COMPARISON OF SURVIVAL ANALYSIS USING ACCELERATED FAILURE TIME MODEL AND COX MODEL FOR RECIDIVIST CASE

  • Nuraziza Arfan Actuarial Master Study Program, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia https://orcid.org/0009-0006-9559-8552
  • Asrul Irfanullah Actuarial Master Study Program, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia https://orcid.org/0009-0007-5398-8488
  • Muhammad Rozzaq Hamidi Actuarial Master Study Program, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia https://orcid.org/0009-0002-1498-5958
  • Utriweni Mukhaiyar Statistics Research Division, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Indonesia https://orcid.org/0000-0003-2353-7110
DOI: https://doi.org/10.30598/barekengvol19iss1pp629-642
Keywords: Accelerated Failure Time, AIC, BIC, Cox PH, Recidivist, Regression, Stratified Cox

Abstract

Recidivists, or ex-prisoners who commit crimes after serving a prior sentence, pose a critical challenge to the criminal justice system. This study examines social and economic factors that may reduce the likelihood of recidivists being re-arrested. Using survival analysis, the probability that a recidivist could survive in society without being re-arrested could be estimated. The purpose of this work is to compare the AFT and Cox models to determine which provides a better fit to identify factors affecting the likelihood of re-arrest within one year after release and to statistically assess the impact of these factors. This study utilizes a stratified Cox model to address variables that violate the proportional hazards (PH) assumption. The analysis is limited to four types of AFT models: Weibull, log-normal, log-logistic, and exponential. Results show that the stratified Cox model provides the best fit, based on Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). This demonstrates the Cox model's robustness in analyzing survival data, accurately approximating the distribution of survival times without restrictive assumptions, unlike AFT models. The study reveals that recidivists who received financial aid upon release have a  lower risk of re-arrest compared to those who did not, and each additional prior theft arrest increased the risk of re-arrest by 1.09193 times.

Downloads

Download data is not yet available.

References

L. Sugiharti, R. Purwono, M. A. Esquivias and H. Rohmawati, "The Nexus between Crime Rates, Poverty, and Income Ineqality: A Case Study of Indonesia," MDPI, vol. 11, no. 62, pp. 1-15, 2023.

D. G. Kleinbaum and M. Klein, Survival Analysis: A Self Learning Text, New York: Springer, 2012.

M. J. V. Rekstad, A Comparison of Methods for Modelling Survival for Cancer Patients, Norway: NMBU, 2023.

M. Parsa and I. Van Keilegom, "Accelerated Failure Time vs Cox Proportional Hazards Mixture Cure Models: David vs Goliath?," Springer Nature, vol. 64, pp. 835-855, 2023.

J. Qi, Comparison of Proportional Hazards and Accelerated Failure Time Models, Saskatchewan: University of Saskatchewan, 2009.

K. Sarkar, R. Chowdhury and A. Dasgupta, "Analysis of Survival Data: Challenges and Algorithm-Based Model Selection," Journal of Clinical and Diagnositc Research, pp. 14-20, 2017.

D. W. Hosmer, S. Lemeshow and S. May, Applied Survival Analysis, New Jersey: John Wiley & Sons, Inc, 2008.

J. D. Kalbfleisch and R. L. Prentice, The Statistical Analysis of Failure Time Data, New Jersey: John Wiley & Sons, Inc., 2002.

N. Mishra and S. Misra, "Comparison of Cox Proportional Hazard Model and Accelerated Failure Time (AFT) Models: an Application to Under-five Mortality in Uttar Pradesh," Indian Journal of Applied Research, vol. 12, no. 12, pp. 83-87, 2022.

K. Mardhiah, N. Wan-Arfah, N. N. Naing, M. R. Abu Hassan and H.-K. Chan, "Comparison of Cox proportional hazards model, Cox proportional hazards with time-varying coefficients model, and lognormal failure time model: Application in time to event analysis of melioidosis patients," Asian Pacific Journal of Tropical Medicine, vol. 15, no. 3, pp. 128-134, 2022.

R. Mutambayi, A. Azeez, J. Ndege and Y. Qin, "Statistical Analysis of Risk factors of Malaria re-infection among Outpatients in DR, Congo: A Comparison Approach of AFT and Cox PH Models," Journal of Communicable Diseases, vol. 51, no. 1, pp. 22-28, 2019.

A. Khamis, C. A. Che Hamat and M. A. A. Abdullah, "Modeling Students' Performance using Cox and Parametric Accelerated Failure Time Models," Scientific Research Journal (SCIRJ), vol. VIII, no. VII, pp. 44-49, 2020.

D. Melak, S. Wedajo and R. Dewau, "Time to Viral Re-suppression and Its Predictors among Adults on Second-Line Antiretroviral Therapy in South Wollo Zone Public Hospitals: Stratified Cox Model," HIV/AIDS - Research and Palliative Care, vol. 15, pp. 411-421, 2023.

P. H. Rossi, R. A. Berk and K. J. Lenhan, Money, Work, and Crime: Work Experimental Evidence, New York: Academic Press, 1980.

R. De la Cruz, P. Oslando, M. A. V. Valle and G. A. Ruz, "Modeling Recidivism through Bayesian Regression Models and Deep Neural Networks," MDPI, vol. 9, no. 6, p. 639, 2021.

L. Waldeck, "How to Predict Recidivism in Young Offenders: Comparing Logistic Regression, Random Forest, and Boosted Classification Trees and Examining Possible Risk Factors," Maastricht University, Maastricht, 2020.

I. Lijarcio, F. J. Llamazares, E. Valle, L. Montoro and S. A. Useche, "Assessing Risk Perception over Recidivist Traffic Offenders from a Multi-group Approach: How Gendered Could It Be?," The European Journal of Psychology Applied to Legal Context, vol. 14, no. 1, pp. 33-41, 2022.

A. Singh and S. Mohapatra, "Automatic Criminal Recidivism Risk Estimation in Recidivist Using Classification and Ensemble Techniques," Soft Computing for Problem Solving, pp. 279-289, 2023.

J. Fox, An R and S-Plus Companion to Applied Regression, Thousand Oaks: SAGE Publications, Inc, 2011.

Y. Anastasya, N. Satyahadewi and H. Perdana, "Model Stratified Cox untuk Mengatasi Nonproportional Hazard," Buletin Ilmiah Math. Stat. dan Terapannya, pp. 311-320, 2016.

D. R. Cox and D. Oakes, Analysis of Survival Data, London: CRC Press, 1998.

T. M. Therneau and P. M. Grambsch, Modeling Survival Data: Extending The Cox Model, New York: Springer, 2000.

R. Saikia and M. P. Barman, "A Review on Accelerated Failure Time Models," International Journal of Statistics and Systems, vol. 12, no. 2, pp. 311-322, 2017.

M. Mills, Introducing Survival and Event History Analysis, London: SAGE Publications Ltd, 2011.

P. Allison, Survival Analysis Using SAS: A Practical Guide, Cary: SAS Institute, 2010.

Published
2025-01-13
How to Cite
[1]
N. Arfan, A. Irfanullah, M. R. Hamidi, and U. Mukhaiyar, “COMPARISON OF SURVIVAL ANALYSIS USING ACCELERATED FAILURE TIME MODEL AND COX MODEL FOR RECIDIVIST CASE”, BAREKENG: J. Math. & App., vol. 19, no. 1, pp. 629-642, Jan. 2025.
Issue
Vol 19 No 1 (2025): BAREKENG: Journal of Mathematics and Its Application
Section
Articles

Copyright (c) 2025 Nuraziza Arfan, Asrul Irfanullah, Muhammad Rozzaq Hamidi, Utriweni Mukhaiyar

Creative Commons License

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Authors who publish with this Journal agree to the following terms:

  1. Author retain copyright and grant the journal right of first publication with the work simultaneously licensed under a creative commons attribution license that allow others to share the work within an acknowledgement of the work’s authorship and initial publication of this journal.
  2. Authors are able to enter into separate, additional contractual arrangement for the non-exclusive distribution of the journal’s published version of the work (e.g. acknowledgement of its initial publication in this journal).
  3. Authors are permitted and encouraged to post their work online (e.g. in institutional repositories or on their websites) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published works.