A multi-theoretical framework to explain same-sex intimate partner violence perpetration and victimization: a test of social learning, strain, and self-control

Abstract

Criminological theories used to explain intimate partner violence perpetration and victimization in heterosexual relationships have rarely been extended to same-sex intimate partner violence (SSIPV). As such, this study examines the predictability of social learning, general strain, and self-control to SSIPV perpetration and victimization. Using a sample of college students from two large public universities, results indicate that variables derived from general strain theory better predicted SSIPV perpetration than any other criminological theory (social learning and selfcontrol). Only one variable (depression) was significant in predicting victimization. Implications for SSIPV research and future directions are discussed.

Keywords:

• Social learning
• self-control
• strain
• same-sex intimate partner violence

Acknowledgments

I would like to thank the editor Michael J. Leiber and the anonymous reviewers for their constructive comments and suggestions. An earlier version of this paper was presented at the 2016 meeting of the Academy of Criminal Justice Sciences in Denver, Colorado.

Notes

1. Application forms were submitted to both of the universities’ Institutional Review Board (IRB) for review. The study was approved by both IRBs and all participants gave informed consent. All participants were able to decline the research invitation and were also allowed to discontinue the study at any time without consequences.

2. It is important to note that although the response rate is low, it is consistent with other studies using non-compensated, online surveys (see Fox, Nobles, and Fisher 2016; Ngo and Paternoster 2011; Reyns, Henson, and Fisher 2011).

3. Dichotomizing victimization data are common practice. This is sometimes due to the fact that scholars are interested in the occurrence of violence rather than the frequency of violence. Clark and Watson (1995) pointed out that dichotomous scales can provide valid and reliable results and can capture significant amounts of information. The current method is consistent with prior studies (Cheung, Choi, and Cheung 2014).

4. It is important to note that the Revised Conflict Tactics Scale have been criticized for capturing relatively ‘minor’ forms of violence (Kimmel 2002). An exploratory analysis (not shown) revealed that only a small number of respondents reported perpetrating or being victimized by more severe forms of violence. Only 4 respondents said they were victimized by ‘a knife or other weapon besides a gun’ and another 6 respondents reported that they were threatened by their partner ‘with a knife or other weapon.’ The vast majority of respondents reported what is called ‘common couple violence,’ which includes slapping, pushing, grabbing, hitting, naming calling, etc. Due to this limitation in the data, I was not able to operationalize IPV in a way that would capture more severe forms of IPV.

5. I focused on trait anger (an enduring characteristic of individuals) rather than situational anger (a reaction to a specific set of circumstances) in order to remain consistent with prior test of general strain theory (Hay and Evans 2006; Hay 2003; Pérez, Jennings, and Gover 2008).

6. The decision to measure only impulsivity and risk-seeking behaviors was influenced by studies that have shown these two dimensions of self-control carry the most explanatory power (Arneklev, Grasmick, and Bursik 1999; Lagrange and Silverman 1999). It should be noted that other scholars have used these same 8 items to measure self-control in studies testing Gottfredson and Hirschi’s (1990) general theory of crime (Agnew et al. 2011; Childs, Cochran, and Gibson 2009; Higgins et al. 2009; Jennings et al. 2010).

7. All the variables have less than 8% missing data. The frequencies (%) of missing data on individual items are as follows: perpetration (8%), victimization (4%), mother-to-father violence (6%), father-to-mother violence (6%), internalized homophobia (5%), stigma (7%), antigay violence (2%), discrimination (2%), anger (6%), depression (6%), social support (7%), and self-control (6%).

8. A coefficient comparison test (see, Paternoster et al. 1998) was estimated when the coefficient were significant in both models (Heterosexual vs. Non-Heterosexual). This test was performed on age on the victimization model and on self-control and white on the perpetration model. These tests indicated that there was no different effect for the slopes. The slopes for the heterosexual and non-heterosexual groups on these variables are statistically similar.

9. In an exploratory analysis (not shown) I included problematic alcohol consumption in the statistical model predicting perpetration and victimization given that a number of studies have shown a relationship between the two (Lewis et al. 2015). Results indicated a non-significant relationship in these data. Results are not shown or discussed in order to preserve space.