This study examined the construct validity of the Dutch Life Orientation Test-Revised (LOT-R) in two samples; one sample of patients with psychiatric disorders (n=157) and one sample of patients with rheumatoid arthritis (RA, n=83). Confirmatory factor analyses showed that the LOT-R was not sufficiently unidimensional and could be better explained by two underlying factors consisting of positively and negatively worded items, respectively. This two-factor solution fitted the data significantly better than the onefactor solution in both groups, but satisfied all criteria for good model fit in the psychiatric patient sample only. One-factor models allowing correlated error terms between the positively or negatively worded items performed equally better than the original one-factor solution in both groups, indicating that the two factors may be the result of the specific wording of the items. However, the two factors were differentially associated with other relevant psychological constructs and correlation patterns differed substantially between both populations, indicating possible conceptual differences between optimism and pessimism. Overall, the findings suggest that the positively and negatively worded items of the Dutch LOT-R do not reflect a true unidimensional construct, but two underlying factors which may reflect a complex combination of methodological artefact and substantive differences. Therefore, researchers using the Dutch LOT-R are encouraged to not only rely on total scale scores, but to use additional sub-scores for optimism and pessimism to better examine possible relationships and effects in optimism research.
|Number of pages||12|
|Journal||Psychologie & gezondheid|
|Publication status||Published - 2010|