If the first part of the research followed an exploratory-descriptive pattern, the last one focused on an explanatory approach. It was based on cross-sectional design, by analyzing a sample at one point in time (Bachman, & Schutt, 2003 Maxfield & Babbie, 2008).
The aims of the study focused on: identifying the factors related to the public image of the police, the assessment of the contribution of a number of variables in shaping the public image of Romanian law enforcers (trust in the police, the perception of legitimacy, institutional trust, personal experience with police, prior victimization, media exposure, social cohesion and some demographic …show more content…
Perceived police legitimacy and trust in the police. These two scales were drawn from the European Social Survey (2011) Round 5 Module on Trust in the Police & Courts, proposed by Jonathan Jackson, Mike Hough, Stephen Farrall, Jan de Keijser & Kauko Aromaa. According to the scientific requirements, before their actual use, all instruments must be adapted and validated on local population. Consequently, the two scales were the focus of a pilot study for checking their statistical parameters and the correspondence with the original form.
b. Perceived Social cohesion - assessed through the Social Cohesion Scale developed by Fone, Dunstan, Lloyd Williams, Watkins and Palmer (2007), an adapted version of the Community Cohesion Scale (Bukner, 1988). Prior to its use, the questionnaire has undergone a process of adaptation and validation to Romanian population.
c. Social Desirability - assessed through a shortened and adapted version on Romanian population (Sârbescu Rusu & Costea, 2012) of the Marlow-Crowne Scale on Social Desirability (MCSDS, Crowne & Marlowe, 1960).
d. Institutional trust: proposed by the researcher and conceptualized dimensional.
e. Personal experience with the police, in terms of prior interactions, type and frequency
f. Media exposure
g. Prior victimization
h. Control variables: gender, age and …show more content…
Each of these three variables played the role of the criterion, while predictors were the factors that have obtained significant correlations with the criterion.
The preliminary analysis showed the data conformity with the multiple regression assumptions about normality, linearity and homoscedasticity; the tolerance coefficients above .1 (Sava, 2004) and VIF values (Variation Inflation Factor) below 10 indicated the lack of data multicollinearity. All predictors statistically correlated with the criterion, indicating the adequacy of using hierarchical linear