Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Presentation Icons

Recorded Session

Ticketed Session

164 Views

Adult- Health Psychology / Behavioral Medicine

Personalized, Adaptive, Physical Activity-Based Social Comparisons: A Real-Time Test of the Identification/Contrast Model to Explain Comparison Response

(PS1-95) Personalized, Adaptive, Physical Activity-based Social Comparisons: A Real-time Test of the Identification/contrast Model to Explain Comparison Response

Friday, November 15, 2024

8:30 AM – 9:30 AM EST

Location: Franklin Hall, Level 4, Marriott Downtown

Author(s)

Danielle Arigo, Ph.D.

Associate Professor of Psychology and Family Medicine
Rowan University
Glassboro, New Jersey, United States

Co-Author(s)

EL

Emmanuel Lapitan, B.A.

Ph.D. Student in Clinical Psychology
Rowan University
Glassboro, New Jersey, United States
IB

Iris Bercovitz, M.A.

Ph.D. Student in Clinical Psychology
Rowan University
Glassboro, New Jersey, United States
AS

Anisha Satish, M.A.

Ph.D. Student in Clinical Psychology
Rowan University
Glassboro, New Jersey, United States
AL

Andrea Lobo, Ph.D.

Professor of Computer Science
Rowan University
Glassboro, New Jersey, United States

Abstract Text: The power of social comparison, or self-evaluation relative to others, is often used to promote behavior change in both in-person and digital interventions. Social comparison is particularly popular in digital tools to promote physical activity (PA) such as web or mobile applications. Yet, emerging evidence shows that many users respond negatively to comparison, which can discourage engagement and limit the overall effectiveness of an intervention tool. Critically, what determines response to comparison in real time is unclear; the preponderance of evidence speaks only to global, person-level differences, which do not always reflect how processes unfold in the real world. Failure to understand what happens when people make comparisons (vs. who makes or responds well to them) could explain subsets of poor intervention responses to date. The Identification/Contrast (I/C) Model of social comparison proposes that perceived similarity to comparison targets determines response and could be useful for identifying specific cognitive or affective intervention targets to mitigate negative response. However, this model has not been tested in real time or in populations who particularly benefit from increased PA. The present study took an innovative step in this direction with a new PA web app, designed for insufficiently active women ages 40-60 with risk for cardiovascular disease, who each completed 7 days of app use (N=88, MAge=52, MBMI=31.9 kg/m2, 36% racial/ethnic minority). Each day, women viewed 1 of 3 types of comparison targets via a peer profile: very active (more PA than the user; upward targets), somewhat active (similar PA; lateral targets), or not very active (less PA; downward targets). They selected the type of target to view each day or indicated no preference, where the system showed the type of target they viewed least recently. To promote relevance to the user, peer profiles were personalized to present a woman with the user’s characteristics and adapted to show PA in range of the user’s from the previous day (steps and active minutes, collected from a PA monitoring device). Immediately after viewing the day’s profile, women rated perceived similarity and response to the target peer. Across all use days, no preference was selected most often (33%) and lateral targets were viewed most frequently (36%). Multilevel models showed >60% variability within-person for identification, contrast, and positive/negative affect. Consistent with the I/C Model, positive affect was greater on days when women identified with upward targets they viewed more so than usual; this association was diminished with lateral and downward targets (sr=0.15; p=0.01). Similarly, negative affect was greater on days when women contrasted against upward targets they viewed more so than usual; this association was more complicated with lateral and downward targets (sr=0.14; p=0.06). Thus, real-time responses to personalized, adaptive comparisons generally follow the I/C Model and warrant further investigation, as I/C processes could represent novel intervention targets. Advancing our understanding of comparison exposure, cognitive/affective responses, and PA in real time will improve the effectiveness of digital PA tools that activate social comparison.