Institutional Review Board approval was obtained before examination or manipulation of data
variables. Data were analyzed through descriptive and inferential statistics using the Statistical
Package for the Social Sciences (SPSS) version 18. Although the dataset was reported to have a
larger sample size, the violence perpetration questions were excluded for some respondents. This
significantly reduced the research sample size. Of the recoded variables, statistical analysis
methods were used to determine the appropriate summation or dichotomization recoding of key
variables. Upon completion of the recoding process, the descriptive data were cross-referenced
with the variable descriptive coding data preceding the recoding process. Logistic regression
analyses were conducted to determine the likelihood to predict violence perpetration. Logistic
regression was selected as a more appropriate test given the categorical responses of the dependent
variable. The completed analysis was to first fit a logistic regression model using a maximum of 10
variables, due to the limited sample size (N = 105). The logistic regression analyzed three
independent variables and two dependent variables (threats and physical violence). The primary
focus of data analysis was inferential statistics to determine the fit of a model that would predict
female violence perpetration from a nationally representative sample.