MH - 264

Online Appendix for Law Enforcement and Social Media (Article)

Online Appendix for Law Enforcement and the Depiction of Minorities and Women on Social Media

Sensitivity Analysis of Table 4 (Without Top 25 Cities With Largest Output)OLS Regression Analysis of Female Officer Images
Officer with
Community
(n=2303)
%
General
Officer
(4429)
%
Community
with Officers
(n=5134)
%
BOLO
(n=280)
%
Arrest
(n=198)
%
Missing
Persons
(n=71)
%
White Male58.156.427.422.530.325.3
White Female9.610.527.010.44.519.7
Black Male14.211.511.339.633.315.5
Black Female3.83.510.86.87.614.1
Asian Male1.11.71.31.42.01.4
Asian Female0.50.71.90.00.04.2
Hispanic Male9.812.29.716.018.712.7
Hispanic Female2.83.510.53.23.57.0

 

Variables and Findings

The dependent variables in Table 9 include percentage of total officer images that are female, percentage of general officer images that are female, and the percentage of images of officers interacting with community member that are female. Officers interacting with the community denote images where the officers are simply interacting with the community in a positive manner (not making an arrest, etc.). Control variables for these variables include city population, and police per 100k people (sworn officers per 100,000 people). Percent Female Officers (actual percentage officers that are female) will be the primary variable of interest in this analysis. This variable was collected through departmental web-pages, surveys of police departments, media reports within the last two years, and through departmental Facebook pages.

While there was some explanatory value in Table 8, results of the models that test the percentage of females in images of officers in Facebook pictures are negligible. We did find that the actual percentage of female officers across cities does have a significant positive impact on the percentages of the officer images that consist of female officers across those cities. The model as a whole, however, provides little explanation of female officers in images.

Table of Results

Ordinary Least Squares Regression (OLS) of % of the population that is female, population size, and police force size, % of officer images that are of Female Officers (n= 163 Cities)

Total
Officer
Images

% Female

General
Officer
Images

% Female

Officers With
Community
Images

% Female

Population Size.000
(-.013)
.000
(-.533)
.000
(2.828)
Percent Female Off..005**
(2.676)
.006*
(2.377)
.006*
(2.275)
Police Per 100k.000
(-.816)
.000
(-1.344)
.000
(.377)
Constant.110
(4.166)
.120
(3.912)
.060
(1.702)
Adj. R-Square.030.022.106

*p< .05
**p< .01
(Standard Error in Parentheses)