A Confirmatory Factor Analysis of the Lexington Attachment to Pets Scale

摘要: The purpose of this study was to investigate the extent which word “pet” provided valid results in Lexington Attachment Pets Scale (LAPS) (Johnson, Garrity, & Stallones, 1992). A multiple-groups confirmatory factor analysis, using maximum likelihood (ML) estimation, conducted test hypothesized structures specified measurement model. Respondents Original version (i.e. used pet) and Personalized embedded canine’s name as typed by participants) made up two groups. Features an electronic survey were necessary for personalization systematic alternative assignment LAPS versions. snowball sampling method utilized mail invite self-selected participants meeting following criteria: 18 years or older; lived United States; at least one dog living inside/outside their home whom they are responsible some time. According collected human canine demographics, (n = 1,854) =1849) groups appeared be statistically equivalent (N 3,703). model generated a χ value 4130.242, with 264 degrees freedom probability less than .001 (p <.001), suggestive lack fit. However, goodness-of-fit indices consideration. Comparing CFI (.95 vs. .862), TLI .840) RMSEA (<.05 .063) cut off values from respectively, reveals again, inadequate These imply that not consistent data precluded specific tests differential validity. While it inadvisable further examine validity associated linguistic differences scale items, critical information nonetheless identified. current study, conceptualization “pet attachment” (sic) may valid. Until research provides stronger evidence, use could produce lead invalid inferences. original quantification relationships is still work progress. Moreover, psychometric critically before LAPS.