Knowledge graphs based on meta-analysis papers improve the quality of case formulation: a mixed methods design (Preprint)
Bibliographic record
Abstract
BACKGROUND: Case formulation (CF) is a core skill for therapists; however, creating high-quality CF requires considerable time. OBJECTIVE: This study demonstrates that providing a knowledge graph based on the meta-analytic literature can enhance CF quality. METHODS: Five groups were established, including four large language model (LLM) groups and one human expert group, each generating 25 CFs based on 25 vignettes. The Control group with Claude Sonnet 3.7 produced 25 CFs. The Personalization group served as the control group with additional personalization prompts. The Knowledge Graph group employed an LLM that generated 25 CFs, which was provided with a meta-analysis Knowledge Graph. Further incorporation of additional personalization prompts then comprised the Knowledge Graph with Personalization group. Finally, the Expert Group consisted of 25 CFs generated by a human expert. These 125 CFs in total were evaluated for general quality (i.e., correctness, completeness, feasibility, and consistency) using a 7-point scale and 18 essential elements with binary scores (0 or 1) by another human expert. The CFs were also qualitatively analyzed. RESULTS: The Knowledge Graph and Knowledge Graph with Personalization groups scored significantly higher than the control group in terms of correctness, completeness, and feasibility. The Expert group scored significantly higher on consistency than the machine-generated groups. Additionally, there was no significant difference in the feasibility scores between the Knowledge Graph, Knowledge Graph with Personalization, and expert groups. The qualitative evaluation suggested that human CFs narrow the text to content that is easy for the client to read, whereas machine CFs are more likely to include expressions that are unnatural to the client. CONCLUSIONS: These results indicate that providing knowledge graphs to novice therapists increases the correctness, completeness, and feasibility of CF. Providing experienced therapists with knowledge graphs is suggested to improve the quality of their CF and mental health services. CLINICALTRIAL: None.
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 | 0.131 | 0.274 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.004 |
| Bibliometrics | 0.005 | 0.004 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.003 | 0.004 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.010 | 0.001 |
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".