Social Support Theory

In the article “Applying Differential Coercion and Social Support Theory to Police Officer’s Misconduct”, authors Zavala and Kurtz strive to provide a connection between police officer misconduct and the emerging theory of differential coercion and social support. This is done by giving a brief description outlining the main components of differential coercion and social support theory and then examining coercive forcers, social support, intimate partner violent (IPV), and alcohol consumption among police officers (2016). The authors are able to measure these factors by using data collected from the Police Stress and Domestic Violence in Police Families in Baltimore, Maryland in 1997 – 1999. Furthermore, after running their analysis, the authors

suggest that evidence supports victimization being a form of coercion that can aid to criminal behavior (cite). Although this article proves to provide an interesting new perspective on police officer misconduct, there are some limitations to what is provided, along with some strengths.
The theoretical framework of this article is embedded in differential coercion and social support theory.

The theoretical framework serves a purpose to ensure that the question asked is not based on an individual’s suspicion alone, reduce bias, and can help the reader understand the perspective and context better. Zavala and Kurtz give a brief description of this theory and how it relates to IPV perpetrated by police officers and problematic alcohol consumption, which provides evidence for their research purpose (2016). In short, this theory operates by using its’ two different elements, coercion and social support, to determine if someone will engage in criminal behavior (Zavala & Kurtz, 2016). For instance, the authors use pervious literature to give examples of how coercion and social support impact an individual’s life. Although, there is no research provided directly on this theory and police misconduct, the authors do give examples of studies that show evidence of coercion and social support related to IPV and alcohol consumption among police officers, to provide support for their research purpose. By doing this the authors are able to fill the gaps with adequate support for their reasoning in engaging in this study. After reviewing their theoretical framework, Zavala and Kurtz develop two research hypotheses regarding coercive forces and social support related to IPV perpetrated by police and problematic alcohol consumption

(2016).
Furthermore, to test the hypothesis post the data is gathered from the Police Stress and Domestic Violence in Police Families in Baltimore, Maryland in 1997 – 1999. The authors develop two main dependent variables, perpetration and alcohol consumption. Perpetration is only based off of one very general question asking if the officer has even gotten out of control with their significant other (Zavala & Kurtz, 2016, p.882). Therefore, because this is question is so broad it serves as a limit to the research by not adequately measure perpetration. The independent variables in the study include coercion, social support, and social-psychological deficits (Zavala & Kurtz, 2016, p. 883). Each of these main independent variables include other variables such as coercion including child maltreatment, street victimization, peer victimization, etc. However, many of these variables are hindered in a similar way as perpetration, being that they are only one very broad question being answer. Along with dependent and independent variables, the authors also controlled for sex, race and ethnicity, education, rank, and marital status (Zavala & Kurtz, 2016, p. 884).
When discussing the variables used by Zavala and Kurtz, it is important to identify that their coding and computing of variables may have not been consistent throughout their process. For instance, when doing a re-analysis of this research, identifying how the new variables were computed served to be difficult, being that what worked for one variable didn’t work for the following, being that the sum they would use would be different. Along with coding, the authors failed to give clear instruction of where missing values fall when recoding and computing new variables. For example, some missing values may fall into a “no” category while others may be labeled “missing”, if this placement is not consistent it has the potential to sway the outcomes of the analysis. Although these are small, these issues could serve as a limitation in the research presented by Zavala and Kurtz.
A logistic regression model was used to test these hypotheses.

A logistic regression is normally used when dependent variables are nominal and there is more than one independent variable, which there was in this study (cite). Logistic regression also is used to describe data and explain the relationship between one dependent binary variable and one or more nominal, ordinal, interval, or ratio-level independent variables (cite). When revisiting the two hypotheses that Zavala and Kurtz proposed, they initially wanted to explain the relationship between coercion, social support and IPV and alcohol consumption among police officers. Taking into consideration the variables they used and the way they were coded using a logistic regression for their analysis was an appropriate

decision.
After completing their analysis, Zavala and Kurtz reported their findings. In general, they sum that their findings support previous evidence that victimization is a consistent form of coercion that can lead to criminal behavior (Zavala & Kurtz, 2016). Within the models produced, there are few variables were significant on the dependent variables. When looking at coercive forces were positive and significant in predicting IPV perpetration, only two variables, inherent to policing and peer victimization were significant (Zavala & Kurtz, 2016, p. 886). The authors go on to give more interpretations of their findings such as “police officers who reported child maltreatment had 81% greater odds of reporting perpetrating violence towards their partners than those with no child maltreatment” (Zavala & Kurtz, 2016, p. 884). When the authors are interpreting model two of table two they state that anger was the only thing found to be significant in social-psychological deficits (Zavala & Kurtz, 2016, p. 885). However, when viewing the model, anger is not statistically significant, being that it is over the standard 0.05 acceptance. Although they do have asterisks at the bottom of the table, indicating p < .01 ***, it still serves as a confusing understanding for readers. This interpretation serves as another limitation for this article.