This study assessed repetition as a potential source of a positivity bias in memory of older adults. In line with previous studies that aimed to clarify the nature of positivity effects in memory of older adults [32,33,34], our study suggests that repetition and liking may act together to generate subsequent better memory for items that participants liked most. In particular, we hypothesized better memory for information that received a more positive connotation during encoding especially in older adults. Differently, we did not expect any such priority for liked information in memories of younger adults.
Repetition effects in memory
We found a main effect of type of repetition in both Experiment 1 and 2. Furthermore, the classical spacing effect was detected in both experiments. When stimuli were repeated, memory performance increased in both younger and older adults. In addition, memory was better when stimuli were repeated spaced apart compared with massed presentations. The major explanation advanced to explain these results relies on repetition priming mechanisms that operate during encoding for both younger and older adults. Repetition priming reduces the processing of the second occurrence of massed items, favoring spaced items that, in this way, receive a total greater amount of processing [5, 6].
When looking at recognition performance in both experiments and across the two groups, we found that recognition memory, in general, increased in Experiment 2 compared to recognition data from Experiment 1 indicating that nonwords were more difficult to remember than faces. This pattern of results mirrors previous data on spacing effects in cued-memory tasks [5, 6] which showed lower performance for unfamiliar verbal items compared to pictorial material. This may be due to the fact that unfamiliar faces may somehow remind participants of semantic details from familiar faces and thus memory may benefit.
Altogether, our findings indicated that spaced repetitions may be used at encoding to sustain memory performance in both younger and older adults. This study also adds to previous studies on spacing effects in aging [35,36,37] by showing that there are no age-related differences in terms of better memory for spaced items in a single study phase manipulation.
We also found that participants were more liberal on spaced items compared with massed and single items. This finding is in line with previous studies that showed that healthy adults adopt a more conservative response bias when memory judgments become more difficult (as it happens here with single and massed items) [38, 39].
The mere exposure effects in memory
In both experiments, we replicated mere exposure effects typically found in memory studies [1]. That is, liking for a stimulus increased following repeated exposure to the stimulus.
According to processing fluency theories, this happens because the second time we experience a stimulus, it is processed more easily than novel stimuli. Consequently, repetition increases the ease and speed of processing of the presented stimuli. Such fluency makes stimuli liked better because fluency produces a sense of familiarity and what is familiar is perceived as positive [9]. This is one of the first studies, as far as we know, to show that it is possible to obtain mere exposure effects in a single study phase where items are repeated either consecutively or spaced. In particular, mere exposure effects were obtained only for spaced items, that is, items that were repeated with a number of intervening items in between. This is in line with the mere exposure effect data that evidenced the importance of an interval for the effect to arise. Our data support this hypothesis. In fact, we found that when participants judged target items that were repeated consecutively, the mere exposure effect did not occur. One explanation [16] posits that when items are presented in a massed fashion, the increased fluency associated with consecutive studied items is not surprising and thus it is, generally, not interpreted in terms of increased liking: individuals tend to rate the item in the same way they just did. Conversely, participants do not normally expect to encounter the second occurrence of a spaced item, so when they do, they tend to increase their liking and semantic processing of it.
Another aspect worth noting regards age-related differences. As stated in the Introduction, studies on the mere exposure effect in aging are few [10, 11]. Our study thus adds to literature by showing that it is possible to obtain mere exposure effects in older adults. These data are in line with previous studies on preserved implicit memory mechanisms in aging, e.g., repetition priming [40]. In fact, the mere exposure effect has been considered an example of repetition priming effects [15, 17] that are typically preserved in healthy aging.
A comparison across experiments showed an intriguing pattern of results as older adults tended to show an increase in liking especially with nonwords (Experiment 1), while there were no differences across groups when faces were presented at study (Experiment 2). A tentative hypothesis may be that older adults assigned a greater positive connotation to the second occurrence of a nonword compared to the second occurrence of a face. This may be due to the higher level of fluency and familiarity that a nonfamiliar nonword may convey on its second presentation with respect to faces presentations that may intrinsically be interpreted with a greater level of unfamiliarity than a nonword. However, one may also argue that the opposite should be expected. In fact, we have more experience with perceiving faces in general, so it seems more plausible that the second occurrence of a face would increase feelings of familiarity relative to a nonword. Additional experiments will need to clarify this issue.
Age-related effects of spacing and mere exposure effects on memory
In line with the hypothesis that repetition and positive affective connotation may interact in memory processes, we found a superiority effect in older adults’ memory for spaced items that were liked most. Differently, younger adults showed a general spacing effect independent of liking. In general, these effects seem to resemble the classical age by valence interaction typically found in positivity effects studies [32, 34, 41] if we assume that what is liked is generally perceived as more positive. In fact, older adults showed a processing priority for items they liked most, whereas younger adults’ memory was less sensitive to increases in liking for spaced items. This pattern of results suggests that older adults’ motivational goals influenced the quality of the affective information to be remembered [2, 22, 42, 43]. In particular, the mere exposure effects for spaced items strengthened older adults’ focus on positive information due to their motivational implications (e.g., positive affect). Our results seem to be in line with those from studies that showed the strength of the positivity effect in memory in older adults [34, 44, 45] by highlighting the role of socio-emotional self-relevant goals in influencing memory performance [2].
Taken together, our results are also noteworthy when the type of studied material is considered. In fact, it has been shown that positivity effects in aging occur more often with meaningful emotionally charged verbal and pictorial material (e.g. words and pictures) [46, 47]. In our case, we found a positivity bias with unfamiliar material (such as nonwords and unfamiliar faces) that was not valenced. Our data seem to point to the role of both perceptually and conceptually driven processes in the generation of positivity bias in aging. Given the incidental learning procedure adopted and the unfamiliar material used, we assumed that more perceptually and implicit mechanisms were at work during encoding (e.g., repetition priming effects) giving rise to older adults’ increases in liking. However, at test, older adults selected only items they liked most, underlining the role of their motivational goals in memory (conceptually driven processes).
Limitations
Among the limitations of our study, we did not directly test repetition priming effects. A future study should include a session that experimentally primes participants and later investigates memory performance. We are currently running a reaction time experiment in which target nonwords are repeated either in massed or in spaced fashion to explore differences in priming effects between younger and older adults as a function of liking responses.
Another aspect regards the single study phase manipulation of spacing and mere exposure effects. We do not know whether classical spacing and exposure effects paradigms (e.g., distributed in different sessions coupled with long-term retention [48]) may also orient participants’ memory towards the positive pole.
The robustness of effects with different memory tasks should also be investigated. In fact, spacing effects rely on different mechanism according to the type of memory task [49]. For instance, in free recall, one of the major explanations posits that spaced items benefit from more distinctive contextual retrieval cues than massed items due to their greater context variability. In this case, it would be interesting to study whether the mere exposure effect arises and how memory is influenced. Mather and Knight [32] conducted one of the first studies on free recall and repeated testing with a 2-day interval and still found a positivity effect in memory of older adults indicating that repetition may be an important mechanism in the generation of positivity effects.
Finally, it would be also interesting to adopt a remember/know paradigm to better highlight the role of familiarity and recollection in older adults’ greater focus on items they liked most, e.g., whether older adults consciously retrieve contextual details (in this case their liking judgement) that were associated with the item at the time of encoding or just know that the item was on the list even though they cannot recall no specific information about its prior occurrence. Our results seem to point to a recollection-based hypothesis of positivity effects as older adults’ recognition focused on the items, they liked most among a series of items they remembered overall.