Bibliographic record
Abstract
In the investigative setting, police must often decide whether multiple crimes have been committed by a single offender.Using a variety of statistical techniques, studies have shown that it is possible to link serial crimes in a relatively accurate fashion using behavioural information (i.e., a process often referred to as behavioural linkage analysis; BLA).Despite this, practitioners often resist using these techniques, in a similar fashion to how clinical psychologists often resist actuarial techniques.In an attempt to develop an approach to BLA that may be better received by end users, this dissertation explored how classification trees (CTs) can be used to link serial crimes.Specifically, three variations of a CT approach were explored: a standard, single CT, an iterative CT (ICT), and the combination of multiple standard CTs and/or ICTs (i.e., a multiple model approach).Using separate samples of serial break and enters from Saint John, New Brunswick (N = 170) and serial sexual assaults from Quebec (N = 260), the ability of these approaches to link serial crimes were compared to one of the most commonly employed statistical approaches to BLA: main-effects logistic regression analysis.Generally, results revealed that all statistical approaches achieved high (and similar) levels of predictive accuracy; however, a number of potential advantages of a simple, standard CT approach were identified (e.g., transparency and ease-of-use).The findings reported in this dissertation have implications for BLA researchers (e.g., how behavioural domains are defined, how crime samples are selected, etc.) and police practitioners (e.g., the availability of a userfriendly statistical linking tool, the need for better data collection protocols, etc.).However, before a CT-based approach to BLA is implemented in practice, future research is required to address some of the limitations of the current research.
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 | 0.011 | 0.049 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.007 | 0.006 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 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".