College Students' Television Friends: Parasocial Relationships as Attachment Bonds

Noelle T Liwski, Purdue University


This study examined if parasocial relationships (PSR) with favorite TV characters can function as supplementary attachment bonds, as defined by the four theoretical components of attachment: proximity seeking behaviors, separation protest behaviors, safe haven function, and secure base function (Bowlby, 1969/1982; Hazan, Gur-Yaish, & Campa, 2004). To empirically test if PSRs can fulfill attachment-related functions, I examined the relatedness of PSR with a favorite TV character (i.e., TV-PSR), attachment anxiety and avoidance, loneliness, and stress. Undergraduate and graduate college students ( N = 226) completed (a) demographic items, (b) the WHOTO scale (Fraley & Davis, 1997) adapted for this study, (c) Revised Parasocial Interaction Scale (PSI-R; Rubin & Perse, 1987), (d) Experiences in Close Relationships Inventory - Revised (ECR-R; Fraley, Waller, & Brennan, 2000), (e) Revised UCLA Loneliness Scale (Russell, Peplau, & Cutrona, 1980), and (f) Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983). For this correlational study, I examined five hypotheses. Hypothesis 1 (H1) stated that Romantic Partner Proximity would be negatively correlated with TV Character Proximity. Hypothesis 2 (H2) stated that PSR would be negatively correlated with loneliness and with stress. Hypothesis 3 (H3) predicted that attachment anxiety and avoidance would positively predict both (a) loneliness and (b) stress, and (c) anxiety would uniquely and positively predict TV-PSR, whereas avoidance would negatively predict PSR. Hypothesis 4 (H4) predicted TV-PSR would moderate the positive correlations between attachment anxiety and avoidance and (a) loneliness and (b) stress; and Hypothesis 5 (H5) predicted TV-PSR would moderate the negative correlation between Peer/Friend Proximity and loneliness. In terms of results, Pearson correlation indicated that H1 was not supported; Romantic Partner Proximity was not significantly and negatively related to TV Character Proximity. Pearson correlation indicated that H2 was not supported; TV-PSR was significantly and positively, rather than negatively, related to loneliness and stress. Multiple regressions indicted that H 3 was partially supported. Attachment anxiety and avoidance positively predicted loneliness (H3a), but only anxiety significantly positively predicted stress (H3b); and anxiety uniquely and positively contributed to TV-PSR, whereas avoidance significantly negatively predicted TV-PSR (H 3c). Hierarchical regressions indicated H4 was not supported; TV-PSR did not reduce the positive associations between attachment anxiety and avoidance dimensions and loneliness and stress. Additionally, TV-PSR was significantly positively correlated with loneliness and stress, but when attachment anxiety and avoidance are controlled, the ability of TV-PSR to predict loneliness and stress became non-significant. Hierarchical regression also indicated Hypothesis 5 (H5) was not supported; TV-PSR did not reduce the negative association between Peer/Friend Proximity and loneliness. In addition, in supplemental analyses, Pearson correlations revealed that TV-PSR (as measured by the PSI-R scores) was significantly positively correlated with four items measuring attachment theory's proximity, separation, safe haven, and secure base components, thereby providing convergent validity evidence for the PSR measure (i.e., the PSI-R scores). This research contributes to attachment and PSR literature by finding that (a) college students form parasocial relationships, that is, one-sided cognitive-affective bonds with favorite TV characters, and (b) such bonds are associated with attachment-related feelings. Implications for future research and clinical practice are discussed.




Pistole, Purdue University.

Subject Area

Counseling Psychology|Mass communications

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server