The use of social influence techniques was examined in two studies of
marital conflict. In each study couples discussed two issues in counte
rbalanced order: one in which the husband wanted change and one in whi
ch the wife wanted change. Interactions were coded for the use of infl
uence techniques. Evidence from both studies supported a social role e
xplanation, such that roles based on who wanted change predicted the u
se of influence, whereas gender did not. Secondary hypotheses based on
attribution theory and schema theory were also supported. When discus
sing areas in which they wanted change in partners, spouses made more
partner attributions, were more likely to suggest change, and were mor
e likely to explain their partners' behavior than when discussing area
s in which their partners wanted change. However, they were more likel
y to ask questions to clarify the problem when discussing a change the
ir partners wanted in them than when discussing a change they wanted.
The importance of considering social structural variables such as soci
al roles in examining conflict interactions is discussed.