Results
Two separate 2 x 3 mixed factor ANOVAs were conducted examining the effects
of age (younger vs. older adults) and word type (negative/neutral/positive words) on both
recognition hit rate and d prime, which measures the sensitivity of correctly
discriminating “old” from “new” words presented in the test list. The age variable was a
between-subjects variable, whereas word type was manipulated within-subjects. An alpha
level of 0.05 was used for all F-tests.
Hit Rate
Hit rate was computed for each word type condition by dividing the number of
hits by the number of “old” words presented for each condition. Mean hit rates are
presented in Table 1. The 2 (Age) x 3 (Word Type) ANOVA resulted in a significant
main effect of age group, F(1, 46) = 5.72, p = .013, 2 = .08, with younger adults
exhibiting higher hit rates than older adults. The main effect of word type was also
significant, F(2, 92) = 4.61, p = .021, 2 = .07. Specifically, bonferroni-adjusted
comparisons indicated that when compared to neutral words, both positive, t(47) = 2.98,
p = .009, and negative words, t(47) = 2.64, p = .011, had higher hit rates. Hit rates for
positive and negative words did not differ, t(47) = 1.26, p = .241.
The above main effects should be interpreted with caution given the significant
Age x Word Type interaction, F(2, 92) = 8.49, p = .002, 2 = .11. Simple main effects
showed that the word type effect was significant for both younger, F(2, 92) = 5.77, p =
.021, 2 = .08, and older adults, F(2, 92) = 4.39, p = .026, 2 = .07, though the pattern of
word type effects was different for the two age groups. For younger adults, only positive
words were more memorable than both neutral, t(47) = 3.56, p = .004, and negative
words, t(47) = 3.01, p = .010, whereas hit rates for neutral and negative words were not
statistically different, t(47) = 0.72, p = .492. With older adults, compared to the neutral
condition, both positive, t(47) = 5.48, p = .002, and negative words, t(47) = 9.71, p <
.001, had higher hit rates, though the comparison of positive and negative words was not
statistically significant, t(47) = 0.69, p = .589.
d prime
As with hit rate analyses, a 2 (Age) x 3 (Word Type) mixed factor ANOVA was
conducted on d prime values computed from hit and false alarm rates (see Table 1 for
means). The main effects of Age, F(1, 46) = 4.99, p = .022, 2 = .08, and Word Type,
F(1, 46) = 5.97, p = .011, 2 = .09, were identical in pattern to those found in the hit rate
analyses, and so are not further discussed. An Age x Word Type interaction was again
observed, F(2, 92) = 7.34, p = .007, 2 = .10, though the pattern of word type effects was
slightly different than that found for hit rates. Simple main effects of word type were
again significant for both younger, F(1, 46) = 5.21, p = .019, 2 = .07, and older adults,
F(1, 46) = 5.75, p = .012, 2 = .08. Bonferroni-adjusted simple comparisons for younger
adults showed …..yada, yada, yada
regression: R2 for model, splitting data sets, stbs with ts