Source Memory

Source monitoring is an attribution process of memory records to their original sources. Source monitoring tasks are extensively used to examine both item recognition and source discrimination. However, there are no standard agreed-upon methods for presenting and testing sources in an experiment yet. This could potentially lead to the confounds during source attributions. The main objective of the current study was to investigate the measurement of source monitoring processes by focusing on test formats and source presentation within the same experimental design. We hypothesized more false alarms for new items and fewer source misattributions for the simultaneous source monitoring test format relative to the sequential source monitoring test format. Moreover, we expected fewer source misattributions and better item recognition in the blocked source presentation condition compared to the mixed source presentation condition. Although our planned analyses were not in support of the hypothesized differences, when we analyzed the data in blocks, we observed more false alarms for new items in the simultaneous source monitoring test format compared to the sequential source monitoring test format in the second block. We further investigated data with multinomial source monitoring modeling and discussed our findings in relation to encoding strategies and criterion shifts.

In source monitoring, schematic expectations affect both memory and metamemory. In metamemory judgments, people predict better source memory for items that originated from an expected source (e.g., oven in the kitchen) than for items that originated from an unexpected source (e.g., hairdryer in the kitchen; expectancy effect; Schaper, Kuhlmann, & Bayen, 2019). By contrast, actual source memory is either unaffected by expectations or better for unexpected sources (inconsistency effect; Kuhlmann & Bayen, 2016). Thus, the metamemory expectancy effect is illusory. This research tests the hypotheses that such metamemory monitoring of source memory affects metamemory control (i.e., measures taken to achieve a desired level of memory; Nelson & Narens, 1990) and memory. Due to their expectancy illusion, people should choose to re-study unexpected source–item pairs more often. Three participant groups (n = 36 each) studied expected and unexpected source–item pairs. One group rendered metamemory judgments and chose pairs for re-study. A second group made re-study choices only. These two groups then re-studied the chosen pairs. A third group did not make re-study choices and re-studied a random half of the pairs. All participants completed a source-monitoring test. As predicted, participants chose unexpected pairs more often for re-study based on their illusory conviction that they would remember unexpected sources more poorly. These re-study choices resulted in an inconsistency effect on source memory not shown in the group without re-study choices. Thus, the metamemory illusion affected control and memory in source monitoring.

Emotion-enhanced memory (EEM) describes the robust memory advantage of emotional over non-emotional stimuli. While extensively investigated with emotional items (e.g., pictures or words), EEM has been largely ignored with regard to emotional source information. Filling this gap, we tested if there is a source memory advantage for emotional over neutral source information by systematically manipulating source valance (positive vs. negative) between participants and source arousal (high vs. low) within participants. Specifically, we presented neutral scenery and object pictures as items together with high-arousing or low-arousing sounds of either positive or negative valence (dependent on the experimental group) as sources. We used a neutral low-arousing sound as baseline in both experimental groups (Group Negative: n = 40; Group Positive: n = 40). Multinomial model-based analysis indicated that—contrary to the EEM typically observed in item memory—source memory was substantially reduced for the negative high-arousing source compared to all other sources. There were no other effects of source emotionality on source memory. That is, positive valence did not influence source memory, not even under high arousal. Source emotionality further did not influence memory for the (neutral) items. We propose reduced binding of negative high-arousing sources as a potential explanatory account for the found source memory deficit and discuss possibilities to test this account in future studies.

Facial identification is important because it informs us regarding how to react to an approaching person, who might be friend or foe. In social situations, the emotional expressions of faces are important and very salient aspects of nonverbal communication. Previous work has shown that the facial expression (happy or angry) influences the memory of this face in a later recognition test. It is possible that positive expressions facilitate facial processing, therefore, more cognitive re-sources are available to process facial identity. Research on mood dependent memory shows that the likelihood of recalling something is higher when encoding and retrieval moods match than when they mismatch. We wanted to examine whether memory for positive and negative faces is influenced by the emotional state of the participant. Results indicate that old-new discrimination but not source memory is affected by whether a face was presented with a positive or a negative expression, independently of the emotional state the person was in.

In modern digital environments, people are exposed to information from a wide variety of sources, some of which are trustworthy while others are not. Having to deal with a large proportion of information from untrustworthy sources poses challenges to human information processing. Specifically, discriminating between information from trustworthy and untrustworthy sources places high demands on source monitoring processes. Advertising can be viewed as an untrustworthy source because advertisers have an economic self-interest in presenting the advertised products in a biased way. In two experiments, the proportion of messages from trustworthy and untrustworthy sources was manipulated so that participants encountered a high or low proportion of information from untrustworthy sources. In a third experiment, one group of participants saw only information from trustworthy sources while another group encountered information from a trustworthy as well as from an untrustworthy source. The exposure to a large amount of information from an untrustworthy source stimulated increased source monitoring at the cost of decreased processing of the content of the messages. When a source was not remembered, participants showed a guessing bias towards attributing the message to the untrustworthy source. These findings suggest that having to deal with a large amount of information from untrustworthy sources changes how information is encoded and remembered in a potentially costly way.

In social interactions, it is central to remember to which interaction partner one has said or given something. This process is referred to as destination memory and distortions of the latter can provoke (potentially embarrassing) redundancy in communication (e.g., tell a person the same joke twice). Albeit closely related to source memory, there has been only little research on destination memory in the past. To foster the understanding of this process, we examined to what extent it is affected by familiarity of the interaction partner and by reciprocity of one’s act. In two computerized experiments, individuals assigned everyday objects (e.g., bike, computer) to either of two persons (destinations). In Experiment 1, we manipulated familiarity of destinations between subjects such that individuals gave objects to either of two close relatives/friends or unknown persons. In Experiment 2, we manipulated the reciprocal nature of the interchange within subjects such that individuals lent or gave away objects to either of two close relatives/friends. In a subsequent (destination-) memory test, individuals decided whether, and if so to which destination, they gave the object (and additionally whether they had lent or gave away the object; only in Experiment 2). Individuals remembered better to whom they had assigned objects when they interacted with familiar than with unfamiliar persons. Individuals, however, remembered equally well to whom they had lent/given away the objects. We discuss substantive and methodological commonalities and differences when assessing the cognitive processes of both source and destination memory.