Multiobjective Genetic Programming Can Improve the Explanatory Capabilities of Mechanism-Based Models of Social Systems
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The generative approach to social science, in which agent-based simulations (or other complex systems models) are executed to reproduce a known social phenomenon, is an important tool for realist explanation. However, a generative model, when suitably calibrated and validated using empirical data, represents just one viable candidate set of entities and mechanisms. The model only partially addresses the needs of an abductive reasoning process - specifically it does not provide insight into other viable sets of entities or mechanisms, nor suggest which of these are fundamentally constitutive for the phenomenon to exist. In this paper, we propose a new model discovery framework that more fully captures the needs of realist explanation. The framework exploits the implicit ontology of an existing human-built generative model to propose and test a plurality of new candidate model structures. Genetic programming is used to automate this search process. A multi-objective approach is used, which enables multiple perspectives on the value of any particular generative model - such as goodness-of-fit, parsimony, and interpretability - to be represented simultaneously. We demonstrate this new framework using a complex systems modeling case study of change and stasis in societal alcohol use patterns in the US over the period 1980-2010. The framework is successful in identifying three competing explanations of these alcohol use patterns, using novel integrations of social role theory not previously considered by the human modeler. Practitioners in complex systems modeling should use model discovery to improve the explanatory utility of the generative approach to realist social science.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it