Abstract
The present study tested the validity of the state-trait curiosity distinction, investigated the effect of manipulating state curiosity upon learning performance, and assessed the reliability and validity of state (C-State) and trait (C-Trait) curiosity scales. Three hundred senior secondary school students were assigned randomly to (i) Curiosity Stimulating Instructions (CSI); (ii) Neutral Instructions (NI); or (iii) Boredom Inducing Instructions (BII) groups respectively. These treatments preceded the first administration of a test-battery (comprised of C-State, a state epistemic curiosity measure; SECS, a state curiosity scale; A-State; and C-Trait). A learning task (prose passage) ensued, followed by a second administration of the test-battery. Finally a posttest of immediate retention permitted assessment of each group's learning performance. Alpha coefficients for both C-State and C-Trait scales ranged from .86 to .92. Stability (test-retest) correlations were higher for the C-Trait scale than for C-State. Both concurrent and discriminant validity correlations were significant. Only tentative support for the state-trait curiosity distinction resulted, since the factor analyses were dominated by reversed and nonreversed items and subscales. Only for females in grade 12 did the instructional sets influence C-State scores and subsequent learning performance as predicted.