TMS and Working Memory and Programming
Bibliographic record
Abstract
The goal of this project is to better understand the role of working memory in writing programs (or program synthesis). Working memory can be defined as the small amount of information that can be held in an especially accessible state and used in cognitive tasks. Program synthesis refers to the ability of a programmer to construct a program that provably satisfies a given high-level formal specification. This is a key component in programming. By better understanding the cognitive processes of programming, we can better develop educational interventions and software programming support. A previously published study from our lab used MRI (magnetic resonance imaging) and TMS (Transcranial magnetic stimulation) to examine the relationship between spatial visualization and program comprehension tasks [1] leveraging previous correlative findings [2]. Our study aims to use a similar approach to examine the existence of a causal relationship between previous correlative findings of working-memory related brain regions with code system thesis. In this study, we propose to investigate the causal link of working memory and programming using only TMS for disrupting working-memory-associated regions Research Question: Is there a causal relationship between working memory load and programming ability? Significance: Confirming a causal link between working memory and program comprehension at the neurological level could spur changes in introductory programming education (e.g., including spatial visualization training, presenting material with a focus on spatial diagrams, etc.) and improve student success. Furthermore, as the second TMS study of programming (to the best our knowledge), this project would further the confidence of using TMS in computer science research. Methods: TMS will be used to temporarily disrupt brain areas that are associated with working memory. To localize these brain regions, we will use Montreal Neurological Institute (MNI) coordinates derived from previous studies who sampled more than 700 participants [3]. Participants will be brought in for two TMS sessions, where TMS is used to stimulate a working-memory brain region, and one where TMS is used to stimulate a non-working-memory associated region as an active control. In each session, participants will respond to a series of working memory and programming-related stimuli. We will check to see if there is a difference in accuracy, response time, and/or keystrokes between the sessions. Informally, if disrupting the X region of the brain interferes with the programming task but disrupting the Y region does not, we gain evidence that X activity is causally linked to performing the programming task. [1] Ahmad, H., Endres, M., Newman, K., Santiesteban, P., Shedden, E., & Weimer, W. Causal Relationships and Programming Outcomes: A Transcranial Magnetic Stimulation Experiment. In the International Conference on Software Engineering (ICSE): 2024 [2] Madeline Endres, Zachary Karas, Xiaosu Hu, Ioulia Kovelman, Westley Weimer: Relating Reading, Visualization, and Coding for New Programmers: A Neuroimaging Study: International Conference on Software Engineering (ICSE): 2021 [3] Schicktanz, N., Fastenrath, M., Milnik, A., Spalek, K., Auschra, B., Nyffeler, T., ... & Schwegler, K. (2015). Continuous theta burst stimulation over the left dorsolateral prefrontal cortex decreases medium load working memory performance in healthy humans. PloS one, 10(3), e0120640.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.008 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".