A Convergent and Divergent Validity Study of Scores on the Family Bondedness Scale

Background and Purpose: The Family Bondedness Scale (FBS) is designed to measure the degree to which pet owners are emotionally attached to their pets in a manner commensurate with their attachment to human family members. Although pets are often viewed as family members, there is limited evidence on the validity of measures assessing this type of bond. Especially lacking is evidence for convergent and divergent validity. This study is the first to investigate the convergent and divergent validity of FBS scores and did so in relation to scores on the Lexington Attachment to Pets Scale (LAPS).

Methods: This study used secondary data from a needs assessment survey of 415 pet owners conducted by Meals on Wheels America. Confirmatory factor analysis with WLSMV estimation was used. First, unidimensional models were tested for the FBS and LAPS. Then, a hypothesized two-factor convergent/divergent validity measurement model was tested. Indices specific to convergent and divergent validity that were used included average variance extracted (AVE), construct reliability (CR), the dis-attenuated correlation between FBS and LAPS latent variables, and the Fornell-Larker criterion (FLC). Sentiment analyses of FBS and LAPS items were done investigating the emotional bonding implied by item wordings.

Results: The convergent/divergent validity model fit the data well, with model chi-square, χ²(494) = 754.2, p < 0.001, and relative Chi-square, 1.538. The CFI was 0.99, and the TLI 0.99. The RMSEA was 0.036 (90% CI, 0.031, 0.041), p-close = 1.00, and SRMR = 0.054. Results for convergent validity were AVE > 0.50, and CR > 0.90, for both FBS and LAPS, and for divergent validity, FLC > 0 for both FBS and LAPS, and the dis-attenuated correlation between FBS and LAPS latent variables was 0.69. These results were consistent with established criteria for convergent and divergent validity. Sentiment analyses suggested both FBS and LAPS measure emotional attachment to pets, and that the FBS measures a deeper level of attachment than the LAPS.

Conclusions and Implications: These findings provide an evidentiary basis for the use of the FBS as a measure of family bondedness with pets in social work practice and research and adds to the literature on human-animal bond measurement. Practitioners in social work and related fields have recommended the inclusion of pets, and family bondedness with pets, in assessment. The FBS can be used by social workers, especially veterinary social workers, in clinical assessment as well as in research on the role pets and family bondedness with pets play in health and mental health research. It can also be used in research on the outcomes of pet assisted therapy and the extent family bondedness may play as mediating or moderating variables.