Page 130 - Historical Dictionary of Political Communication in the United States
P. 130
Q
Q SORT METHOD was devised by William Stephenson. The procedure begins
by having individuals respond to items. Usually, respondents are asked to rate
items from 1 to 10, but on a forced normal distribution. So, for example, if you
had 60 items, you might give respondents a forced distribution with 1 1, 2 2s,
3 3s, 8 4s, 15 5s, 15 6s, 8 7s, 4 8s, 2 9s, and 1 10. You would correlate
respondents with each other rather than correlating questions or items with each
other. You then factor analyze, and your factors are groups of people whose
responses are similar. You then compare the groups' responses, usually using z
scores, to find out which responses set each group apart from the other groups
and thus define each group. Stephenson maintained that you could do this with
small, purposive samples because the purpose is to determine what the factors
are, not what percentage of the population is in each one. Meaningful results
have been produced by Q studies with samples of no more than 25. (See also
William Stephenson.)
SOURCE: William Stephenson, The Study of Behavior, 1953.
Guido H Stempel III
