Attitudes Towards Health Care Teams Scale: A Confirmatory Factor Analysis

Purpose: The research dealing with interprofessional collaboration (IPC) has found that attitudes towards health care teams might be important factors in determining whether professionals work with other professionals as a team. To measure attitudes toward interprofessional collaboration, researchers employed the adapted Attitudes Toward Health Care Teams Scale (ATHCTS). However, the factor solutions of the adapted ATHCTS differed across studies as a two-factor structure or a three-factor structure. Cronbach's alphas of some subscales were unreliable at below 0.7. Some items had factor loadings less than 0.3 or double-loaded on two factors. The ATHCTS was completed only by undergraduate students. Thus, it was not supported that ATHCTS was valid to a sample of graduate professional students. The purpose of this study was to assess reliability and validity of the ATHCTS with graduate professional students, and to identify latent factors explaining the variation and covariation among the scale items. 

Methods: Data using a cross-sectional design were collected from 288 graduate students who were enrolled at an urban professional university in the Mid-Atlantic regions. The total sample consisted of 81.6% females and 17.6% males. A confirmatory factor analysis (CFA) was conducted to determine whether the adapted ATHCTS was a reliable and valid instrument in measuring attitudes of graduate professional students toward interprofessional health care teams. To test the construct validity of the scale, the discriminant validity between the ATHCTS and the Interdisciplinary Education Perception Scale (IEPS) was also examined.

Results: A two-factor CFA model (Curran et al., 2008) and a three-factor CFA model (Hayashi et al., 2012) were examined to determine which model better represents the sample, and the two-factor model was determined to be more appropriate to represent the sample (χ²₍₇₆₎ = 221.191; p < .001). This final version of the ATHCTS model was comprised of 14 items with two error covariances between items (χ²₍₇₄₎ = 170.678; p < .001). The factor loadings for 11 items on the quality of care subscale ranged from 0.55 to 0.79, and the loadings for three items on the time constraints subscale ranged from 0.76 to 0.88. Reliability of the ATHCTS with this study's samples was considered good (a = .82). Internal consistency of the quality of care subscale was high at 0.92, and reliability of the time constraints subscale was good at 0.86 in this study. To assess construct validity of the adapted ATHCTS, discriminant validity showed that the ATHCTS was adequately distinct from the IEPS.  

Implications: This study is valuable in that it is the first to assess the psychometric properties of the adapted ATHCTS by a confirmatory factor analysis. The CFA from this study provided evidence that the adapted ATHCTS with a two-factor structure is a valid scale with adequate model fit. This study suggests that the-two factor model should be tested across other groups such as undergraduate students, health care professionals, and persons at other training programs. In higher education, it is important to incorporate content related to quality of care and managing time constraints into coursework or training programs.