The Household Risk Perception instrument and the Self-Efficacy in Environmental Risk Reduction instrument: Psychometric testing using principal component analysis

摘要

Aim: To report the psychometric testing of the Household Risk Perception and Self-Efficacy in Environmental Risk Reduction instruments using principal components analysis. Background: There are limited instruments available to test household risk perception and self-efficacy related to environmental health behaviours. The Household Risk Perception instrument was developed to measure personal perceptions of household environmental health risks. The Self-Efficacy in Environmental Risk Reduction instrument was designed to measure caregivers' confidence in taking steps to reduce household risks. Design: An exploratory analysis of previous data was undertaken. Method: Baseline data from 235 caregivers enrolled in a randomized clinical trial testing a healthy housing intervention were collected between 2006-2009. Principal components analysis was used to determine principal components from measured responses to each instrument. Results: Components were explored and compared to constructs used to design the original instruments. A five-component structure showed the simplest solution and explained 65% of variance in the Household Risk Perception analysis. Cronbach's alpha values indicated satisfactory internal consistency for four of five identified components. Risk perception varied according to available sensory input of the specific risk. A four-component structure explained 64% of the variance in the Self-Efficacy in Environmental Risk Reduction analysis. Cronbach's alpha values were satisfactory. Items mapped to steps in an action-oriented process vs. agent-specific actions. Results from both analyses suggest that environmental tobacco smoke is perceived differently than other household risks. Conclusion: Previously, both instruments relied on item reliability and content validity testing. This study provides a basis for further instrument revision and theoretical testing.