engleski

Preliminary Validation of Repeatedly Measured Alternate Forms of Attitude Toward Science Scale

Two alternate forms of 29 iteams each have been derived from pool of 78 items items of attitudes toward science (Prot, F., Prot, S. and Anderson, C., 2017). Eliminating one pair of equivalently formulated items 28 items in each of the forms remained. In new research, on five occasions, repeated measurement of attitudes toward science with two alternate forms, i.e. A and B were applied on the sample of students of kinesiology at University of Zagreb. For the moment 94 students completed all of five measurements. Means and standard deviations showed that pairwise difference between forms are existing, and that differences are present in between measurements points in forms internally. Preliminary validity of measurements is presented in two steps. For the first step composite score for each of forms and measurement points was derived as unweighted sum of items and their correlations were analyzed. Pairwise validity for each of five measurement of the forms A and B is defined as correlation matrix of order ten. Submatrices of correlations inside repeated measurement of each of forms are symmetric with positive and high values suggesting internal validity and offering possibility for simplex structure testing. Submatrix of correlations between repeated measurements of one of the forms against the other is asymmetric with high positive values having their pairwise validities in main diagonal (r1ab = 0.813 ; r2 ab = 0.866 ; r3ab= 0.897 ; r4ab = 0.901 ; r5ab= 0.887) and are highest values in respective rows and columns. There is one distinctively big eigenvalue 7.507of correlation matrix explaining 75.08% of overall variability by the first principal component. In the second step, synthetic measure of validity between sets of repeated measurement is demonstrated by correlation between the first principal components r12k = 0.954 ; the first pair of canonical correlation components r12c = 0.963 ; and the first pair of quasicanonical correlation components r12q = 0.955.

Validity ; Validation ; Alternate forms ; Attitude toward science scale

SIMPOZIJ 3 / SYMPOSIUM 3 SCIENCE DENIAL, ATTITUDES TOWARDS SCIENTISTS AND PSYCHOLOGY OF FREEDOM Convener: Sara Prot Coventry University, Coventry, United Kingdom Contact: ac2215@coventry.ac.uk Science denial is a common phenomenon occurring in various areas such as climate change, vaccine safety, tobacco and violent media effects. It is often accompanied by anger and personal attacks on individual researchers (Lewandowski et al., 2013) as well as development of global negative attitudes towards science and scientists (Prot et al. 2017). This symposium brings together researchers examining science denial, attitudes towards scientists and psychology of freedom. The first talk reports four correlational and experimental studies examining causal factors leading to science denial and attacks on scientists. Next, three talks examine global negative attitudes towards science and scientists. The second talk discusses metric properties of a global measure of attitudes towards scientists (the Attitudes towards Science scale, ATS), demonstrating reliability and validity across two parallel forms of the scale. The third talk examines the factor structure of the Attitudes towards Science scale