The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.
Comparing effect sizes in follow-up studies: ROC Area, Cohen's d, and r.
Abstract
In order to facilitate comparisons across follow-up studies that have used different measures of effect size, we provide a table of effect size equivalencies for the three most common measures: ROC area (AUC), Cohen's d, and r. We outline why AUC is the preferred measure of predictive or diagnostic accuracy in forensic psychology or psychiatry, and we urge researchers and practitioners to use numbers rather than verbal labels to characterize effect sizes.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The magnitude of the effects of these risk domains ranged from small (i.e., r= 0.099 for “low IQ/learning difficulties”) to large (i.e., r= 0.553 for “having a negative school attitude”) based on the criteria of Rice and Harris (2005) for interpreting effect sizes.”
From Risk Factors for School Absenteeism and Dropout: A Meta-Analytic Review (2019) · cites this paper as Rice and Harris (2005)
In the citing paper
“The AUC of 0.62 is also a moderate effect size [14], but is less than that for deadly violence (but this difference was not statistically significant).”
From The Domestic Abuse, Stalking and Harassment and Honour-Based Violence (DASH) Risk Assessment Instrument in Predicting Deadly or Persistent Domestic Abuse (2025) · cites this paper as [14]
In the citing paper
“The AUC of 0.67 for deadly violence is regarded as a “moderate” effect size [14].”
From The Domestic Abuse, Stalking and Harassment and Honour-Based Violence (DASH) Risk Assessment Instrument in Predicting Deadly or Persistent Domestic Abuse (2025) · cites this paper as [14]
In the citing paper
“AUC values of 0.56, 0.64, and 0.71 indicate small, medium, and large effects (Rice & Harris, 2005).”
From Strategies for monitoring mentoring relationship quality to predict early program dropout (2022) · cites this paper as Rice & Harris, 2005
Citations
Cites
Meta-Analytic Procedures for Social Research
Rosenthal, R. (1991) · doi:10.4135/9781412984997 · Cited by 4,635
Interpreting the magnitudes of correlation coefficients.
Hemphill, J. F. (2003) · American Psychologist · doi:10.1037/0003-066x.58.1.78 · Cited by 1,237
A common language effect size statistic
McGraw, K. O., & Wong, S. P. (1992) · Psychological Bulletin · doi:10.1037/0033-2909.111.2.361 · Cited by 934 · cited 1×
“AUC also has a plain language interpretation because it is conceptually and mathematically similar to the common language effect size or CL ( McGraw & Wong, 1992; see also Delaney & Vargha, 2002).”
Flagged first, then by citations
Research works printed but not linked: 13
- Berlin, F. S., Galbreath, N. W., Geary, B., & McGlone, G. (2003). The use of actuarials at civil commitment hearings to predict the likelihood of future sexual violence. Sexual Abuse: A Journal of Research and Treatment, 15, 377–382. · DOI
- Cohen, J. (1969). Statistical power analysis for the behavioral sciences. New York: Academic Press.
- Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
- Cohen, J. (1992). A power primer. Psychological Bulletin, 122, 155–159. · DOI · Paper page
- Delaney, H. D., & Vargha, A. (2002). Comparing several robust tests of stochastic equality with ordinally scaled variables and small to moderate sized samples. Psychological Methods, 7, 485–503. · DOI
- Harris, G. T., & Rice, M. E. (2003). Actuarial assessment of risk among sex offenders. In R. A. Prentky, E. S. Janus, & M. C. Seto (Eds.), Understanding and managing sexually coercive behavior, Vol. 989 (pp. 198–210). New York: Annals of the New York Academy of Sciences. · DOI
- Mossman, D. (1994). Assessing predictions of violence being accurate about accuracy. Journal of Consulting and Clinical Psychology, 62, 783–792. · DOI
- Pearson, E. S., & Hartley, H. O. (Eds.). (1954). Biometrika tables for statisticians, Vol. 1 (1st ed.). Cambridge: Cambridge University Press.
- Rice, M. E., & Harris, G. T. (1995). Violent recidivism: Assessing predictive validity. Journal of Consulting and Clinical Psychology, 63, 737–748. · DOI
- Rosenthal, R. (1990). How are we doing in soft psychology? American Psychologist, 45, 775–777. · DOI
- Rosenthal, R., & Rubin, D. B. (1982). A simple, general purpose display of magnitude of experimental effect. Journal of Educational Psychology, 74, 166–169. · DOI
- Swets, J. A. (1986). Indices of discrimination or diagnostic accuracy: Their ROCs and implied models. Psychological Bulletin, 99, 100–117. · DOI
- Swets, J. A., Dawes, R. M., & Monahan, J. (2000). Psychological science can improve diagnostic decisions. Psychological Science in the Public Interest: A Journal of the American Psychological Society, 1, 1–26. · DOI
Cited by
Sorted by citations
Items 801–900 of 1,001
Validation of the Juvenile Sexual Offense Recidivism Risk Assessment Tool-II.
Ralston, C. (2008) · doi:10.31274/rtd-180813-17009 · Cited by 10
Cano-Lozano, M. C., León, S. P., & Contreras, L. (2024) · Criminal Justice and Behavior · doi:10.1177/00938548241233967 · Cited by 9
“In the field of law and human behavior, it is recommended to interpret AUC values as 0.56 for small, 0.64 for medium, and 0.71 for large detection ability (Rice & Harris, 2005).”
BackgroundInfluentialRoaldset, J. O., Gustavsen, C. C., Lockertsen, Ø., Landheim, T., & Bjørkly, S. (2023) · Frontiers in Psychiatry · doi:10.3389/fpsyt.2023.1210871 · Cited by 9
Reindl, M., Lechner, I., Holzknecht, M., Tiller, C., Fink, P., Oberhollenzer, F., Mayr, A., Troger, F., Pamminger, M., Henninger, B., et al. (2023) · European heart journal. Acute cardiovascular care · doi:10.1093/ehjacc/zuad069 · Cited by 9
Dynamic Risk Scales Degrade Over Time: Evidence for Reassessments
Lee, S. C., Babchishin, K., Mularczyk, K. P., & Hanson, R. K. (2023) · Assessment (Odessa, Fla.) · doi:10.1177/10731911231177227 · Cited by 9
“Given its similarity to the area under the curve (AUC), similar interpretations of effect size magnitudes are applicable (i.e., the effect of .56 is small, .64 is moderate, and .71 is large; Helmus & Babchishin, 2017; Rice & Harris, 2005).”
BackgroundLiu, M., Yin, D., Wang, Y., Wang, W., Fu, T.-T., Duan, Y.-T., Hu, M., & Huang, K. (2023) · International Journal of COPD · doi:10.2147/copd.s403603 · Cited by 9
“6 “poor”.(15) Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), correct classification ratio, positive likelihood ratio (+LR) and negative likelihood ratio (-LR) were established for the different questionnaires and cut-off values.”
BackgroundGondek, D., Howe, L. D., Gilbert, R. E., Feder, G. S., Howarth, E., Deighton, J., & Lacey, R. E. (2023) · JAMA Network Open · doi:10.1001/jamanetworkopen.2023.1175 · Cited by 11
Quantified Canine: Inferring Dog Personality From Wearables
Meegahapola, L., Constantinides, M., Radivojevic, Z., Li, H., Quercia, D., & Eggleston, M. S. (2023) · International Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581088 · Cited by 9
“We report the top five features for each personality trait with: a) the highest t-statistic9 [51], and b) the highest Cohen’s-d10 [75].”
BackgroundEsser, E., Zimmermann, J., Storp, J., Eter, N., & Mihailovic, N. (2022) · Graefe's Archive for Clinical and Experimental Ophthalmology · doi:10.1007/s00417-022-05946-6 · Cited by 9
Janković, M., Boxtel, G. J. M. V., & Bogaerts, S. (2022) · International Journal of Offender Therapy and Comparative Criminology · doi:10.1177/0306624x221133013 · Cited by 10
“To interpret the magnitude of AUC values, effect sizes of 0.56, 0.64, and 0.71 were used as thresholds for low, moderate, and high, respectively, with effect sizes 0.56 considered to be negligible (Rice & Harris, 2005).”
MethodInfluentialLiu, W.-P., Li, P., Zhan, X., Qu, L., Xiong, T., Hou, F.-F., Wang, J.-K., Wei, N., & Liu, F.-Q. (2022) · Frontiers in Genetics · doi:10.3389/fgene.2022.870222 · Cited by 9
“The distinction was considered good when the area under the curve (AUC) value was between 0.8 and 0.9 and exceptional when the AUC value was greater than 0.9 (Rice and Harris, 2005).”
BackgroundInfluentialGyorda, J. A., Nemesure, M., Price, G. D., & Jacobson, N. (2022) · Journal of Affective Disorders · doi:10.2139/ssrn.4019107 · Cited by 9
“Similar to the corresponding continuous outcomes, for both of these outcomes, the models performed poorly on both the validation and test sets (e.g., AUC values indicate a small predictive relationship; Rice and Harris, 2005), and feature importance is not reported as a result.”
MethodResultInfluentialHenderson, C. M., & Fantuzzo, J. W. (2022) · Journal of Education for Students Placed at Risk (JESPAR) · doi:10.1080/10824669.2022.2065636 · Cited by 18
“To facilitate understanding of the magnitude of the AUC across fields, Rice and Harris (2005) translated AUC into a more readily understood measure of effect size (Cohen’s d).”
Methodde Jong, Y., Boon, A. E., Gouw, D., van der Gaag, M., & Mulder, C. L. (2022) · Child and Adolescent Psychiatry and Mental Health · doi:10.1186/s13034-022-00459-w · Cited by 10
“70 are considered clinically useful [36].”
Morris, C. S., Ingram, P. B., & Armistead-Jehle, P. (2022) · Military Psychology · doi:10.1080/08995605.2021.2013059 · Cited by 9
Shaik, A. S., & Dhir, S. (2021) · Journal of Indian Business Research · doi:10.1108/jibr-09-2020-0309 · Cited by 9
Antoniou, M., Estival, D., Lam‐Cassettari, C., Li, W., Dwyer, A., & Neto, A. D. A. (2022) · JMIR Formative Research · doi:10.2196/33036 · Cited by 17
Prioritizing Police Investigations of Intimate Partner Violence Using Actuarial Risk Assessment
Henning, K., Campbell, C. M., Stewart, G., & Johnson, J. M. (2021) · Journal of Police and Criminal Psychology · doi:10.1007/s11896-021-09466-7 · Cited by 15
“The full 10-item NIBRS-DV assessment evidenced a “large” association with DV recidivism (AUC = 0.72; Rice and Harris 2005).”
BackgroundMethodInfluentialAzizian, A., Olver, M., Rokop, J., & D'Orazio, D. M. (2021) · Sexual abuse. A Journal of Research and Treatment · doi:10.1177/10790632211019726 · Cited by 9
“…with values that range from 0 to 1.0, representing the probability that a randomly selected recidivist for a given outcome has a higher score than a randomly selected nonrecidivist; values of .56, .64, and .71 correspond to small, medium, and large effect sizes, respectively (Rice & Harris, 2005).”
BackgroundInfluentialAndersen, L., Saal, W., Joska, J., Safren, S., Bantjes, J., O'Cleirigh, C., Witten, J., Lee, J. S., & Kagee, A. (2021) · Aids and Behavior · doi:10.1007/s10461-021-03305-w · Cited by 9
Critical review of the use of the Rorschach in European courts
Areh, I., Verkampt, F., & Allan, A. (2021) · Psychiatry, Psychology and Law · doi:10.1080/13218719.2021.1894260 · Cited by 9
Paquette, S., & Fortin, F. (2021) · International Journal of Offender Therapy and Comparative Criminology · doi:10.1177/0306624x211013523 · Cited by 17
“AUCs of .54, .63, and .71 are considered small, moderate, and large effect sizes, respectively (Rice & Harris, 2005).”
BackgroundPREDICTIVE MODELLING AND ANALYTICS FOR DIABETES USING A MACHINE LEARNING APPROACH
Mishra, P., Sharma, A., & Badholia, A. (2021) · Information Technology in Industry · doi:10.17762/itii.v9i1.121 · Cited by 9
Schafers, C., Olver, M., & Wormith, J. (2020) · Criminal Justice and Behavior · doi:10.1177/0093854820980498 · Cited by 9
“With values of .50 representing chancelevel accuracy, AUC values of .56, .64, and .71 correspond to small, medium, and large effects, respectively (Rice & Harris, 2005).”
BackgroundThapa, S., Brown, S. L., & Skilling, T. A. (2020) · Criminal Justice and Behavior · doi:10.1177/0093854820977577 · Cited by 18
“CRiminal aTTiTudes, self-esTeem, and ReCidivism ResulTs BY GendeR As Table 2 illustrates, the attitudinal measures (MCAA: ATV and PIDS) predicted recidivism in the moderate range (Rice & Harris, 2005) for males and females (AUCs ranged between .60 and .71).”
BackgroundBiedermann, L., & Rettenberger, M. (2020) · Monatsschrift für Kriminologie und Strafrechtsreform · doi:10.1515/mks-2020-2059 · Cited by 9
“Die ermittelten AUC-Werte dieser beiden Instrumente sind nach internationalen Standards (Rice & Harris 2005) als hoch einzustufen, was für einen Einsatz derselben vor der Lockerungsentscheidung in der sozialtherapeutischen Anstalt sprechen würde.”
BackgroundPredictors of Residential Care Admission in Community-Dwelling Older People With Dementia
Jamieson, H. A., Abey‐Nesbit, R., Nishtala, P. S., Allore, H. G., Han, L., Deely, J. M., & Pickering, J. W. (2020) · Journal of the American Medical Directors Association · doi:10.1016/j.jamda.2020.04.021 · Cited by 8
Indirect and Neuropsychological Indicators of Pedophilia
Rosburg, T., Pflueger, M. O., Mokros, A., Boillat, C., Deuring, G., Spielmann, T., & Graf, M. (2020) · Sexual abuse. A Journal of Research and Treatment · doi:10.1177/1079063220931822 · Cited by 9
“AUC >0.8 indicate large effect sizes (Rice & Harris, 2005).”
BackgroundInfluentialBallesteros, A., Torres, A. S. S., López-Ilundain, J., Mezquida, G., Lobo, A., González-Pinto, A., Pina-Camacho, L., Corripio, I., Vieta, E., de la Serna, E., et al. (2020) · Psychological Medicine · doi:10.1017/s003329172000080x · Cited by 9
“The significant results showed small to moderate strength according to the Cohen criteria (Rice & Harris, 2005) in both cases (r values ranged from −0.24 to −0.31 regarding CPZ and the significant cognitive domain correlations, whereas the r value was 0.34 for the correlation between CPZ burden and…”
InfluentialRussell, P. S. S., Chikkala, S. M., Earnest, R., Viswanathan, S. A., Russell, S., & Mammen, P. M. (2020) · World Journal of Psychiatry · doi:10.5498/wjp.v10.i4.71 · Cited by 13
Psychometric properties of the Adulthood Trauma Inventory.
Wittbrodt, M. T., Vaccarino, V., Shah, A. J., Mayer, E. A., & Bremner, J. D. (2020) · Health Psychology · doi:10.1037/hea0000856 · Cited by 10
“10.1037/a0022294 [PubMed: 21517189] Cohen S, Kamarck T, & Mermelstein R (1983).”
BackgroundInfluentialPluck, G. C., Crespo‐Andrade, C., Parreño, P., Haro, K. I., Martínez, M. A., & Pontón, S. C. (2020) · Psychology & Neuroscience · doi:10.1037/pne0000195 · Cited by 18
The HCR-20 for Predicting Violence in Adult Females: A Meta-Analysis
Rossdale, S.-V., Tully, R. J., & Egan, V. (2019) · Journal of Forensic Psychology Research and Practice · doi:10.1080/24732850.2019.1681875 · Cited by 15
“AUC values ≥.714 are generally considered as having a “large” effect size, .639 as “moderate” and ≤.556 as “small” (Rice & Harris, 2005).”
Personalized change awareness: Reducing information overload in loosely-coupled teamwork
Amir, O., Grosz, B., Gajos, K. Z., & Gultchin, L. (2019) · Artificial Intelligence · doi:10.1016/j.artint.2019.05.005 · Cited by 9
5-HTTLPR and COMT Val158Met are not associated with alexithymia: New evidence and meta-analyses
Yang, Y., Li, X., Zhao, J., Xue, M., Zhang, M., Wang, C., Song, H., He, L., Guo, W., & Gong, P. (2019) · Progress in Neuro-Psychopharmacology and Biological Psychiatry · doi:10.1016/j.pnpbp.2019.01.015 · Cited by 12
“The Hardy-Weinberg equilibrium test was conducted with the FINETTI software (Sasieni, 1997).”
Zhou, H., Li, C.-P., Huang, Y., Zou, X., Luo, X.-R., Wu, L.-J., Zhang, L., Xu, X., Yan, W.-L., & Wang, Y. (2018) · World Journal of Pediatrics · doi:10.1007/s12519-018-0201-3 · Cited by 9
Olver, M., & Kingston, D. A. (2018) · Criminal Justice and Behavior · doi:10.1177/0093854818804601 · Cited by 9
Fuermaier, A. B. M., Tucha, O. M., Koerts, J., Send, T., Weisbrod, M., Aschenbrenner, S., & Tucha, L. (2018) · Journal of Clinical and Experimental Neuropsychology · doi:10.1080/13803395.2018.1457139 · Cited by 9
“ROC analysis allows for determination of diagnostic accuracy as measured by the area under the curve (AUC), indicating the probability that a score drawn at random from the first sample (e.g., people feigning ADHD) is higher than a score drawn at random from the second sample (e.g., people with genuine ADHD; Rice & Ha…”
InfluentialInterspecies Scaling in Blast Pulmonary Trauma
Wood, G. W., Panzer, M., Cox, C. A., & Bass, C. (2018) · Human Factors and Mechanical Engineering for Defense and Safety · doi:10.1007/s41314-018-0013-1 · Cited by 9
Pornari, C., Dixon, L., & Humphreys, G. (2018) · Journal of Interpersonal Violence · doi:10.1177/0886260518755487 · Cited by 9
Marriott, R., O'Shea, L., Picchioni, M., & Dickens, G. (2017) · Psychiatry Research · doi:10.1016/j.psychres.2017.07.009 · Cited by 9
“Rice and Harris (2005) reported that AUC values of 0.56, 0.64 and 0.71 are equivalent to small (0.2), moderate (0.5) and large (0.8) Cohen's d values (Cohen, 1992), respectively.”
BackgroundInfluentialMartens, R., Rettenberger, M., & Eher, R. (2017) · Legal and Criminological Psychology · doi:10.1111/lcrp.12080 · Cited by 9
“…results are consistent with previous studies, showing that the Static-2002 provided large effects in predicting general criminal, violent, and sexual recidivism (Bengtson, 2008; Hanson et al., 2010; Langton et al., 2007), corresponding to the recommended critical values by Rice and Harris (2005).”
Actuarial Risk Assessment of Sexual Offenders
Rettenberger, M., & Craig, L. A. (2016) · doi:10.1002/9781118574003.wattso029 · Cited by 9
Vrij, A., Taylor, P., & Picornell, I. (2015) · Wiley eBooks · doi:10.1002/9781118769133.ch12 · Cited by 9
Cohen, J. R., Sheshko, D. M., Ames, A. M., Young, J. F., Hansford, A., Zhu, X., Yao, S., & Abela, J. R. Z. (2014) · Journal of Research on Adolescence · doi:10.1111/jora.12146 · Cited by 10
van der Put, C. E. (2014) · doi:10.1177/1073191113495710 · Cited by 9
“The AUC is regarded as the most important measure for this (Rice & Harris, 2005).”
BackgroundMethodPredictive Validity of the Washington State Juvenile Court Pre-Screen Assessment in the Netherlands
van der Put, C. E., Stams, G. J. J. M., Deković, M., & van der Laan, P. H. (2012) · Assessment · doi:10.1177/1073191112436666 · Cited by 18
Self-Structures, Negative Events, and Adolescent Depression
Cohen, J. R., Spiegler, K. M., Young, J. F., Hankin, B. L., & Abela, J. R. Z. (2013) · The Journal of Early Adolescence · doi:10.1177/0272431613503217 · Cited by 18
Chagigiorgis, H., Michel, S., Seto, M., Laprade, K., & Ahmed, A. (2013) · doi:10.1080/14999013.2013.857740 · Cited by 9
“Studies demonstrated the predictive validity with moderate to large effect sizes (Rice & Harris, 2005) of the HCR-20 in inpatient samples (Dernevik, Grann, & Johansson, 2002 (AUC = .68); Dolan & Fullam, 2007 (AUC = .72); Gray et al., 2003 (AUC = .81); McDermott et al., 2008 (AUC = .65)).”
Backgroundvan der Put, C. E., & Stams, G. (2013) · Psychological Assessment · doi:10.1037/a0033453 · Cited by 9
The great methods bake-off: Comparing performance of machine learning algorithms
Kigerl, A., Hamilton, Z., Kowalski, M., & Mei, X. (2022) · Journal of Criminal Justice · doi:10.1016/j.jcrimjus.2022.101946 · Cited by 8
Laake, A. L. W., Roaldset, J. O., Husum, T., Bjørkly, S., Gustavsen, C. C., Grenabo, S. T., & Lockertsen, Ø. (2025) · European psychiatry · doi:10.1192/j.eurpsy.2025.3 · Cited by 8
“An AUC value of .714 converts to Cohen ’ s d of 0.8, which is typically considered a large effect size [48].”
InfluentialUsing machine learning to detect noncredible cognitive test performance
Finley, J. A., Robinson, A. D., Soble, J. R., & Rodriguez, V. J. (2024) · The Clinical Neuropsychologist · doi:10.1080/13854046.2024.2440085 · Cited by 8
“Classification accuracy was defined as low-to-moderate for AuC values of .65-.69, fair for values of .70-.79, good for values of .80-.89, and excellent for values of ≥.90 (Cohen et al., 2018; Rice & harris, 2005).”
Exploring determinants of school dropout across regions in India: a comprehensive meta-analysis
Venkatesan, R. G., Karmegam, D., & Mappillairaju, B. (2024) · Journal of Computational Social Science · doi:10.1007/s42001-024-00285-4 · Cited by 11
“According to the criteria established by Rice and Harris [69], the size of the highly significant total effect varied from modest (r = 0.122 for "birth order") to large (0.133 for "siblings") in a positive direction to negative (r = − 0.081 for "household size") and positive (r = − 0.062 for "age").”
InfluentialSmeekens, M. V., Robbé, M. D. V., Popma, A., & Kempes, M. (2024) · Frontiers in Psychology · doi:10.3389/fpsyg.2024.1359535 · Cited by 8
Shahriyaripoor, R., Shahhosseini, Z., Pourasghar, M., Hoseinnezhad, Z., Shahriyaripoor, R., & Ganji, Z. (2023) · Journal of Nursing and Midwifery Sciences · doi:10.5812/jnms-137116 · Cited by 8
“Effect size was used to evaluate the strength of the intervention effect (27).”
MethodWen, X., Zhou, Y.-P., Li, Y.-N., Lv, Y., Han, S.-Y., & Zhao, J. (2023) · Frontiers in Psychology · doi:10.3389/fpsyg.2023.1118481 · Cited by 8
Overestimation of Balance Ability Among Older Adults at Risk for Falls
Ickert, E. C., Hughes, T., Berg-Carramusa, C. A., Dudash, S., & Kerns, L. X. (2023) · Journal of Aging and Health · doi:10.1177/08982643231186630 · Cited by 8
“These effect size thresholds are based on the assumption that both values of the binary variable are equally likely to occur (McGrath & Meyer, 2006; Rice & Harris, 2005).”
BackgroundPekmezian, L. B., von Hagen, A., Piñeyro, S., & Senabre, I. (2023) · Psychology Crime and Law · doi:10.1080/1068316x.2023.2214661 · Cited by 13
“Following Rice and Harris (2005), AUC values equal to or greater than 0.56, 0.64, and 0.71 can be interpreted as small, moderate, and large e ff ects, respectively.”
InfluentialUsing serial dependence to predict confidence across observers and cognitive domains
Mei, N., Rahnev, D., & Soto, D. (2023) · Psychonomic Bulletin & Review · doi:10.3758/s13423-023-02261-x · Cited by 8
“However, ROC is arguable the best measure of predictive performance compared with other measures of effect size (Rice & Harris, 2005), and the current ROC confidence prediction scores of ~0.65 for withindomain and slightly lower for cross-domain generalization, indicate that confidence prediction…”
BackgroundInfluentialOpioid‐involved overdose trainings delivered using remote learning modalities
Giordano, N. A., Febres-Cordero, S., Baker, H., Pfeiffer, K. M., Walsh, L. M., Gish, A., & Axson, S. (2023) · Nursing Open · doi:10.1002/nop2.1615 · Cited by 8
“A sample size as small as 6 provides sufficient power to detect a large effect size, specifically a Cohen's d > 1.40, using paired t- test (Rice & Harris, 2005).”
MethodInfluentialPham, A. T., Hilton, N., Ennis, L., Nunes, K., & Jung, S. (2023) · Criminal Justice and Behavior · doi:10.1177/00938548221143535 · Cited by 8
“…the most commonly used and recommended effect size statistic for evaluating the prediction of recidivism (Helmus & Babchishin, 2017; Rice & Harris, 2005) and is typically preferred to other measures of predictive validity that are more affected by the base rate of recidivism (Rice & Harris, 2005).”
BackgroundMethodInfluentialLee, S. C., Hanson, R. K., & Yoon, J. S. (2022) · Sexual Abuse · doi:10.1177/10790632221139173 · Cited by 20
“According to Rice and Harris (2005), AUCs of .56 would be considered small, .64 would be moderate, and .71 would be large.”
BackgroundMethodShura, R., Ingram, P. B., Miskey, H., Martindale, S., Rowland, J., & Armistead-Jehle, P. (2022) · Clinical Neuropsychologist · doi:10.1080/13854046.2022.2131630 · Cited by 8
“Large effects for classification accuracy were determined by AUC values greater than .714, which approximate large effect sizes (Cohen, 1988; Rice & Harris, 2005).”
Ladányi, E., Novakovic, M. M., Boorom, O. A., Aaron, A. S., Scartozzi, A., Gustavson, D. E., Nitin, R., Bamikole, P. O., Vaughan, C., Fromboluti, E. K., et al. (2022) · Neurobiology of Language · doi:10.1162/nol_a_00082 · Cited by 15
“The AUC of slowest motor tempo was 0.75, which can be interpreted as an acceptable performance in predicting DLD status (values between 0.7 and 0.8 are considered acceptable, while values between 0.8 and 0.9 are considered excellent; Rice & Harris, 2005).”
InfluentialDistortions in Spatial Mental Representation Affect Sketch Maps in Young Adults
Lopez, A., & Bosco, A. (2022) · Neurology International · doi:10.3390/neurolint14040064 · Cited by 8
Delinquency between siblings: A meta-analysis
Walters, G. (2022) · Aggression and Violent Behavior · doi:10.1016/j.avb.2022.101792 · Cited by 8
Fernandes, L., Villadsen, A., Holm, C., Sørensen, M. S., Zebis, M., Andersen, L., & Petersen, M. M. (2022) · Disability and Rehabilitation · doi:10.1080/09638288.2022.2101696 · Cited by 8
Tong, J., Alonso-Caneiro, D., Kalloniatis, M., & Zangerl, B. (2022) · Ophthalmic & physiological optics · doi:10.1111/opo.12997 · Cited by 8
“80 indicating excellent diagnostic accuracy.(51,52) However, poorer accuracy at peripheral macular locations was observed, where curves approached the no- discrimination line and AUROCs ranged between 0.”
Ng, L. C., Oblath, R., Brigham, R., Tai, M. H., & Coles, M. S. (2022) · Child and Adolescent Psychiatry and Mental Health · doi:10.1186/s13034-022-00501-x · Cited by 11
Forecasting childhood adversities from conditions of birth
Green, M. J., Watkeys, O. J., Kariuki, M., Hindmarsh, G., Whitten, T., Dean, K., Laurens, K. R., Harris, F., & Carr, V. J. (2022) · Paediatric and Perinatal Epidemiology · doi:10.1111/ppe.12828 · Cited by 8
Prospective Field Validation of the START:AV in a Dutch Secure Youth Care Sample
Beuf, T. L. F. D., de Vogel, V., Broers, N., & de Ruiter, C. (2021) · Assessment (Odessa, Fla.) · doi:10.1177/10731911211063228 · Cited by 8
“For each AUC value, we also report the approximate Cohen’s d based on calculations by Rice and Harris (2005; see Table, p. 616), with d values of .20, .50, and .80 representing a small, medium, and large effect size (Rice & Harris, 2005).”
MethodAssessing the Risk of Repeat Victimization Using Structured and Unstructured Police Information
Geurts, R., Raaijmakers, N., Delsing, M. J. M. H., Spapens, T., Wientjes, J. A. M., Willems, D., & Scholte, R. H. J. (2021) · Crime & Delinquency · doi:10.1177/00111287211047533 · Cited by 10
“The AUC value is one of the most well-known and widely used indicators to assess an model’s predictive validity (Messing & Thaller, 2013; Rice & Harris, 2005).”
BackgroundMethodZentner, M. R., Biedermann, V., Taferner, C., da Cudan, H., Möhler, E., Strauß, H., & Sevecke, K. (2021) · Frontiers in Psychiatry · doi:10.3389/fpsyt.2021.667503 · Cited by 9
“80 are considered to be realistic of a good test (34).”
BackgroundDepression Symptoms are Linked to Music Use
Kanagala, S. C., Schäfer, T., Greenberg, D. M., & Gabińska, A. (2021) · Music & Science · doi:10.1177/20592043211057217 · Cited by 15
“McFerran and Saarikallio (2014) demonstrated in their study that young people at risk of depression are more likely to rely on music to improve their affective states and feel content about themselves.”
BackgroundKoh, L. L., Day, A. S., Klettke, B., Daffern, M., & Chu, C. (2020) · International journal of offender therapy and comparative criminology · doi:10.1177/0306624x20970887 · Cited by 8
“AUCs of 0.56, 0.64 and 0.71 are generally be considered to represent “small,” “moderate,” and “large” effect sizes respectively (Rice & Harris, 2005).”
BackgroundInfluentialStokes, M. J., Witchalls, J. B., Waddington, G. S., & Adams, R. D. (2020) · Physical Therapy in Sport · doi:10.1016/j.ptsp.2020.09.009 · Cited by 10
Dupuis, M., Meier, E., Gholam-Rezaee, M., Gmel, G., Strippoli, M., & Renaud, O. (2020) · Personality and Individual Differences · doi:10.1016/j.paid.2020.109812 · Cited by 8
Toward Optimal Selection of Information Retrieval Models for Software Engineering Tasks
Rahman, M. M., Chakraborty, S., Kaiser, G. E., & Ray, B. (2019) · IEEE Working Conference on Source Code Analysis and Manipulation · doi:10.1109/scam.2019.00022 · Cited by 8
“05) with medium to large Cohen’s d effect size [56].”
BackgroundNevill, R. E., Hedley, D., & Uljarević, M. (2019) · Journal of Autism and Developmental Disorders · doi:10.1007/s10803-019-04153-3 · Cited by 8
Cartwright, J. K., Desmarais, S. L., Johnson, K. L., & Dorn, R. A. V. (2018) · Psychological Services · doi:10.1037/ser0000183 · Cited by 8
Arai, K., Takano, A., Nagata, T., & Hirabayashi, N. (2017) · Criminal Behaviour and Mental Health · doi:10.1002/cbm.2007 · Cited by 8
Wientjes, J., Delsing, M., Cillessen, A., Janssens, J., & Scholte, R. (2017) · Journal of Criminological Research Policy and Practice · doi:10.1108/jcrpp-01-2017-0008 · Cited by 8
“…OF CRIMINOLOGICAL RESEARCH, POLICY AND PRACTICE j VOL. 3 NO. 4 2017 D ow nl oa de d by 1 43 .1 79 .1 23 .1 3 A t 0 3: 57 2 3 N ov em be r 20 17 ( PT ) performance indicator to measure the predictive validity of structured risk assessments (Babchishin and Helmus, 2016; Rice and Harris, 2005).”
BackgroundMcCafferty, J. T., & Scherer, H. (2017) · doi:10.1177/0032885517734493 · Cited by 8
“The use of ROC curve analyses has been strongly advocated for by risk assessment scholars as a means to bring consistency to the methodology of correctional predictive validity research (see Rice & Harris, 1995, 2005).”
MethodWard, E., Hillier, S., Raynor, A., & Petkov, J. (2017) · Pediatric Physical Therapy · doi:10.1097/pep.0000000000000423 · Cited by 8
Vasiljević, Z., Berglund, M., Öjehagen, A., Höglund, P., & Andersson, C. (2017) · Psychiatry, Psychology and Law · doi:10.1080/13218719.2017.1308219 · Cited by 8
“However, whether a specific AUC value is considered small or large depends on the context, and other more liberal classification values have been suggested in the context of recidivism risk assessments (Rice & Harris, 2005).”
BackgroundNg, T. H., Burke, T. A., Stange, J. P., Walshaw, P. D., Weiss, R. B., Urošević, S., Abramson, L. Y., & Alloy, L. B. (2017) · Journal of Abnormal Psychology · doi:10.1037/abn0000255 · Cited by 13
“The AUC values of .56, .64, and .71 represent small, medium, and large effects, respectively (Rice & Harris, 2005).”
BackgroundZhang, J.-W. (2016) · International journal of offender therapy and comparative criminology · doi:10.1177/0306624x14557471 · Cited by 8
“AUC coefficients greater than .71 are regarded as large effects (Rice & Harris, 2005), thus the AUC coefficient of .733 in the current study can be placed in the large range.”
BackgroundDynamic risk and sexual offending: the conundrum of assessment
Casey, S. (2016) · Psychology Crime and Law · doi:10.1080/1068316x.2015.1111366 · Cited by 8
“It is worth noting that Rice and Harris (2005) acknowledged that the categories were provided as a tentative rule of thumb.”
BackgroundImplementing violence and incident reporting measures on a forensic mental health unit
Woods, P., Olver, M., & Muller, M. (2015) · doi:10.1080/1068316x.2015.1077247 · Cited by 8
“Using the guidelines of Rice and Harris (2005), AUC magnitudes can be interpreted in terms of small (0.56–0.63), medium (0.64–0.70), and large (0.71+).”
BackgroundMethodInfluentialBabchishin, K., Nunes, K., Hermann, C. A., & Malcom, J. R. (2015) · Journal of Sexual Aggression · doi:10.1080/13552600.2013.836575 · Cited by 8
“Corresponding values for AUC according to Rice and Harris (2005) are .56, .64 and .71.”
MethodSex Offender Modus Operandi Stability and Relationship With Actuarial Risk Assessment
Lasher, M. P., McGrath, R., & Cumming, G. F. (2015) · Journal of Interpersonal Violence · doi:10.1177/0886260514539757 · Cited by 8
“The AUC statistic represents the probability that a randomly selected recidivist will have a higher score on a risk measure than a randomly selected non-recidivist, with .5 representing chance-level prediction and 1.0 representing perfect prediction (Rice & Harris, 2005).”
BackgroundInfluentialRettenberger, M., Gaunersdorfer, K., Schilling, F., & Eher, R. (2009) · doi:10.1007/s11757-009-0016-0 · Cited by 8
Davies, S. J. C. (2019) · doi:10.26686/wgtn.17139935.v1 · Cited by 8
“The RoC*RoI and three DRAOR subscale scores significantly predicted general recidivism over both follow-up periods.”
BackgroundInfluentialHurting Others vs. Hurting Myself, a Dilemma for our Autonomous VehiclePreprint
Luzuriaga, M., Heras, A. J., & Kunze, O. (2019) · Social Science Research Network · doi:10.2139/ssrn.3345141 · Cited by 8
Zhu, N., Pan, X.-Q., & Zhao, F. (2025) · Children · doi:10.3390/children12040478 · Cited by 7
“An AUC between 0.556 and 0.639 is considered a small effect size, between 0.639 and 0.714 is a medium effect size, and 0.714 or above is a large effect size [36].”
InfluentialAn Empirical Study on the Characteristics of Bias upon Context Length Variation for Bangla
Sadhu, J., Khan, A. A., Bhattacharjee, A., & Shahriyar, R. (2024) · Annual Meeting of the Association for Computational Linguistics · doi:10.48550/arxiv.2406.17375 · Cited by 7
“(Rice and Harris, 2005)”
Associations Between Gait Speed and Fat Mass in Older Adults
Šagát, P. (2024) · Clinical Interventions in Aging · doi:10.2147/cia.s456724 · Cited by 7
“29 To be able to correctly classify the accuracy of the test, Rice and Harris 30 proposed a classification of the AUC as follows: 1) 0.55– 0.62 (small), 2) 0.63–0.71 (moderate), and 3) >0.71 (large).”
MethodInfluentialMiller, H. A., Toman, E., & Pederson, K. M. (2024) · Criminal Justice and Behavior · doi:10.1177/00938548241240991 · Cited by 7
“Area under the curve values also indicate effect sizes, with small effect sizes being between .556 and .637, medium between .639 and .712, and large values equal or above .714 ( Rice & Harris, 2005).”
InfluentialExamining Trajectories of Change on the Dynamic Risk Assessment for Offender Re-Entry (DRAOR)
Rieger, D. J., Perley-Robertson, B., & Serin, R. C. (2024) · International journal of offender therapy and comparative criminology · doi:10.1177/0306624x241240701 · Cited by 7
“AUC analysis provides a measure of overall predictive accuracy of a scale instrument on a binary outcome by examining rates of true and false positives across decision thresholds (Rice & Harris, 2005).”
Metadata
- Authors
- Rice, M. E., & Harris, G. T.
- Affiliations
- Waypoint Centre for Mental Health Care (CA)
- Venue
- Law and Human Behavior
- Published
- 2005-10
- Volume
- 29
- Issue
- 5
- Pages
- 615-620
- ISSN
- 0147-7307 (ISSN-L)
- 1573-661X
- OpenAlex
- W2060774914
- Cited by
- 1,608
- References
- 0
- Citations by year
- 2026: 60 · 2025: 84 · 2024: 90 · 2023: 121 · 2022: 151 · 2021: 150 · 2020: 174 · 2019: 137 · 2018: 112 · 2017: 80 · 2016: 88 · 2015: 65 · 2014: 64 · 2013: 58 · 2012: 51
- FWCI
- 11.55
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Law and Human Behavior (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Psychology Research and Bibliometrics
- Keywords
- effect size, Cohen's d, diagnostic accuracy, prediction accuracy, forensic psychology, follow-up study
- MeSH terms
- Forecasting, Humans, ROC Curve, Risk Assessment
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
