The Psychological Assessment of Political Leadersity of the leader's scores, trying to ascertain how sensitive the leader
is to the political context. Both kinds of reliability have been deter-
mined for the coding system described here.
Across a number of studies (e.g., Hermann 198oa, 198ob, 1984a,
1987^ Hermann and Hermann 1989), the intercoder agreement for
the seven traits described in this chapter has ranged from .78 to i.oo
between a set of coders and the author. Where there were disagree-
ments, the discussions that followed between coders permitted
refinements of the coding system. Generally, a coder currently is not
permitted to content analyze a leader's interview responses to be
included in the larger data set until he or she achieves intercoder reli-
abilities with the author on all traits that are .90 or higher. As the
automated coding system is being developed, similar types of relia-
bility coefficients are being calculated to determine how accurately
that coding system is in reflecting the original intent of the author.
Information about the leader can be gained by assessing trait reli-
abilities for that particular person. By correlating a leader's odd- and
even-numbered interview responses, the analyst can ascertain how
stable the traits are across time and issues. This index provides
another way of determining how open and closed the leader is likely
to be to contextual information (see, e.g., Hermann 198oa, 1984a).
One of the questions often raised about content analysis coding
systems that use translated material, which we are often forced to do,
regards the effect of the translation on the resulting scores. To ascer-
tain whether there were any effects and the nature of such effects, in
several instances intercoder agreement has been calculated between a
native speaker coding text in the original language and the author
focusing on the translated text. The languages were Russian and
French. In both cases, agreement averaged .92 across the seven traits
(Hermann 198ob, 198ya, 1987^.
Validity of This Profiling Technique
How valid is this particular way of determining leadership style?
That is, how accurate is it in capturing the leadership styles that
heads of state and others in leading party and bureaucratic positions
actually exhibit? Although I have received numerous suggestions
about how to determine the validity of this technique, ranging from
running experiments with college students to participant observa-