17
strategy in the vignette was not verbally aggressive. The hypothesis was supported. The
correlation between high verbal aggression and negative stereotype activation when
presented with a not verbally aggressive message is .28 (p = .006). The correlation
(Table 1) for each negative superordinate category (without a Bonferroni-Sidak
adjustment) is .24 (p = .02) for Severely Impaired, .27 (p = .007) for Shrew/ Curmudgeon,
.21 (p = .03) for Despondent, and .186 (p = .05) for Vulnerable.
The second hypothesis asserts that positive quality interactions with older adults
will be positively correlated with positive stereotype activation and negatively correlated
with negative stereotype activation, regardless of relational level or message strategy.
The hypothesis was not supported. There was a negative, but insignificant, correlation
(Table 2) between positive interactions and positive stereotypes. The negative
correlations between positive interactions and each positive superordinate category
except Perfect Grandparent. There were negative correlations between positive
interactions and negative stereotypes none of them at a significant level.
Effect of Message Strategy
Hypothesis 3 considers the association between a verbally aggressive message
and the activation of negative stereotypes. Analysis was limited to those respondents
who had the vignette in which a verbally aggressive message appeared. This hypothesis
was supported. Respondents receiving a verbally aggressive message had more negative
stereotypes activated (t = -13.93, p < .001), and fewer positive stereotypes (t = 9.41, p <
.001) (Table 3). The hypothesis was also supported for the seven superordinate
categories; for Perfect Grandparent (t = 13.18, p < .001), John Wayne Conservative (t =
3.34, p = .001), Golden-Ager (t = 7.81, p < .001), Severely Impaired (t = -10.9, p < .001),
Convention