Working memory
Sunday, October 02nd, 2011 [18:00 - 20:00]
PS_3.054 - Transfer effects from working memory training to executive control processes
Salminen, T. 1 , Strobach, T. 1 , Sorg, C. 2 , Müller, H. 1 & Schubert, T. 1
1 Department of Psychology. Ludwig-Maximilians-University. Munich, Germany.
2 Department of Psychiatry/Klinikum rechts der Isar. Technical University. Munich, Germany.
Recent studies have reported an increase in fluid intelligence following extensive and adaptive working memory (WM) training. However, it is still unclear which components of such training can generalize to other, untrained tasks. In the present study, we investigated transfer effects from a demanding WM task, which requires simultaneous performing of a visual and an auditory n-back task, to other, untrained tasks tapping different cognitive domains: WM updating, coordination of performing multiple tasks, and attentional processing which, taken together, constitute executive control processes. Compared to an untrained control group, it was found that training led to improvements in the trained task as well as in the transfer WM updating task, and to enhancements in attentional processing. Transfer to the coordination of performing multiple tasks was marginal. Thus, these results confirm previous findings that WM can be trained and, importantly, they show that these training effects can generalize to other tasks that are not part of the training regimen.
PS_3.055 - Age difference in affective bias : The effect of context valence in working memory
Katsuhara, M. 1 , Osaka, M. 2 & Osaka, N. 1
1 Kyoto University
2 Osaka University
In previous studies, younger adults demonstrated an affective bias for negative context, and an impairment of memory for neutral target words simultaneously-presented (Rabinowitz & Carlson, 2010). Though older adults seem to be even more susceptible to the influence of negative and highly arousing stimuli, it remains unclear whether this affective bias would also occur in negative and low arousal context (Mather & Knight, 2005). We explored the effect of negative, low arousal context using three different reading span tasks (RST): positive-context, negative-context, and control (neutral-context) conditions. In the positive-context condition, the sentences had positive content, whereas the negative-context condition sentences had negative content. We selected sentences that had low arousal levels to remove the effect of automatic processing and attention. Additionally, target words had a relatively neutral content. In each RST, participants were asked to read sentences and targeted words aloud, while memorizing the targeted words. The results suggest that there was no affective bias for negative or low arousal context in older adults. In younger adults, recall accuracy was worse when the sentences proceeding target words had negative valence as opposed to positive. However, in older adults, there was no performance difference between negative and positive conditions.
PS_3.056 - Serial coding of verbal information in working memory
Ginsburg, V. 1 , Van Dijck, J. 2 , Van Opstal, F. 2 , Majerus, S. 3 , Fias, W. 2 & Gevers, W. 1
1 Unité de recherche en Neurosciences Cognitives, Université Libre de Bruxelles, Brussels, Belgium
2 Department of Experimental Psychology, University of Ghent, Ghent, Belgium
3 Département de Sciences Cognitives, Université de Liège, Liège, Belgium
Recently, a lateralized position effect in working memory was observed (Van Dijck & Fias, 2011). Words (fruits/vegetables) were sequentially presented and were to be remembered for later recall. During the retention interval, participants performed a categorization task with lateralized responses embedded in a go-nogo procedure such that participants only responded to the words inside the WM sequence but not to the words outside the WM sequence. Words presented early in the sequence were faster responded to with the left hand whereas words later in the sequence were responded to faster with the right hand.
First, we replicated this lateralized position effect using the same go-nogo paradigm but with different categories of words on each sequence (e.g. with 10 different categories of words instead of one) and with a smaller working memory load. In the second experiment, participants responded to all words, both inside and outside the WM sequence. This manner, the WM sequence was no longer relevant for the categorization task. The results show that in this case the lateralized position effect disappeared. We conclude that the lateralized position effect is a robust observation that crucially depends on the relevance of the working memory sequence for the lateralized response categorization.
PS_3.057 - Slowing down after an error is related to working memory updating and rule rehearsal
Dorchin, S. & Meiran, N.
Ben-Gurion University of the Negev
Responses usually slow after an error, indicating Post-Error Slowing (PES). We propose that PES is associated with Working Memory (WM) updating and the rehearsal of task-rules in WM in order to prevent future errors. Using a task-switching paradigm, we manipulated the need to update WM and rehearse rules by varying the information presented in the cues. Half of the participants received mapping cues providing information regarding both the relevant task identity and its rules, and half were shown dimensional cues providing only task identity information. As predicted, larger PES was observed with dimensional cues as compared with mapping cues.
PS_3.058 - Working memory capacity compensates hearing related phonological processing deficit
Classon, E. , Rudner, M. & Rönnberg, J.
Linnaeus Centre HEAD, The Swedish Institute for Disability Research, Department of Behavioural Sciences and Learning. Linköping University. Linköping, Sweden.
Post-lingually acquired hearing impairment is associated with decreasing phonological processing abilities. This study aimed to examine whether working memory capacity (WMC) compensates for this effect. Individuals with acquired hearing impairment (HI) and with normal hearing (NH) performed a visually presented rhyme judgment task in four conditions. Word pairs consisted of rhymes (R+) and non-rhymes (R-) that were orthographically similar (O+) or dissimilar (O-). The groups were matched on age, education level, WMC and verbal as well as non-verbal abilities. Each group was subdivided into high- and low-WMC individuals by a median split of reading span scores. In agreement with earlier studies NH performed significantly better than HI. Further, high-WMC individuals performed better than low-WMC individuals and this effect interacted with group and condition. Specifically, WMC had an impact on the performance of HI, but not NH. HI with high WMC performed on a par with persons with normal hearing. In contrast, HI with low WMC made significantly fewer correct judgments than HI with high WMC and NH with high or low WMC in the phonologically most demanding conditions (R+O-, R-O+). Results thus indicate good WMC can compensate for hearing related phonological processing deficit.
PS_3.059 - Maintaining cross-domain information in working memory, what resources are involved?
Langerock, N. , Vergauwe, E. & Barrouillet, P.
Developmental Cognitive Psychology Team, University of Geneva, Geneva, Switzerland
Domain-general attentional as well as domain-specific resources have been shown to be involved in maintaining verbal (e.g., letters) or visuo-spatial (e.g., locations) information in working memory However, it is not clear whether these same kind of resources are involved in maintaining cross-domain information (e.g., letters in locations). The involvement of domain-general resources was tested using a complex span task in which cross-domain storage of letters presented in different locations was combined with a neutral processing task (tone discrimination) involving either a low, a medium, or a high cognitive load. Span decreased as the cognitive load of the neutral processing task increased, which is in line with the involvement of domain-general resources. The involvement of domain-specific resources, over and above domain-general resources, was tested using the same cross-domain storage task but combined with either a verbal (semantic judgments) or a spatial (fit judgments) processing task, in which the cognitive load was manipulated in the same way. Interference with domain-specific resources was found only in the verbal domain. This suggests that domain-general attentional resources are clearly involved in maintaining cross-domain information, as well as verbal domain-specific resources.