Establishing Scale Structure, Reliability and Construct Validity of a Forced-Choice Measure of Preschoolers' Illness Knowledge

Wednesday, 1 August 2012: 8:50 AM

Catherine Reisenberg, PhD
Department of Nursing Quality, Safety and Education, Washington Hospital Center, Washington, DC

Learning Objective 1: The learner will be able to identify four dimension of illness knowledge that preschoolers recognize.

Learning Objective 2: The learner will be able to evaluate the utility of a newly developed measureof prescholers' illness knowledge.

Purpose: To establish scale structure, reliability and construct validity of a newly developed instrument to evaluate preschoolers’ illness knowledge called the Illness Knowledge Questionnaire (IKQ). The IKQ was a 32-item questionnaire that included four brief vignettes describing a child who had an illness. The vignettes were followed by forced-choice questions that focused on four illness dimension (i.e., identify, cause, consequence, and cure). Simple line drawings provided visual cues of vignettes and choices.

Methods: Two hundred and thirty children from three age groups (i.e., 4-5, 8-9 and 11-12 years) participated in this cross-sectional descriptive study. School age children were included to assess the construct validity of the IKQ.

Correlation of items, item difficulty, and item discrimination index were used to evaluate the performance of the individual items. Alpha coefficient for IKQ scale and subscales were evaluated. The construct validity of the IKQ was evaluated using hypothesis testing. Exploratory factor analysis was performed to empirically determine the number of underlying constructs.

Results: Twenty items were retained because the items had significant correlations, discrimination index of + .3 or greater and difficulty levels between 30 and 95 percent. Twelve items were dropped from the IKQ for either poor correlations and/or poor difficulty/discrimination ability. Coefficient alpha for the IKQ was .88 and coefficient alphas for three dimension subscales were greater than .70.  The construct validity of the IKQ was supported by two of the three hypotheses. The factor structure of the IKQ included four factors; however, the factors were not completely consistent with the four illness dimensions.

Conclusion: The newly developed 20-item IKQ fills an important gap in the literature in that it measures illness knowledge (i.e., an individual’s understanding or familiarity with facts about illnesses) in preschoolers in a developmentally appropriate format that can be administered in a relatively short period of time.