Demographic Portrait and Comparison of Means Analyses
Table 1a reveals that participating students in this study were predominantly female
(60%, n = 151), non-White (52%, n = 132), and had a cumulative grade point average of 2.5
or higher (87%, n = 199). No significant perception differences were discovered for gender
and reported grade point average, but analysis did determine that non-White students
reported significantly higher agreement with increased police misconduct (t = 3.37, p < .01).
A significant portion of the sample also consisted of students who were majoring in various
academic disciplines (74%, n = 193) and were non-seniors (57%, n = 148). No significant
perception differences emerged along academic discipline lines, but significant perception
differences were discovered between the two levels of academic maturation (t = 2.23, p
< .05), with non-seniors reporting higher corruption perceptions than seniors. A majority of
respondents also reported not having completed an ethics course (62%, n = 160). Lastly, a
majority of respondents reported having no family members in law enforcement (58%, n =
150). Of those who did have such a family member, though, most (58%, n = 51) were either
a father or uncle.
Table 1a
Demographic Frequencies and Means Comparisons
________________________________________________________________________
Variable N P M SD SE Test p
________________________________________________________________________
Gender
Male 100 40 3.35 0.90 .09 t = 1.64 .10
Female 151 60 3.54 0.92 .08
Race
t = 3.37
**
White 123 48 3.24 0.85 .08 .01
Non-White 132 52 3.63 0.97 .08
Student GPA
<2.5 30 13 3.53 0.97 .18 F = 1.51 .22
2.5-2.99 100 44 3.47 0.94 .10
3.0+ 99 43 3.27 0.91 .10
Student Major
Forensic Science 67 26 3.58 0.81 .10 t = -1.42 .16
Other 193 74 3.41 0.97 .07
Student Classification
t = 2.23*
Senior 112 43 3.31 0.98 .09 .03
Non-Senior 148 57 3.57 0.85 .07
Completed Ethics Course
Yes 98 38 3.33 1.05 .10 t = 1.72 .09
No 160 62 3.53 0.84 .07
Family in Law Enforcement
Yes 109 42 3.46 0.93 .09 t = -0.16 .87
No 150 58 3.44 0.93 .08
Family Relation in LE
Father 20 23 3.50 1.00 .22 F = 0.82 .97
Uncle 31 35 3.42 0.89 .16
Cousin 17 19 3.47 0.94 .23
Other 20 23 3.55 0.99 .22
________________________________________________________________________
*Percentages are rounded to the nearest whole number.
*Means range from 1(Strongly Disagree) to 5(Strongly Agree).
*p<.05 ** p<.01.
Neither of these variables, however, produced any significant perception differences.
In addition to descriptive data (see Table 1a), Table 1b provides analyses for each
significant perception variable. Six (of twelve) perception variables were statistically
significant, with five of the six variables significant at the .01 alpha level. Specifically, a
large portion of respondents agreed (43%, n = 113) or were neutral (35%, n = 90) that
police officers bend rules during daily operations. A one-way analysis of variance test
produced significant mean differences (F = 7.71, p < .01) between those who strongly
disagreed (M = 1.00, SD = 0.00) and all other respondents. Likewise, a similar portion of
the sample indicated some level of agreement that police discretion significantly
influences unethical decision-making (52%, n = 135). Tukey analysis revealed that
respondents who strongly agree (M = 3.95, SD = 1.06) that police discretion negatively
influences decision-making differed significantly from all other groups (F = 5.39, p <
.01). In addition, a majority of respondents agreed that racial profiling is of equal concern
(71%, n = 183). A one-way analysis of variance test indicated that respondents who strongly
agreed (M = 3.79, SD = 0.89) held significantly higher mean scores than those who
disagreed and agreed (F = 4.82, p < .01). Respondents also tended to be neutral (39%, n =
103) or in some agreement (35%, n = 92) that criminal justice courses teach valuable lessons
not taught in the police academy. Tukey analysis indicated respondents who disagreed (M =
2.73, SD = 1.03) differed significantly from all other groups (F =
3.91, p < .01). Most respondents agreed (38%, n = 98) or strongly agreed (21%, n = 55)
that a college degree promotes a more positive image in the community. Significant
mean differences were discovered, revealing that those who agreed (M = 3.24, SD =
0.81) differed most significantly from respondents who were neutral response
Table 1b
Perceptions of Misconduct
________________________________________________________________________
Variable N P M SD SE Test p
________________________________________________________________________
Unethical decision-making creates negative perception with community
F = 2.87*
Strongly Disagree 3 1 2.33 1.15 .67 .02
Disagree 5 2 3.00 1.00 .45
Neutral 20 8 3.35 0.75 .17
Agree 97 37 3.32 0.91 .09
Strongly Agree
137
52 3.60 0.93 .08
Discretion influences unethical decision-making
F = 5.39
**
Strongly Disagree 6 2 2.83 1.17 .48 .00
Disagree 32 12 3.03 0.89 .16
Neutral 87 33 3.43 0.79 .08
Agree 97 37 3.46 0.90 .09
Strongly Agree 38 15 3.95 1.06 .17
Racial profiling is a problem
F = 4.82**
Strongly Disagree 8 3 3.13 1.13 .39 .00
Disagree 17 7 2.88 0.93 .23
Neutral 53 20 3.38 0.90 .12
Agree
111
43 3.37 0.89 .09
Strongly Agree 72 28 3.79 0.89 .11
Police officers bend the rules
F = 7.71**
Strongly Disagree 2 <1 1.00 0.00 .00 .00
Disagree 12 5 2.83 0.94 .27
Neutral 90 35 3.32 0.83 .09
Agree
113
43 3.51 0.92 .09
Strongly Agree 43 17 3.81 0.91 .14
CJ courses teach valuable lessons
F = 3.91**
Strongly Disagree 6 2 4.33 0.82 .33 .00
Disagree 15 6 2.73 1.03 .27
Neutral
103
39 3.43 0.86 .09
Agree 92 35 3.52 0.83 .09
Strongly Agree 45 17 3.47 1.12 .17
Table 1b (continued).
________________________________________________________________________
Variable N P M SD SE Test p
________________________________________________________________________
Police with college degree have more positive image
Strongly Disagree 2 <1 3.50 2.12 1.50 F = 3.57** .01
Disagree 23 8 3.26 1.01 .21
Neutral 83 32 3.73 0.81 .09
Agree 98 38 3.24 0.81 .08
Strongly Agree 55 21 3.49 1.10 .15
Percentages are rounded to the nearest whole number.
Means range from 1(Strongly Disagree) to 5(Strongly Agree).
*p < .05, ** p < .01
(M = 3.73, SD = 0.81) (F = 3.57, p < .01). Lastly, a majority of respondents reported
strong agreement (52%, n = 137) that unethical decision making creates a negative
perception with the community (F = 2.87, p < .05); however, no significant mean
differences emerged upon comparison of groups.
Factor Analysis
Table 2 illustrates two factors produced from the 14 perception variables in this
study: one scale (Cronbach’s alpha = .81) comprised of six variables about college
education’s positive influence on policing and professionalism, and a second scale
(Cronbach’s alpha = .37) comprised of two variables about unethical actions by officers,
bending rules and racial profiling. Due to the second scale’s low reliability coefficient, it
was excluded from all regression analysis.
Correlation Analyses
Table 3 provides the results of an intercorrelation matrix presenting the associated
correlations between the dependent variable and (1) all significant demographic variables
and (2) significant perception variables. The variable unethical actions produced the
Table 2
Exploratory Factor Analysis for Continuous Independent Variables
________________________________________________________________________
Factor Loadings
________________________________________________________________________
Variable College Unethical
Education Actions
________________________________________________________________________
CJ college courses teach valuable lessons in ethics… .41
Officers with college degree have more positive image… .57
Officers with college degree exhibit less racial profiling. .62
Officers with college education better understand law. .65
A college degree enhances professionalism in policing. .80
A college degree enhances ethical conduct in policing. .82
Racial profiling is a problem within policing. .56
Police officers bend the rules during daily operations. .60
Eigenvalues 2.63 1.11
Cronbach’s Alpha 0.81 0.37
Cumulative % of Variance 26.28 37.35
strongest correlation with the dependent variable (r = .28, p < .01).
Cross Tabulation Analyses
Tables 4a and 4b provide results of cross-tabulations between all demographic variables
which exhibited significant mean differences (see Table 1). Table 4a indicates
overabundance (Cramer’s V = .22) of White male students and an overrepresentation of
non-White female students (χ2 = 11.39, p < .01). Additionally, a strong relationship was
revealed between student race and their respective grade point averages (Cramer’s V
= .22). White respondents predominantly held a 3.0 grade point average or higher, while
Table 3
Intercorrelation Matrix for Demographic Variables and Significant Perception Variables
________________________________________________________________________
Variables 1 2 3 4 5 6 7 8 9 10 11
________________________________________________________________________
1. Corruption (DV) 1.00
2. Student Gender -.10 1.00
3. Student Race -.21**
.22
**
1.00
4. Student GPA -.11 .02
.22
**
1.00
5. Student Major .08
-.24
**
.06 .11 1.00
6. Student Classification -.14*.05 .12
-.16
**
-.26** 1.00
7. Ethics Course -.11 .06
-.04
-.13
**
-.22
**
.40
**
1.00
8. Decision-making .19** .07 .00 .03 .00 .11 .04 1.00
9. Discretionary Power .26** -.05 -.11 -.05 -.07 .11 .07 .18** 1.00
10. Unethical Actions .28** -.13*
-.34
**
-.20
**
.04 .05 .05 .28** .28** 1.00
11. College Ed Perceptions -.10 .10
.18
**
.05 -.04 .08 .07 .06 .02 -.091.00
*p < .05. ** p < .01.
aPolice corruption is increasing (DV) measured on 5 point scale (5= Strongly Agree, 1= Strongly Disagree).
Coding: Student Gender (Female=0, Male=1), Student Race (White=0, Non-White=1), Student GPA (<2.50=0, 2.5-2.99=1, 3.0+ =
2), Student Major (Other=0, Forensic Science=1), Student Classification (Non-Senior=0, Senior= 1), Ethics Course Completion
(No=0, Yes=1)
Variables 8-11 measured on 5 point scale (5= strongly agree, 1= strongly disagree)
Note: Variables 10 and 11are scaled factors (see Table 2).
students who were non-White tended to possess grade point averages of 2.50 to 2.99 (χ2
= 10.75, p < .01). Lastly, the findings indicate that a moderate relationship existed
between student race and whether respondents reported having a family member in law
enforcement (Cramer’s V = .13). White respondents tended to have fewer familial
relationships to law enforcement (χ2 = 4.01, p < .05).
Table 4b indicates a moderate relationship between student classification and
grade point average (Cramer’s V = .16). The findings suggest that non-seniors held grade
point averages of 3.0 or higher, while seniors tended to hold grade point averages above
3.0 (χ2 = 6.08, p < .05). Moreover, results indicated a strong relationship (Cramer’s V =
.25) between student race and major. Here, findings indicate an overabundance of
Table 4a
Contingency Table for Race against Other Independent Demographic Variables
________________________________________________________________________
Variable White Non-White χ2
________________________________________________________________________
Gender (Cramer’s V = .22)
11.39**
Male 48 (E) 49(E)
61 (O) 36(O)
Female 73
(E)
74(E)
60 (O) 87(O)
Student GPA (Cramer’s V = .22)
10.75**
<2.5 15
(E)
15(E)
9 (O) 21(O)
2.5-2.99 49
(E)
51(E)
44 (O) 56(O)
3.0+ 48
(E)
49(E)
59 (O) 38(O)
Student Major (Cramer’s V = .06)
FSC 32
(E)
34(E) 0.82
35 (O) 31(O)
Other 91
(E)
98(E)
88 (O) 101(O)
Student Classification (Cramer’s V = .12)
Senior 54
(E)
57(E) 3.64
61 (O) 49(O)
Non-Senior 70
(E)
74(E)
62 (O) 81(O)
Completed Ethics Course (Cramer’s V = .04)
Yes 47
(E)
49(E) 0.48
44 (O) 52(O)
No 75
(E)
80(E)
78 (O) 77(O)
Family in Law Enforcement (Cramer’s V = .13)
4.01*
Yes 51
(E)
54(E)
43 (O) 62(O)
No 71
(E)
76(E)
79 (O) 68(O)
Table 4a (continued).
________________________________________________________________________
Variable White Non-White χ2
________________________________________________________________________
Family Relation in LE (Cramer’s V = .22)
Father 8 (E) 11(E) 4.30
10 (O) 9(O)
Uncle 13
(E)
18(E)
16 (O) 15(O)
Cousin 7
(E)
9(E)
4 (O) 12(O)
Other 9
(E)
11(E)
9 (O) 13(O)
*p < .05; ** p < .01.
Note: E = Expected Count; O = Observed Count.
seniors who majored in other disciplines, while non-seniors tended to be forensic
science majors (χ2 = 16.28, p < .01). Lastly, a strong relationship (Cramer’s V = .39)
existed
between student classification and whether respondents had completed a formal ethics
course. Specifically, seniors were more likely to have completed an ethics course, while
Table 4b
Contingency Table for Classification against Other Independent Demographic Variables
________________________________________________________________________
Variable Senior Non-Senior χ2
________________________________________________________________________
Gender (Cramer’s V = .05)
Male 43 (E) 57(E) 0.63
46 (O) 54(O)
Female 64
(E)
85(E)
61 (O) 88(O)
Student Race (Cramer’s V = .12)
White 54 (E) 70(E) 3.64
61 (O) 62(O)
Non-White 57 (E) 74(E)
49 (O) 81(O)
Table 4b (continued).
________________________________________________________________________
Variable Senior Non-Senior χ2
________________________________________________________________________
Student GPA (Cramer’s V = .16)
6.08*
<2.5 14 (E) 16(E)
18 (O) 11(O)
2.5-2.99 46 (E) 53(E)
50 (O) 49(O)
3.0+ 46 (E) 53(E)
38 (O) 61(O)
Student Major (Cramer’s V = .25)
16.28**
FSC 29 (E) 38(E)
15 (O) 52(O)
Other 83 (E) 108(E)
97 (O) 94(O)
Completed Ethics Course (Cramer’s V = .39)
40.69**
Yes 42 (E) 55(E)
67 (O) 30(O)
No 70 (E) 89(E)
45 (O) 114(O)
Family in Law Enforcement (Cramer’s V = .02)
Yes 47 (E) 62(E) .06
48 (O) 61(O)
No 64 (E) 84(E)
63 (O) 85(O)
*p < .05; ** p < .01.
Note: E = Expected Count; O = Observed Count.
fewer non-seniors had completed the course (χ2 = 40.69, p < .01).
Ordinary Least Squares Regression Analyses
Tables 5a and 5b provide results of the ordinary least squares regression analyses
which examined the relative influence of each independent variable on the dependent
variable. Table 5a includes all perception and demographic variables in a forced entry
model, while Table 5b included only the variables that most closely predicted the
dependent variable. Completion of an ethics course was included in both models
despite its lack of significance (see Table 1).
As shown in Table 5a, one fifth (R2 = .20) of the variance in the dependent
variable can be explained through examination of all 14 independent variables (F = 3.25,
p < .01). Results indicate that only two variables significantly predict the dependent
variable. The most important contributor (β = .26, p < .01) pertained to students who
perceive police discretion to influence unethical conduct. Students who agree with the
negative influence of police discretion were more likely (b = .26) to agree that corruption
is increasing. Meanwhile, student classification was the second largest contributor to the
variance of the dependent variable (β = -.15, p < .05). Specifically, senior students were
significantly less likely to perceive police corruption to be increasing (b = - .28) when
compared with non-senior students.
Table 5a
Regression for Police Corruption Perception Subset
________________________________________________________________________
Variables b SE b β t VIF
________________________________________________________________________
Gender -.12 .13 -.07 -0.92 1.18
Race -.17 .14 -.09 -1.22 1.39
GPA -.12 .09 -.09 -1.32 1.11
Major .11 .15 .05 0.74 1.23
Classification -.28 .14 -.15 -2.02*1.35
Completed Ethics Course -.09 .14 -.05 -0.72 1.25
Family in LE -.02 .12 -.00 -0.12 1.07
Unethical Decision-making .12 .08 .10 1.42 1.25
Corruption Inhibits .02 .07 .02 0.29 1.22
Duration of Service .03 .07 .03 0.35 1.24
Discretion Influences .26 .07 .26 3.76** 1.12
Officer College Education -.13 .09 -.09 -1.32 1.19
Racial Profiling .03 .07 .03 0.41 1.42
Table 5a (continued).
________________________________________________________________________
Variables b SE b β t VIF
________________________________________________________________________
Officers Bend Rules -.09 .08 .09 1.19 1.25
Constant
2.30 .57 4.02**
R=.44 R2 = .20 F = 3.25**
**p < .01. *p < .05.
DV: Police corruption is increasing among today’s police officers (1=Strongly Disagree to 5= Strongly Agree).
Coding: Gender (1 = Male, 0 = Female), Race (1 = White, 0 = Non-White), GPA (0 = <2.50, 1 = 2.50-2.99, 2 = 3.0+), Major (1
= Forensic Science, 0 = Other), Student Classification (1 = Senior, 0 = Non-Senior), Completed Ethics Course (1= Yes,
0 = No), Family in Law Enforcement (1 = Yes, 0 = No), All other variables (1 = Strongly Disagree to 5 =
Strongly Agree)
Table 5b provides results for a best fit regression model that predicts the dependent
variable ‘police corruption is increasing among today’s police officers’. As indicated, 19%
of the total variance can be explained through examination of six independent variables.
Results indicate that student perceptions regarding discretionary powers (β = .28, p < .01) is
the most influential predictor. Essentially, students who agree (b = .27) that discretion
influences future unethical conduct perceive police corruption to be a growing issue.
Meanwhile, regression results indicate that the influence of unethical decision-making was
the second most influential predictor (β = .18, p < .01). Here, students who agreed (b = .20)
with the perception measuring the influence of unethical decision-making also perceive
police corruption to be increasing. Furthermore, results indicate that (β = -.17, p < .01)
student race was the third most influential predictor, suggesting that non-white students are
more likely to agree (b = -.32) that police corruption is increasing in contrast to their non-
white counterparts. Lastly, the results of the model provide that student classification (β =
-.15, p < .05) was the least significant predictor, in that seniors are less likely (b = -.27) to
perceive police corruption to be increasing; non-seniors are more apt to agree with the
notion of the increased problem. Table 5b
Best Fit Model for Police Corruption Perception Subset
________________________________________________________________________
Variables b SE b β t VIF
________________________________________________________________________
Race -.32 .11 -.17 -2.87** 1.07
Classification -.27 .12 -.15 -2.22*1.25
Completed Ethics Course -.11 .12 -.06 -0.93 1.22
Unethical Decision-making .20 .07 .18
2.98
**
1.04
Discretion Influences .27 .06 .28
4.61
**
1.05
Officer College Education -.07 .08 -.06 -0.92 1.04
Constant 2.25 .42
5.31
**
R=.43 R2 = .19 F = 8.98**
**p < .01. *p < .05.
DV: Police corruption is increasing among today’s police officers, (1=Strongly Disagree to 5= Strongly Agree).
Coding: Race (1 = White, 0 = Non-White), Student Classification (1 = Senior, 0 = Non-Senior), Completed Ethics Course
(1= Yes, 0 = No), All other variables (1 = Strongly Disagree to 5 = Strongly Agree)
