Towards a methodology for cluster searching to provide conceptual and contextual “richness” for systematic reviews of complex interventions: case study (CLUSTER)
Bibliographic record
Abstract
BACKGROUND: Systematic review methodologies can be harnessed to help researchers to understand and explain how complex interventions may work. Typically, when reviewing complex interventions, a review team will seek to understand the theories that underpin an intervention and the specific context for that intervention. A single published report from a research project does not typically contain this required level of detail. A review team may find it more useful to examine a "study cluster"; a group of related papers that explore and explain various features of a single project and thus supply necessary detail relating to theory and/or context.We sought to conduct a preliminary investigation, from a single case study review, of techniques required to identify a cluster of related research reports, to document the yield from such methods, and to outline a systematic methodology for cluster searching. METHODS: In a systematic review of community engagement we identified a relevant project - the Gay Men's Task Force. From a single "key pearl citation" we conducted a series of related searches to find contextually or theoretically proximate documents. We followed up Citations, traced Lead authors, identified Unpublished materials, searched Google Scholar, tracked Theories, undertook ancestry searching for Early examples and followed up Related projects (embodied in the CLUSTER mnemonic). RESULTS: Our structured, formalised procedure for cluster searching identified useful reports that are not typically identified from topic-based searches on bibliographic databases. Items previously rejected by an initial sift were subsequently found to inform our understanding of underpinning theory (for example Diffusion of Innovations Theory), context or both. Relevant material included book chapters, a Web-based process evaluation, and peer reviewed reports of projects sharing a common ancestry. We used these reports to understand the context for the intervention and to explore explanations for its relative lack of success. Additional data helped us to challenge simplistic assumptions on the homogeneity of the target population. CONCLUSIONS: A single case study suggests the potential utility of cluster searching, particularly for reviews that depend on an understanding of context, e.g. realist synthesis. The methodology is transparent, explicit and reproducible. There is no reason to believe that cluster searching is not generalizable to other review topics. Further research should examine the contribution of the methodology beyond improved yield, to the final synthesis and interpretation, possibly by utilizing qualitative sensitivity analysis.
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 | 0.550 | 0.711 |
| Meta-epidemiology (narrow) | 0.003 | 0.005 |
| Meta-epidemiology (broad) | 0.010 | 0.013 |
| Bibliometrics | 0.050 | 0.047 |
| Science and technology studies | 0.008 | 0.014 |
| Scholarly communication | 0.018 | 0.018 |
| Open science | 0.010 | 0.021 |
| Research integrity | 0.007 | 0.008 |
| Insufficient payload (model declined to judge) | 0.010 | 0.003 |
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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".