Predicting Crime Severity Among Individuals Deemed Not Criminally Responsible on Account of Mental Disorder (NCRMD): The Creation and Initial Validation of the Crime Severity Scale (CSS)
Bibliographic record
Abstract
Individuals deemed Not Criminally Responsible on Account of Mental Disorder (NCRMD) typically receive indeterminant treatment sentences that fall under the authority of a provincial Review Board (RB; Crocker et al., 2015). The RB makes decisions on the conditions and length of sentences based on the risk of severe harm the individual poses to the public. RB decisions are made under consultation with forensic clinicians (e.g., psychologists and/or psychiatrists) who use violence risk assessment measures to estimate the risk of harm the individual poses to the public. However, present measures are devised to estimate risk of any future violent behaviour, without regard for the severity of the behaviour. The present study sought to improve our understanding of how to predict future crime severity to be able to better inform RBs regarding risk of harm to the public, by addressing two major objectives and research questions (RQs): (1) clarify the most salient predictors of crime severity (RQ1) and (2) if salient predictors are found in RQ1, refine current risk assessment practices by devising and psychometrically testing a scale devised to predict future crime severity, among those deemed NCRMD (RQ2). A total of 315 archived NCRMD files were coded and analyzed in the present study. Moderate-to-high interrater reliability among the three raters who coded the files was revealed. To address RQ1, postdictive methods were employed, such that historical (e.g., history of offending), social (e.g., gang affiliation), psychological (e.g., previous diagnoses), and cognitive (e.g., previous IQ) variables that would have been reasonably available to a clinician at the time someone was deemed NCRMD, were used to predict the index offence(s) that led to the NCRMD verdict. In direct test of RQ1, data was analyzed in two ways: (1) by measuring crime severity dichotomously, considering the most egregious of violent offences (i.e., homicide, attempted homicide, assault with a weapon and/or assault leading to bodily harm) as severe—and all other offences as not severe; and (2) by adopting the standards and methods employed by Statistics Canada’s Crime Severity Weights, and measuring crime severity continuously, based on sentence length. The results of the analyses revealed that the answer to what predicts crime severity, substantially depended on how crime severity was operationally defined and measured. Given the discrepancy of definitions, and the result of only one factor being predictive of crime severity, RQ2 could not be meaningfully tested in this research. The results of the present study then suggested that while rigorous efforts were employed to clarify the predictors of crime severity, research is required to clarify how to operationally define crime severity, before further investigations on predicting crime severity should be employed. Future investigations of how to operationally define crime severity are imperative, given that the RB consults with forensic clinicians to understand whether an individual deemed NCRMD poses a risk of severe harm to the public. With unclear and discrepant operational definitions of crime severity, forensic clinicians’ ability to inform RBs as to the risk of severe harm an individual poses to the public are hampered. This is concerning given that the answer to whether an individual poses a risk of severe harm to the public impacts the length, conditions, and freedoms of these indefinite treatment sentences—ultimately impacting the lives, liberty and autonomy of individuals deemed NCRMD. It also impedes the ability for forensic clinicians to support the RBs ability to balance the liberty and autonomy of the individual, with the liberty and safety of the general public. Limitations and further legal, clinical, and research implications are also discussed.
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 | 0.003 | 0.015 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".