Predicting sex offender recidivism. I. Correcting for item overselection and accuracy overestimation in scale development. II. Sampling error-induced attenuation of predictive validity over base rate information.
Bibliographic record
Abstract
The authors demonstrate a statistical bootstrapping method for obtaining unbiased item selection and predictive validity estimates from a scale development sample, using data (N = 256) of Epperson et al. [2003 Minnesota Sex Offender Screening Tool-Revised (MnSOST-R) technical paper: Development, validation, and recommended risk level cut scores. Retrieved November 18, 2006 from Iowa State University Department of Psychology web site: http://www.psychology.iastate.edu/ approximately dle/mnsost_download.htm] from which the Minnesota Sex Offender Screening Tool-Revised (MnSOST-R) was developed. Validity (area under receiver operating characteristic curve) reported by Epperson et al. was .77 with 16 items selected. The present analysis yielded an asymptotically unbiased estimator AUC = .58. The present article also focused on the degree to which sampling error renders estimated cutting scores (appropriate to local [varying] recidivism base rates) nonoptimal, so that the long-run performance (measured by correct fraction, the total proportion of correct classifications) of these estimated cutting scores is poor, when they are applied to their parent populations (having assumed values for AUC and recidivism rate). This was investigated by Monte Carlo simulation over a range of AUC and recidivism rate values. Results indicate that, except for the AUC values higher than have ever been cross-validated, in combination with recidivism base rates severalfold higher than the literature average [Hanson and Morton-Bourgon, 2004, Predictors of sexual recidivism: An updated meta-analysis. (User report 2004-02.). Ottawa: Public Safety and Emergency Preparedness Canada], the user of an instrument similar in performance to the MnSOST-R cannot expect to achieve correct fraction performance notably in excess of what is achievable from knowing the population recidivism rate alone. The authors discuss the legal implications of their findings for procedural and substantive due process in relation to state sexually violent person commitment statutes and the Supreme Court's Kansas v. Hendricks decision regarding the constitutionality of such statutes.
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 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 | 0.001 | 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 | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".