Date of Award


Document Type


Degree Name

Doctor of Psychology (PsyD)


Graduate Department of Clinical Psychology

First Advisor

Rodger K. Bufford, Ph.D.

Second Advisor

Nancy Thurston, Psy.D.

Third Advisor

Ken Logan, Psy.D.


This dissertation examined the relationship between self-reported political ideology and the relational variables of attachment, shame, and grace. Previous research on political ideology has focused on conservatism, as well as morality and personality differences across the political spectrum. This study combined two samples of 155 adult participants gathered through Qualtrics and 79 undergraduate students. Each participant completed a demographics questionnaire, Attachment Style Questionnaire (ASQ; Feeney et al., 1994), Dimensions of Grace Scale (DGS; Bufford et al., 2017), Duke University Religion Index (DUREL; Koenig et al., 1997), and Internalized Shame Scale (ISS; Cook, 1988). Additionally, 19 of the undergraduate students completed the Thurston-Cradock Test of Shame (TCTS; Thurston & Cradock, 2009). Prior to analysis participants were grouped into five categories based on self-reported social political ideology (GSPI): Very Liberal, Liberal, Neither Liberal nor Conservative, Conservative, and Very Conservative. Internal consistency was evaluated for each measure used and correlations were calculated to examine the relationship among GSPI, select demographic variables, and each measure. Next, a regression analysis was conducted and established that shame and attachment characteristics do not account for a significant amount of variance in GSPI. However, this study found that age, education, and religiosity did account for significant variance in GSPI. Following this, an ANOVA was completed and found significant between group differences for DGS: Experiencing God’s Grace, DGS: Costly Grace, DGS: Grace from Others, DGS: Grace to Others, ASQ: Avoidance, DUREL: Organizational Religiosity, DUREL: Non-organizational Religiosity, and DUREL: Intrinsic Religiosity. Significant within group differences were also discovered across each of the measures, with the exception of ASQ: Avoidance. Lastly, a cluster analysis was conducted and established that political ideology can be viewed from a two, three, six, or eight-cluster perspectives, with eight clusters likely providing the most accurate depiction of group differences. Results of the study indicate that individuals across the political spectrum do not differ in shame and attachment characteristics and do display significant differences in religiosity and grace. In addition, political ideology may be more accurately represented as eight clusters rather than a liberal-conservative dichotomy that is used in much of political ideology research.

Included in

Psychology Commons