Introduction (PDF & R-Code) An outlier is a case, datapoint, or score meaningfully removed from the mass of the distribution as to be recognizably different from the remainder of cases, datapoints, or scores. Consistent with this definition, outliers are conditioned upon the observed data and can vary between samples. For example, an individual with 10Continue reading “Entry 4: The Normality Assumption- Outliers”
Author Archives: Ian A. Silver
Entry 3: The Collinearity Assumption
Introduction (PDF & R-Code) Regression based techniques are one of the most frequently used statistical approaches for hypothesis testing. The primary benefit of regression techniques is the ability to adjust estimates for the variation across multiple independent variables – otherwise known as the statistical control approach. This is extremely valuable when we are interested inContinue reading “Entry 3: The Collinearity Assumption”
Entry 2: The Homoscedasticity Assumption
Introduction (PDF & R-Code) Homoscedasticity and heteroscedasticity are not just difficult words to pronounce Homoscedasticity and heteroscedasticity are not just difficult words to pronounce (homo·sce·das·tic·i·ty & hetero·sce·das·tic·i·ty), but also terms used to describe a key assumption about the distribution of error in regression models. In statistics we regularly make assumptions about the structure of error.Continue reading “Entry 2: The Homoscedasticity Assumption”
Entry 1: The Linearity Assumption
Introduction (PDF & R-Code) Satisfying the assumption of linearity in an Ordinary Least Squares (OLS) regression model is vital to the development of unbiased slope coefficients, standardized coefficients, standard errors, and the model R2. Simply put, if a non-linear relationship exists, the estimates produced from specifying a linear association between two variables will be biased.Continue reading “Entry 1: The Linearity Assumption”
A failure to maintain regulated exposure: Developing an understanding of the predictors and recidivistic effects of experiencing program interruptions
Abstract: At the forefront of correctional scholarship is the reestablishment of the rehabilitative ideal and the ways in which correctional departments can influence recidivism. Of importance is the interruption of rehabilitative treatment due to highly frequent movement of inmates between correctional facilities. To address this concern, the current study evaluated the predictors and recidivistic effectsContinue reading “A failure to maintain regulated exposure: Developing an understanding of the predictors and recidivistic effects of experiencing program interruptions”
Ensnarement during imprisonment: Re-conceptualizing theoretically driven policies to address the association between within-prison sanctioning and recidivism.
Article Summary: The current study used data collected during the Evaluation of Ohio’s Prison Programs. The analytical sample of N = 63,772 inmates represents one of the largest samples used to assess the association between within-prison sanctioning and recidivism. Latent class growth analysis (LCGA) demonstrated that five guilty sanctioning clusters existed within the data: PersistentContinue reading “Ensnarement during imprisonment: Re-conceptualizing theoretically driven policies to address the association between within-prison sanctioning and recidivism.”
Academic Achievement and the Implications for Prison Program Effectiveness and Reentry
Article Summary: The current study examines how academic achievement—measured as verbal and math performance—is associated with prison programming and reentry. We assess how academic achievement might be directly associated with recidivism and whether this occurs through indirectly by moderating the effectiveness of in-prison programs. Using a statewide subsample of incarcerated individuals (N = 13,536) theContinue reading “Academic Achievement and the Implications for Prison Program Effectiveness and Reentry”
ianasilver.com 