computer scientists are, and how they rate computer science as an interesting profession.

Similar to the "Interest in Computer Science" construct, we did not see an overall increase in student ratings of computer science as being a creative subject after their CAPE Summer experience. However, as with the Interest construct, students came to CAPE Summer already demonstrating that they felt that computer science was creative. Thus, there was little room for increase in their responses on the post CAPE Summer questionnaire.

Although there were no significant changes for the overall creativity construct, two of the items did demonstrate significant changes. It is important to note that these two items were dropped from the overall measure of creativity because they didn't sufficiently correlate with the other items. However, as individual items, responses demonstrated potentially important shifts. More specifically, students were more likely to agree that "There is more to computer science than programming" after CAPE than they did prior to CAPE. Students were also statistically more likely to disagree that "Computer programs are only good if they help solve a problem" after CAPE Summer than they were before. (See Table 1a) Thus, despite little change in overall perceptions of computer science as a creative subject (likely due to already high ratings at the outset), student attitudes did increase slightly in these two very specific items.

As with perceptions of computer science as interesting and creative, CAPE students came into the CAPE Summer experience already having very high perceptions of CS as useful. Thus, it was not surprising that there was no demonstrated increase in their perceptions of computer science as being more useful overall after the CAPE experience.

Table 1: Summary Statistics for Before and After CAPE Analyses

Note: All items are on a 6 point scale, with 6 being the most positive. Analyses were conducted using a Paired-Samples T test and Cohen’s D, with no additional control variables. P values of .05 or less are widely considered to demonstrate statistically significant differences.

Table 1a. Additional Item-Specific Statistics for Before and After CAPE Analyses

*Note: Negatively-framed items have been reverse-coded. Higher means always indicate more positivity toward CS.

Table 2: Items by Construct

* Exploratory factor analysis findings resulted in the elimination of some items that did not fit well with each construct. Final scale Cronbach’s Alpha statistics generally ranged from above .6 to .9, with one scale (Creativity Pre-CAPE) having poorer internal consistency at .522, likely due in part to the small size (3 items) of the scale.