Nearly 700 students from five institutions enrolled in seven introductory programming courses provided more than twentyfive hundred responses to the “for me, programming is ...” question. Overall, 48% of all responses were positive, 32% were negative, and the rest could not be interpreted as either positive or negative. Similar ratios were observed in five of the courses. In the remaining two courses, positive responses made up only 37% and 42% of the total, while the negative response rates were 49% and 40%, respectively.
The most common response indicated that many students believe that programming is fun or cool. Many students also describe programming as interesting or rewarding. But, the students also indicate that they think that learning to program is difficult or frustrating.
As noted above, in two of the courses we observed lower positive response rates. These courses were taught using media computation or gaming-focused approaches, which might lead us to conclude that these approaches are not effective. However, one of the other courses was also taught using the media computation approach, and the positive response rate in this course did not differ from that seen in the courses that were taught using traditional approaches. These results suggest that in some situations students may not find contextualized approaches to teaching programming as compelling as we might have hoped.
We also found significant differences in the response distributions between majors and non-majors. Non-majors tended to provide fewer positive and more negative comments than majors. They were less likely to view programming as fun or cool, but more likely to see it as something that was useful or interesting. They were also more likely to view it as a miserable or pointless activity, even though they did not indicate that it was hard or difficult as frequently as majors did. The media computation approach course with lower positive response rates was also comprised solely of non-majors.
Based on these findings, we have proposed and piloted a multiple-choice question that can be used to assess students’ affective beliefs about programming. We hope that the wide-scale use of this question in computer science education research will lead to better evaluation of the impact of proposed pedagogical interventions on students’ attitudes and beliefs about programming.