8.3 METHOD
8.3.1 SAMPLE
The sample used for this study is a subset of the sample used for the previous study. For this study, as
the primary interest was subordinate¶ job satisfaction, only those managers were included that had
been rated by their subordinates. Managers that were rated only by peers and/or superiors, were
discarded from the sample. :KHUU\DQG%DUWOHWW¶V(1982) theory of rating indicates that there are three
bURDGW\SHVRIIDFWRUVWKDWLQIOXHQFHSHUIRUPDQFHUDWLQJV WKHUDWHH¶VDFWXDOMRESHUIRrmance, various
rater biases in the perception and recall of that performance, and measurement error. The second, rater
bias, refers to the systematic variance in performance ratings that is associated in some way with the
rater and not with the actual performance of the ratee. Important type of rater bias concerns the effects
DVVRFLDWHGZLWKWKHUDWHU¶VRUJDQL]DWLRQDOSHUVSHFWLYH LHVHOIGLUHFWUHSRUWSHHURUERVV Borman
(1997) advanced three reasons why it is plausible to hypothesize that perspective-related biases affect
performance ratings. First, raters from different organizational perspectives might focus their attention
on different aspHFWV RI WKH UDWHH¶V SHUIRUPDQFH 6HFRQG UDWHUV IURP GLIIHUHQWSHUVSHFWLYHV PLJKW
attend to the same aspects of performance but attach different weights to them. Third, raters from
different perspectives may REVHUYHGLIIHUHQWVDPSOHVRIDUDWHH¶VEHKDYLRU When interested in the
UHODWLRQVKLSEHWZHHQVXERUGLQDWH¶MREVDWLVIDFWLRQDQGOHDGHUVKLSbehavior of the superior, it is most
appropriate to base both these scores on the ratings of the subordinates themselves.
The sample for this study included 182 supervisor-subordinate pairs. This number surpasses the total
number of managers in the sample, due to the fact that some managers have been rated by more than
one subordinate and hence count for more than one supervisor-subordinate pair. See Figure 8.2 for a
graphic display the nested sample.