Ratio Scaling of Tat Need for Achievement through Application of a Paired-Comparisons Method, with a Test of Internal Consistency
A method for ratio scaling of projective personality variables was devised and tested in two studies, using the TAT and achievement motivation (n Ach). In Exp. I, 4 judges made direct ratio estimates of relative n Ach in all possible pairs of 8 Ss' TAT protocols (28 pairs). In Exp. II, 3 judges used the same procedure to estimate relative n Ach for all pairs of TAT stories told to the same card, rather than for four-story protocols (112 pairs). Ratio scale values correlated with traditional n Ach scoring between .86 and .91 in Exp. I, and .90 and .95 in II. The protocols represented two criterion groups with different n Ach. Ratio scores from Exp. II, but not from I, differentiated these groups as well as traditional scoring. A model was developed for testing the internal consistency of judges' ratio estimates, based on Comrey's work. It showed over 90% of the ratio estimates' variance accounted for by the linearity assumptions of the ratio scaling model. Advantages and problems in using the ratio scaling method were discussed.