Initial Validation of the Constructs for a Newly Created Pet Attachment Scale

Studies indicate that pets have tremendous influence on pet owners’ decision to receive social work services. For example, older adults with pets had resisted evacuation efforts when pets were in jeopardy during disasters (Torgusen & Kosberg, 2006). Battered women are often reluctant to receive social work services; these women hesitate to leave their pets with their abusers (Faver & Strand, 2003a, 2003b). Paying attention to the roles pets play in clients’ lives is critical to providing clients with maximum help. In order to reach these clients, social workers should recognize that clients have various needs that social workers may not have explored in the assessment process. The theoretically and statistically robust pet attachment scale will help social work practitioners/researchers assess its impact of the client-pet relationship on the decision of the use of social work services in an emergency.

Over the past several decades, scholars in the human-animal bond studies have attempted to measure the strength/quality of people’s attachment to their pets. However, they often lack of construct validity. A theory-based pet attachment scale, with 24 items, was newly created for a dissertation study in accordance with attachment theory. This pet attachment scale consists of four sub-constructs: Proximity Seeking, Separation Distress, Safe Haven, and Secure Base (Ainsworth, 1989).

Since the pet attachment scale was developed with an a priori hypothesis of the relationship among its items, a series of confirmatory factor analysis (CFA) using Mplus version7, was conducted for investigation of construct validity. The CFA was completed in a sample of social workers in a northeastern chapter of NASW (N=675, white=90.5 %, female=82.0 %, Mean age=50.3 (SD=10.8)). The best-fit model with the data set was examined through comparison with two-, three-, four-, and five- factor structures. The results suggest that the five-factor model with three original sub-constructs (Proximity Seeking, Separation Distress, Secure Base) and two ramifications from Safe Haven (“Pets’ Responsiveness to Owner’s Needs”, and “Sense of Security”), was statistically better than other models. Further analysis/implications using data from diverse populations are under way to validate these results in-depth. These studies will contribute to development of psychometric research and an early intervention with clients who view their pets as important family members.

Ainsworth, M. D. S. (1989). Attachment beyond infancy. American Psychologist, 44, 709-716.

Faver, C. A. & Strand, E. B. (2003a). Domestic violence and animal cruelty: Untangling the web of abuse. Journal of Social Work Education, 39, 237-253.

Faver, C. A. & Strand, E. B. (2003b). To leave or to stay? Battered women’s concern for vulnerable pets. Journal of Interpersonal Violence, 18(12), 1367-1377.

Torgusen, B. L. & Kosberg, J. I. (2006). Assisting older victims of disasters. Roles and responsibilities for social workers. Journal of Gerontological Social Work. 47(1/2), 27-44.

(Primary Presenter)

Aki Sato earned her PhD degree from the University of Connecticut School of Social Work in 2011. Her dissertation title was "Social workers' attachment to their pets, organizational structures, and their impact on professional assessments regarding the roles pets play in clients' lives." Since June 2012, she works as a visiting scholar at the University of Kansas Center for Research Methods and Data Analysis. During 2007-2010 and Spring 2012, she worked as a research specialist in the Connecticut Department of Mental Health and Addiction Services. In Summer 2004, she worked as a farm intern at Green Chimneys.

(Co-author)

Dr. Todd Little is professor of the University of Kansas (KU) Department of Psychology, and director of KU Center for Research Methods and Data Analysis. He has studied developmental changes in how children's/adolescents' action-control processes influence (a) their adjustment/achievement in school settings, (b) their peer and friendship relationships, (c) their reasons for aggression and victimization, and (d) their ability to cope with challenging and stressful events. He provides graduate-level training in an array of multivariate statistical techniques, and works examining the use of structural equation modeling techniques as a general data analytic approach to studying individual, developmental, and sociocontextual differences.