RT Article
T1 Quantifying the Likelihood of False Positives: Using Sensitivity Analysis to Bound Statistical Inference
JF Journal of quantitative criminology
VO 35
IS 4
SP 631
OP 662
A1 Thomas, Kyle J.
A2 McGloin, Jean Marie
A2 Sullivan, Christopher J.
LA English
YR 2019
UL https://krimdok.uni-tuebingen.de/Record/169173960X
AB Criminologists have long questioned how fragile our statistical inferences are to unobserved bias when testing criminological theories. This study demonstrates that sensitivity analyses offer a statistical approach to help assess such concerns with two empirical examples—delinquent peer influence and school commitment.
K1 Theory testing
K1 Unobserved bias
K1 False positives
K1 sensitivity analysis
K1 Sensitivity analysis
DO 10.1007/s10940-018-9385-x