Cognitive load on working memory both encourages and discourages reasoning bias regarding the mental states of others

Theory of mind (ToM) refers to the ability to reason about the mental states of others. An increasing number of studies have revealed that working memory (WM) plays an important role in ToM. The present study applied WM loads to adults during a ToM task in order to investigate the impact on mental‐state reasoning performance. The task required participants to estimate the probabilities of several possible behaviours for a protagonist following the presentation of a ToM story. Participants were also required to maintain a meaningless two‐ (light WM load) or seven‐letter English alphabet string (heavy WM load) during story comprehension and mental‐state reasoning. The results show that the combination of light WM load applied during story comprehension with heavy WM load during mental‐state reasoning results in an overestimation of the probability that the protagonist's behaviour will accord with a participant's knowledge. Conversely, a heavy WM load applied during story comprehension, regardless of the type of WM load during mental‐state reasoning, did not result in this probability estimation bias. We discuss these findings from the perspective of a WM representation account.     

Breaking a habit: A further role of the phonological loop in action control

Recent research has suggested that keeping track of a task goal in rapid task switching may depend on the phonological loop component of working memory. In this study, we investigated whether the phonological loop plays a similar role when a single switch extending over several trials is required after many trials on which one has performed a competing task. Participants were shown pairs of digits varying in numerical and physical size, and they were required to decide which digit was numerically or physically larger. An experimental cycle consisted of four blocks of 24 trials. In Experiment 1, participants in the task change groups performed the numerical-size judgment task during the first three blocks, and then changed to the physical-size judgment task in the fourth. Participants in the continuation groups performed only the physical-size judgment task throughout all four blocks. We found negative effects of articulatory suppression on the fourth block, but only in the task change groups. Experiment 2 was a replication, with the modification that both groups received identical instructions and practice. Experiment 3 was a further replication using numerical-size judgment as the target task. The results showed a pattern similar to that from Experiment 1, with negative effects of articulatory suppression found only in the task change group. The congruity of numerical and physical size had a reliable effect on performance in all three experiments, but unlike the task change, it did not reliably interact with articulatory suppression. The results suggest that in addition to its well-established role in rapid task switching, the phonological loop also contributes to active goal maintenance in longer-term action control.     

Two modes of confidence rating: An effect of episodic information of participant's own past responses in a repeated-question paradigm

Abstract: General knowledge questions with two answer alternatives were employed in experimental session 1; in session 2, the same questions were presented together with participants' own session-1 responses. In order to examine whether or not the episodic information of participant's own responses would suppress standard confidence-rating mode in session 2, rates of answer- and confidence-changes between sessions were analyzed. In session 2, participants were able to change the confidence value to another, if they thought the initial value inadequate. They then had a chance to change the answer to the other and rated their confidence in the new answer. The major results were as follows: (a) Between-session answer change rate was very low; (b) Between-session answer change rate was not a monotonic decreasing function of confidence; (c) However, the rate depended on confidence change from session 1 to before-answer-change rating. These results clearly contrasted with a previous study ( Saito, 1998 ) in which episodic information of participant's own session 1 answers and confidence values was not presented. It was argued that the episodic information triggered another mode of confidence rating or a decision inertia effect.     

I see into your mind too well: working memory adjusts the probability judgment of others' mental states

Although an increasing number of studies on adults have indicated that working memory (WM) contributes to the ability to understand the mental states of others (i.e., theory of mind), the detailed mechanism by which WM contributes to successful reasoning has not previously been revealed. This study shows that WM modulates the degree of attribution of one's own knowledge to others' mental states. Participants were asked to read a story twice (Experiment 2) or as carefully as possible (Experiment 3) and to estimate the probability percentages of possible choices for a naive protagonist's behavior. The participants were then asked to maintain either a two- or seven-letter alphabet string (i.e., a light or heavy WM load, respectively) during the probability estimation but not during the story comprehension. The results showed that compared to the participants with a light WM load, those with a heavy WM load estimated a significantly higher probability of the choice indicating that the protagonist would behave on the basis of a fact that the participants knew but the protagonist did not. This result indicates that WM moderates the extent to which adults attribute their own knowledge to others' mental states. The role of WM in theory of mind and in heuristic strategy for making probability judgments was then discussed.     

Disruption of visual feature binding in working memory

In a series of five experiments, we studied the effect of a visual suffix on the retention in short-term visual memory of both individual visual features and objects involving the binding of two features. Experiments 1A, 1B, and 2 involved suffixes consisting of features external to the to-be-remembered set and revealed a modest but equivalent disruption on individual and bound feature conditions. Experiments 3A and 3B involved suffixes comprising features that could potentially have formed part of the to-be-remembered set (but did not on that trial). Both experiments showed greater disruption of retention for objects comprising bound features than for their individual features. The results are interpreted as differentiating two components of suffix interference, one affecting memory for features and bindings equally, the other affecting memory for bindings. The general component is tentatively identified with the attentional cost of operating a filter to prevent the suffix from entering visual working memory, whereas the specific component is attributed to the particular fragility of bound representations when the filter fails.     

Does test delay eliminate collaborative inhibition?

Earlier research has demonstrated that collaborative groups recall more than individuals, but less than nominal groups (pooled performance of individuals), thus exhibiting collaborative inhibition. In two experiments, all participants were first asked to recall story material on their own. Some participants were then assigned to pairs and recalled the material collaboratively. On the other hand, the participants in the individual recall condition were asked to recall the material once again on their own. In Experiment 1, the collaborative pairs recalled less than the nominal pairs in accordance with previous studies. In Experiment 2, the timing of the initial individual recall was manipulated by inserting one week between the learning and the recall. The collaborative inhibition was eliminated in this situation. Sources of the collaborative inhibition in immediate recall and its disappearance in delayed recall are discussed in terms of the effect of cross cueing in collaborative remembering.     

The effect of number of kanji radical companions in character activation with a multi-radical-display task

The majority of Japanese complex kanji consist of two subword units (radicals), and each radical is differentiated according to its combinability. Radical-type frequency is defined as the number of radical companions (NC) functioning as the number of characters containing that radical. The current study is intended to address the question of whether the NC plays a role in kanji character activation under a multi-radical-display (MRD) instead of a tachistoscopic display. An MRD task was used in which the NC was manipulated and the frequency of use of the whole characters was matched on average across the experimental conditions. Participants were divided into two groups (good knowledge vs poor knowledge) according to the score of a kanji completion task. The participants' task was to select an appropriate combination of radicals to compose a legitimate character under the MRD condition. The poor-knowledge group yielded relatively more errors than did the good-knowledge group in the low-NC condition, whereas the good-knowledge group yielded more errors in the high-NC condition than in the low-NC condition. These results demonstrate the different influences of NC as metaknowledge (tacit knowledge of radical productivity) in both good and poor readers.     

Verbal representation in task order control: An examination with transition and task cues in random task switching

Recent task-switching studies in which a predictable task sequence has been used have indicated that verbal representation contributes to the control of task order information. The present study focused on the role of verbal representation in sequential task decisions, which are an important part of task order control, and examined the effects of articulatory suppression in a random-task-cuing paradigm with two different types of cues presented just before the presentation of a stimulus: a transition cue and a task cue. The former cue provided information only about switching or repeating the task, whereas the latter was associated directly with the identity of the task (i.e., indicating a parity or a magnitude task). In Experiment 1, in which transition cues guided task sequences, articulatory suppression impaired performance in both repetition and switch trials, thereby increasing the mixing costs. In Experiment 2, in which a task cue, rather than a transition cue, was presented to examine the influence of a cue-decoding process, articulatory suppression had no specific effect on task performance. Experiment 3, in which the transition cue and the task cue were randomly presented in the same block to equalize the memory load and task strategy for the two types of cues, confirmed that articulatory suppression significantly increased the mixing costs only in transition cue trials. The results from the three experiments indicated that the use of verbal representation is effective in sequential task decision—that is, in selecting a task set on the basis of transient task order information in both repetition and switch trials.     

Power imbalance,its legitimacy,and aggression

Sixty Japanese female students were asked to exchange shocks in electric roulette games with female opponents. The subjects were assigned to either the 80%, 50%, or 20% win conditions. Half of them were then led to judge that the power unbalance was legitimate by being informed that the assignment was based on a prior performance contest, with the good performer being assigned to the advantageous position. The other subjects were led to perceive the power imbalance as illegitimate by being informed that the assignment was randomly decided. The opponents always severely attacked them. The retaliation by the subjects was analyzed by a two-way ANOVA with Power Imbalance and Legitimacy. It was found that the subjects both in the 20% and 50% win conditions set more intense shocks to their opponents than those in the 80% win condition. This is not consistent with the fear of retaliation hypothesis which had predicted that the subjects would refrain from intensely aggressing against the opponent who had a greater aggressive capacity. It was also found that in the 80% win condition, the subjects set more intense shocks when the unbalance had been determined by their performance than when determined by luck, whereas in the 20% win condition, they set more intense shocks when the imbalance had been determined by luck than when determined by their performance. These suggest that retaliation depended upon perceived justifiability of aggression which was predicated on the legitimacy of the power imbalance. Finding in the 50% win condition that the shock settings were higher when the power imbalance had been determined by their performance than when determined by luck was interpreted in terms of their heightened competitiveness.