Assessing Leadership Style: Trait Analysis

The Psychological Assessment of Political Leaders

any other source, based on any other kind of material.^3 The decision
of which comparison group to use is critical in two respects. First, it
establishes the normative population against which the particular
person being studied will be compared. Second, the standardization
group decision also affects the selection of actual documents to be
scored, since the scores of all persons in the standardization group
should be based on the same kinds of documents. The process of
identifying comparison groups is subsequently illustrated with some
examples.

Identifying Comparison Groups

Often the appropriate comparison group will be obvious. Thus the

motive imagery scores of any particular president's first inaugural

address can be compared to the scores of all other first inaugural

addresses, as shown in table 7.3. Or the scores of one presidential

candidate's announcement speech can be compared to the speeches of

all other candidates for that year (as in Winter 1982). Hermann

(1979, 198oa, 1988b) has used her accumulated sample of world

leaders as a comparison group for present and future scorings of addi-

tional individual leaders.

Sometimes comparisons can be drawn among different subgroups

of a single larger group: for example, in his study of twenty-two

southern Africa leaders, Winter (1980) compared the motive

imagery scores of white leaders, "front line" heads of state, Zimbab-

wean nationalist leaders, South African nationalist leaders, and South

African "homeland" leaders.

Sometimes a single leader can be studied over time, using similar

kinds of verbal material. For example, Winter (1998b) studied

changes in the motive imagery levels of Bill Clinton's State of the

Union speeches from 1993 through 1996, in effect using Clinton as

his own comparison group.

The Case of Ross Perot as an Illustration of Problems

in Identifying Comparison Groups

Sometimes the appropriate comparison group is not obvious or is

difficult to construct. Such constraints may then affect the kinds of

questions that can be asked and limit the hypotheses that can be for-

mulated. The case of Ross Perot's 1992 candidacy for president can