A 2×5 Mixed Factorial ANOVA showed a significant main effect of Trial [F (4, 192) = 137.93, p < 0.01], Group [F (1, 48) = 13.91, p < 0.01], and an interaction between Trial and Group [F (4, 192) = 5.66, p < 0.01] indicating that both groups’ recall increased across Trials 1-5, but the older group had poorer recall accuracy from Trials 2-5 (Figure 3.1d). A series of independent t-tests were carried out to determine where the differences lay between the two groups across trials 1-5. Independent t-tests revealed no significant difference between the two groups for Trial A1, [t (50) = 1.188, p > 0.05]; however, they indicated significantly better accuracy for the young group on each of the other trials: A2 [t (50) = 3.176, p < 0.01]; A3 [t (49) = 4.474, p < 0.01]; A4 [t (49) = 3.728, p < 0.01]; A5 [t (50) = 3.971, p < 0.01] indicating that both groups began with similar recall scores but as the trials progressed, the younger group were better equipped to learning and recalling the words.
Significant effects were found for retroactive interference [F (1, 49) = 58.19, p <
0.01], group [F (1, 49) = 34.5, p < 0.01] and the interaction between retroactive interference and group [F (1, 49) = 12.32, p < 0.01]. Bonferroni corrected dependent t-tests showed that both groups’ accuracy decreased from Trial 5 to Trial 6 [Young: t (25)
= 3.302, p < 0.01; Older: t (24) = 7.076, p < 0.01] and independent t-tests revealed that
the young group had higher accuracies on Trials A5 [t (50) = 3.971, p < 0.01] and A6 [t (49) = 5.876, p < 0.01], therefore the older group showed more interference, i.e. a greater decrease from Trials 5 to 6.
No significant main effect was found for proactive interference, i.e. between Trials A1 and B [F (1, 50) = 0.330, p > 0.05]; a group effect was found [F (1, 50) = 4.913, p < 0.05] and this difference was seen at Trial B, where the young had a higher recall [t (50) = 2.627, p < 0.05]; and no interaction was found between proactive interference and group [F (1, 50) = 1.305, p > 0.05].
A statistically significant effect was found for the delay trials [F (1, 49) = 8.405, p < 0.01] and group [F (1, 49) = 31.974, p < 0.01], while no significant effect was found for the interaction between delay trials and group [F (1, 49) = 0.244, p > 0.05].
Bonferroni corrected dependent t-tests showed that the younger group had a statistically significantly better performance on Trial 6 than Trial 7 [t (25) = 2.953, p < 0.05] while the older group did not show the same [t (24) = 1.455, p > 0.05]. Independent t-tests revealed that the young group had higher recall scores on Trials A6 [t (49) = 5.876, p <
0.01] and A7 [t (50) = 4.516, p < 0.01].
Table 3.1: Means and standard deviations for NART based IQ estimates, Digit Span,
Figure 3.1: (a) Predicted Full Scale, Verbal Scale and Performance Scale IQs based on NART performance for young and older adults +/- SEM, (b) Forward, backward and total digit span scores for young and older adults +/- SEM, (c) CFQ scores for young and older adults +/- SEM, (d) RAVLT accuracy scores for Trials A1-A7 for young and older adults +/- SEM
(* p < 0.05, ** p < 0.01)
3.3.2 Opposition Task 3.3.2.1 Accuracy
A 2×3×2 Mixed Factorial ANOVA was conducted to determine whether there was a difference between the two groups on accuracy at the lag lengths for the different word types. No main effect was found for Word Type. However, a significant main effect was found for Lag [F (2, 94) = 5.555, p < 0.01] where accuracy decreased across the lag lengths for Distractor words and increased for Target words. The main effect for Group was also significant [F (1, 47) = 4.971, p < 0.05] with the young group showing better accuracy. Significant interactions were observed between Word Type and Group [F (1, 47) = 7.02, p < 0.05], Word Type and Lag [F (2, 94) = 70.67, p < 0.01], and Word Type, Lag and Group [F (2, 94) = 4.101, p < 0.01]. No significant interaction was found between Lag and Group.
Subsequent independent t-tests were carried out to determine if the groups differed on the individual lag lengths. The groups differed significantly at lag 0 [young:
mean = 9.68, SD = 0.56; older: mean = 9.04, SD = 1.16; t (47) = 2.439, p < 0.05], lag 4 [young: mean = 8.48, SD = 1.7; older: mean = 6.36, SD = 2.23; t (50) = 3.875, p < 0.01]
and lag 16 [young: mean = 7.29, SD = 2.05; older: mean = 5.84, SD = 2.5; t (50) = 2.305, p < 0.05] for distractor words, with the young group showing greater accuracies in both cases (Figure 3.2). No other between group differences were significant [all t <
1.874, all p > 0.05].
Furthermore, Bonferroni-corrected dependent t-tests were carried out to determine where the differences lay for the word types and lags for each group. For the young group a significant difference was seen between Distractors at lag 0 [mean = 9.68, SD = 0.56] and lag 4 [mean = 8.48, SD = 1.7; t (24) = 3.464, p < 0.01], lag 0 [mean = 9.68, SD = 0.56] and lag 16 [mean = 7.30, SD = 2.05; t (24) = 3.6, p < 0.01], and lag 4
[mean = 8.48, SD = 1.7] and lag 16 [mean = 7.30, SD = 2.05; t (26) = 3.6, p < 0.01]
with lag 0 having the highest accuracy followed by lag 4 (see Figure 3.2). In addition, the young group had a higher accuracy for targets at lag 16 [mean = 7.96, SD = 1.5]
compared to lag 0 [mean = 6.96, SD = 1.81; t (26) = -2.762, p < 0.05]. For the older group, significant differences were found between distractors at lag 0 [mean = 9.04, SD
= 1.16] and lag 4 [mean = 6.36, SD = 2.23; t (23) = 7.092, p < 0.01], and at lag 0 [mean
= 9.04, SD = 1.16] and lag 16 [mean = 5.84, SD = 2.5; t (23) = 6.608, p < 0.01] with a higher accuracy at lag 0 for both comparisons. Significant differences were also seen for the older group for targets at lag 0 and lag 4 [lag 0: mean = 6.44, SD = 1.85; lag 4:
mean = 7.8, SD = 1.8; t (23) = -3.369, p < 0.01], and at lag 0 [mean = 6.44, SD = 1.85]
and lag 16 [mean = 8.64, SD = 1.08; t (23) = -5.336, p < 0.01] with lag 0 having a lower accuracy for both comparisons. All other comparisons within groups were not statistically significant [all t < 2.068, all p > 0.05].
Figure 3.2: Mean total accuracy +/- SEM for Distractors and Targets words at Lags 0, 4 and 16 for young and older adults on the Opposition task
(* p < 0.05, ** p < 0.01)
3.3.2.1 Response Times for Correct Responses
A 2×3×2 Mixed Factorial ANOVA found statistically significant main effects for Word Type [F (1, 43) = 29.982, p < 0.01], Lag [F (2, 86) = 87.467, p < 0.01] and Group [F (1, 43) = 23.946, p < 0.01] where response times increased across the lag lengths and were faster for the young group (see Figure 3.3). The interactions of Word Type and Group [F (1, 43) = 1.588, p > 0.05], Lag and Group [F (2, 86) = .418, p >
0.05], and Word Type, Lag and Group [F (2, 86) = 2.277, p > 0.05] were not statistically significant. However, a significant interaction was found between Word Type and Lag [F (2, 86) = 2.277, p < 0.01].
Independent t-tests showed that the young group responded significantly faster for correct responses on each of the Lags for Distractor words: Distractor Lag 0 [young:
mean = 520.42, SD = 55.29; older: mean = 826.82, SD = 199.38; t (49) = -7.415, p <
0.01], Distractor Lag 4 [young: mean = 1038.07, SD = 286.10; older: mean = 1420.45, SD = 458.73; t (50) = -3.574, p < 0.01] and Distractor Lag 16 [young: mean = 1091.78, SD = 392.89; older: mean = 1433.17, SD = 439.37; t (47) = -2.869, p < 0.01]. The older adults had significantly slower response times compared to the young group for the Target words at each lag: Targets Lag 0 [young = 537.40, SD = 95.75; older: mean = 897.51, SD = 275.02; t (49) = -6.196, p < 0.01], Targets Lag 4 [young: mean = 837.14, SD = 206.08; older: mean = 1066.11, SD = 275.02; t (48) = -3.636, p < 0.01] and Targets Lag 16 [young: mean = 840.97, SD = 250.36; older: mean = 1058.41, SD = 293.16; t (50) = -2.883, p < 0.01].
Figure 3.3: Mean response times in ms +/- SEM for both Distractors and Targets words at lags 0, 4 and 16 on the Opposition Task for young and older adults for correct
responses
(* p < 0.05, ** p < 0.01)
3.3.2 VPAc Task