The accounts were arranged in the same order for each

judge. Matched pairs of subjects were placed consecutively with the order of experimental followed by control, or vice versa, determined by a coin-flip. The judges were instructed to read all the accounts through at one sitting and then to proceed with the individual scoring of each account one by one. With enough time allowed to finish all the categories in any account which was started, each judge gave a qualitative score to each subcategory or category on a 0-3 scale, as described in the training manual. (Twenty-two items for each account, or 440 in all, were used.) The attempt to score each qualitative level of each item quantitatively was dropped. Thé scoring procedure was designed so that each account was read five times and different categories scored each time. The easiest categories (e.g., loss of time and space, and positive mood) were scored first. The more difficult categories were then scored after the account became more familiar through repeated readings. This technique was meant to reduce the chance that data would be missed. Each judge estimated that between 15 and 20 hours were required for the scoring of all 20 accounts. The Kendall Rank Correlation Coefficient OO3^ was found by com-

0 0

Post a comment