The purpose of this study was to construct a Big-Five personality inventory. As a pretest, 95 items were administered to 236 students, along with Goldberg's bipolar Big-Five marker inventory and the MINI Personality Inventory. The markers proved to be a good criterion for the five factor model (FFM), and 69 items were selected. In the main study, 496 students responded to 300 items, including the 69 pretest and the 43 MINI items that correlated highly with the markers, together with the Big-Five marker and the MINI inventories. Using MINI scores as criteria, those with poor insight were dropped, keeping 443 for further analyses. First, with the Big-Five markers as criteria, 150 items were chosen. Then, with group principal component analysis, 60 items were selected, and principal factor analysis with orthomax and factor parsimony criteria was applied, yielding a simple five-factor structure. Finally, Correct Attitudes items of the MINI Inventory were added to make 70 items, which were reordered for a final version. The correlations between the new scales and the Big-Five markers were between .510 and .774, and one-week test-retest reliability with 227 students ranged between .853 and .953, showing high reliability and validity.
View full abstract