The dangers of extreme positive responses in Likert scales administered to young children

Abstract

Likert scales are commonly used in questionnaires to examine participant attitudes towards a range of factors. If the reliability of the scales being examined, however, is in question then the results reported also become questionable. The current study examines the reliability of scales with young children in a longitudinal study. It explores the variability in response patterns of older (11-12 years) versus those of younger students (7-8 years) for baseline data and shows that reliability increases with the age of the students. It further shows that younger students are more likely than older students to respond positively to, and to miss items. The consequences for studies reliant on baseline or cross-sectional data with younger students is discussed, and implications for design are examined.