Rating episodes phase
Following the free recall test, the participants were provided with a sheet containing all the target words. Each participant was asked to indicate whether each target word reminded them of a previous episode. If a particular target word reminded them of an episode, they were asked to rate the degree of three attributes of each episode. These in­clude pleasantness on a scale of 1 (most unpleasant) to 6(most pleasant), vividness on a scale of 1 (most dull) to 6 (most vivid), and nostalgia on a scale of 1 (never nostalgic) to 6 (most nostalgic) at their own pace. This phase lasted approximately 6 minutes.
RESULTS
The mean scores of the three attributes of pleasantness, vividness, and nostalgia are shown in Table 1 as functions of the type of target (recalled vs. non-recalled). An analysis of variance was conducted for each attribute score to investigate the difference between the recalled and non-recalled targets. The main effect of the type of target (recalled vs. non-recalled) was significant, F (1, 27) = 11.72, p < .01, η2 = .15 for pleasantness rating, F (1, 27) = 21.63, p < .001, η2 = .12 for vividness rating,and F (1, 27) = 10.86, p < .01, η2 = .13 for nostalgia ratings. The correlations between each pair (e.g., pleasantness vs. vividness) are shown in Table 2. Only a significant correlation between vividness and nostalgia rating scores was observed for episodes of recalled targets. Cronbach’s coefficient alpha among the three attributes was calculated and non-recalled targets to investigate internal consistency in each episode. The alphas are shown on the right-hand side of Table 2. The difference in alphas
between the two indicated that the internal consistency in an episode in the recalled target was better than that in non-recalled targets.
Table 1
Mean rating scores of pleasantness, vividness, and nostalgia as a function of the type of targets (recalled vs. non-recalled)