IDEEAS Lab

Every reason people gave for leaving computer science showed up at every stage, from degree program to industry job

In 263 Reddit posts about leaving computer science, all six reasons the study identified appeared at all four points of departure, from students leaving a degree program to professionals leaving computing jobs, though the mix shifted by stage.

Every reason for leaving appeared at every stage Grid of six reasons for leaving computer science by four departure stages, from 263 Reddit posts. Every one of the 24 cells has posts in it. Degree abandonment (123 posts): job dissatisfaction 51 (41%), aspirations and motivation for another field 38 (31%), psychological and emotional factors 71 (58%), academic struggles 74 (60%), health and well-being 31 (25%), industry concerns 25 (20%). Post-degree academic shift (35 posts): job dissatisfaction 13 (37%), aspirations and motivation for another field 14 (40%), psychological and emotional factors 13 (37%), academic struggles 11 (31%), health and well-being 6 (17%), industry concerns 6 (17%). Post-degree industry shift (36 posts): job dissatisfaction 15 (42%), aspirations and motivation for another field 14 (39%), psychological and emotional factors 19 (53%), academic struggles 9 (25%), health and well-being 9 (25%), industry concerns 11 (31%). Industry exit (69 posts): job dissatisfaction 44 (64%), aspirations and motivation for another field 25 (36%), psychological and emotional factors 49 (71%), academic struggles 18 (26%), health and well-being 32 (46%), industry concerns 21 (30%). Left a degree program 123 posts Graduated, then another degree 35 posts Graduated, then other work 36 posts Left a computing job 69 posts Job dissatisfaction Job dissatisfaction, degree abandonment: 51 of 123 posts (41%) 41% 51 Job dissatisfaction, post-degree academic shift: 13 of 35 posts (37%) 37% 13 Job dissatisfaction, post-degree industry shift: 15 of 36 posts (42%) 42% 15 Job dissatisfaction, industry exit: 44 of 69 posts (64%) 64% 44 Aspirations for another field Aspirations and motivation for another field, degree abandonment: 38 of 123 posts (31%) 31% 38 Aspirations and motivation for another field, post-degree academic shift: 14 of 35 posts (40%) 40% 14 Aspirations and motivation for another field, post-degree industry shift: 14 of 36 posts (39%) 39% 14 Aspirations and motivation for another field, industry exit: 25 of 69 posts (36%) 36% 25 Psychological and emotional factors Psychological and emotional factors, degree abandonment: 71 of 123 posts (58%) 58% 71 Psychological and emotional factors, post-degree academic shift: 13 of 35 posts (37%) 37% 13 Psychological and emotional factors, post-degree industry shift: 19 of 36 posts (53%) 53% 19 Psychological and emotional factors, industry exit: 49 of 69 posts (71%) 71% 49 Academic struggles Academic struggles, degree abandonment: 74 of 123 posts (60%) 60% 74 Academic struggles, post-degree academic shift: 11 of 35 posts (31%) 31% 11 Academic struggles, post-degree industry shift: 9 of 36 posts (25%) 25% 9 Academic struggles, industry exit: 18 of 69 posts (26%) 26% 18 Health and well-being Health and well-being, degree abandonment: 31 of 123 posts (25%) 25% 31 Health and well-being, post-degree academic shift: 6 of 35 posts (17%) 17% 6 Health and well-being, post-degree industry shift: 9 of 36 posts (25%) 25% 9 Health and well-being, industry exit: 32 of 69 posts (46%) 46% 32 Industry concerns Industry concerns, degree abandonment: 25 of 123 posts (20%) 20% 25 Industry concerns, post-degree academic shift: 6 of 35 posts (17%) 17% 6 Industry concerns, post-degree industry shift: 11 of 36 posts (31%) 31% 11 Industry concerns, industry exit: 21 of 69 posts (30%) 30% 21 Percent of each stage's posts that gave the reason, with the number of posts. Each bar runs from 0 to 100%.

Each cell is the share of posts at that stage that gave the reason, with the number of posts beside it. No cell is empty, but the mix shifts: academic struggles lead among people leaving a degree program, and psychological and emotional factors lead among people leaving a computing job.

Redrawn from Ross and Katz (2025), Figure 4, using the counts printed in the paper and stage totals from its Figure 3. Original figure CC BY-NC 4.0.

People who leave computer science do not leave for reasons that belong to one stage. In 263 Reddit posts where people described leaving or wanting to leave the field, every one of six reasons appeared at all four points of departure: dropping a degree program, switching to another degree after finishing one, taking a non-computing job after graduating, and leaving a computing job.

The mix shifted by stage. Among people leaving a degree program, the most common reason was academic struggles (74 of 123 posts, 60 percent). Among people leaving industry jobs, psychological and emotional factors such as stress, burnout, and imposter feelings led (49 of 69 posts, 71 percent), followed by job dissatisfaction (44 posts) and health and well-being (32 posts). Academic struggles still came up in 18 posts from people leaving industry, as they reflected on earlier experiences with the field. When people weighed where to go next, what another career would be like was the most discussed factor at every stage.

Why it matters

Few studies of computing attrition look beyond the degree program. These posts suggest that what pushes students out, including failed courses, gaps in preparation, and weak support, keeps shaping decisions later in people’s careers. The authors argue that this makes classroom interventions relevant to the workforce too, and that support for self-efficacy should not stop at graduation.

How the lab did it

Amanda Ross, then a PhD student in the lab and now an assistant professor at Rose-Hulman Institute of Technology, led the study with Andrew Katz. They collected 10,384 posts from 25 subreddits that mentioned both computer science and leaving. An open-source language model (Qwen2.5-32b, run with temperature zero) helped filter these down to the 263 that were actually about someone leaving the field, then sorted each post into a departure stage. The lab’s GATOS workflow drafted a codebook from summaries of the posts. Ross then read the model’s themes against the original posts, revised them, and wrote the final codebook herself, interpreting it through social cognitive career theory.

Ross also checked the model against her own labels on samples of the data. For filtering, agreement was moderate (Cohen’s kappa 0.52 when the model gave a definite answer). For sorting posts into stages, it was substantial (kappa 0.72).

The paper is direct about what these posts cannot show. A post shows an intent to leave, not that the person left. Reddit gives no demographic information, may amplify the most negative voices, and allows no follow-up questions. The posts are not limited to a time window, so some problems they describe may since have changed. The model’s filtering was imperfect, so some relevant posts were likely dropped and some irrelevant ones kept, and the final interpretive step rested on a single researcher.

The question this opens is what happens next: whether the careers people move to actually fix what drove them out, and whether anyone comes back.

The paper

Ross, A. & Katz, A. (2025). Using generative AI for large-scale qualitative analysis of social media posts to understand why people leave computer science. Journal of Engineering Education. https://doi.org/10.1002/jee.70036

All findings