The Vaccine Hesitancy Scale: Psychometric Properties and Validation

"A standardized, validated measurement tool of vaccine hesitancy beliefs would aid in the advancement of research and immunization policy."

There is concern that public confidence in vaccines is decreasing and that anti-vaccine movements are becoming stronger. In 2015, Heidi Larson and colleagues from the SAGE Working Group on Vaccine Hesitancy developed the Vaccine Hesitancy Scale (VHS) in an effort to standardise the measurement of vaccine hesitancy. (See Related Summaries, below.) The present paper examines the VHS's structure and internal consistency, construct validity, criterion validity, and socio-demographic differences in parents' vaccine hesitancy. This is in line with the field of psychometrics, which is concerned with the theory and technique of psychological measurement, including the measurement of knowledge, abilities, attitudes, and personality traits.

Using a cross-sectional design, the researchers collected self-reported survey data from a national sample of 3,779 Canadian parents from August to September 2016 through an online questionnaire. They used exploratory and confirmatory factor analysis to identify latent constructs underlying parents' responses to 10 VHS items (response scale 1-5, with higher scores indicating greater hesitancy). In addition to the VHS, measures included socio-demographics items, vaccine attitudes, parents' human papillomavirus (HPV) vaccine decision-making stage, and vaccine refusal.

Exploratory and confirmatory factor analysis revealed that a 2-factor structure best explained the data, consisting of "lack of confidence" (mean (M) = 1.98, standard deviation (SD) = 0.72) and "risks" (M = 3.07, SD = 0.95). Significant Pearson correlations were found between the scales and related vaccine attitudes. ANOVA analyses found significant differences in the VHS sub-scales by parents' vaccine decision-making stages (p<.001). The VHS sub-scales were found to be associated with HPV vaccine refusal and refusing another vaccine (p<.001). Socio-demographic differences in the VHS were found; however, effect sizes were small.

The construct validity of the VHS sub-scales was emphasised through a medium to large relationship with related vaccine attitudes, including a negative relationship with trust in vaccines and a positive relationship with vaccine conspiracy beliefs. This indicates that the VHS is appropriately measuring the construct of interest. Furthermore, a medium to large relationship between the VHS sub-scales and vaccine refusal indicates that the VHS has criterion validity.

Among the researchers' reflections on the findings is that, though parents overall were not very vaccine hesitant, more parents endorsed vaccine "risks" associated with new vaccines and adverse reactions (M = 3.07) compared to "lack of confidence" (M= 1.98). This result corresponds to findings from a Canadian survey that found that few parents did not trust vaccine-related information provided by doctors (7%) and public health officials (12%), whereas a larger number of parents believed or were uncertain about a link between vaccines and autism (28%).

As reported here, Larson et al. (2015) acknowledged that a limitation of the VHS was that the identified questions do not address all determinants of vaccine hesitancy; therefore, they recommended that additional questions be developed and validated. The authors of the present study hope that the results will help guide future development of the VHS. It would also be helpful, they suggest, for future research to compare the psychometric properties of the VHS to other vaccine attitude scales.