(1983) Markham, Steven E.; Dansereau, Fred, Jr.; Alutto, Joseph A.; Dumas, MacDonald
Problems in drawing inferences about leadership
phenomena when multiple units of analysis (groups
and individuals) simultaneously exist in a data set are
addressed. Using a technique recommended by Dansereau
and Dumas (1977), within-unit and between-unit
sources of covariation are examined for data containing
matched superior-subordinate reports. In this
data set matched superior-subordinate reports were not
significantly correlated at the individual level. When
supervisory group differences were held constant,
however, the relationships between these matched reports
were significantly greater than zero. This convergent
validity within supervisory units suggests an approach
to validity which is not included in traditional
theories of leadership.
