IOE EPrints

Mental mechanisms, contextual effects and the processing of word-attributes.

Jessel, John. (1992) Mental mechanisms, contextual effects and the processing of word-attributes. PhD thesis, Institute of Education, University of London.

[img]
Preview
Text
128583.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (3029Kb) | Preview

Abstract

It is known that the presence of information related to, but not necessary for the completion of a speeded experimental task can affect performance. Previous studies involving such 'contextual' effects with information in verbal and nonverbal form have produced mixed results which have been attributed to differing underlying processing mechanisms. The present work examines the extent to which some possible mechanisms could, in common, account for within- and crossmodality contextual effects. 'Priming' information relating to typographic case was presented over time intervals varying up to 480 msec before, and up to 480 msec after, a target to which a speeded category-decision was indicated by a left- or right-hand key-press. Within-attribute' case-priming was effected using a letter string printed either in upper- or lower-case (e.g., `dghej', or 'YFLRB') paired with a similar target for case-decision. Within-attribute name-priming upon name-decision was similarly achieved using the case-denoting words `upper' and 'lower'. 'Cross-attribute' priming was possible through name information priming case-decisions (e.g., `upper-YFLRB'), or, case information priming name-decisions (e.g., 'YFLRB-upper'). The combined effects of case- and name-priming upon case-decision were also investigated using case-denoting words printed in either one or other of the two typographic cases (e.g., `upper-yflrb', or `UPPER-yflrb'). Relatively short decision latencies accompanied by relatively low error rates occurred when targets were preceded by congruent within-attribute primes. Incongruent priming led to slower responses with higher error rates. Similar effects were obtained with name-priming of case-decisions although, in comparison to within-attribute priming, these occurred after longer prime pre-exposures. Case-priming effects upon name-decisions were obtained for simultaneous as well as pre-exposed primes, however, congruent as well as incongruent case-priming led to relatively slow decision latencies and higher error rates. Combined congruent case and name information speeded case-decisions in comparison to congruent case information alone. However the converse did not occur; increase in case-decision latency with incongruent case and name information combined was no greater than with incongruent case information alone. It has been argued that the results are consistent with models which propose that input is initially subject to encoding where 'activation' may spread automatically among interconnected verbal or nonverbal representations. In particular, this could partly account for speeded responses obtained with congruent primes. Relatively shorter decision latencies can also be attributed to subsequent decision processes and the generation of responses. The results also suggest that relatively longer latencies could arise from competing incompatible responses having to be suppressed. It has been further argued that relative speed of processing different stimulus attributes and the form of experimental task can be major determinants in the detection of contextual effects.

Item Type: Thesis (PhD)
Controlled Keywords: Human information processing,Symbolic processes,Error analysis,Models,Cognitive style,Memory
Subjects: Departments > Psychology and Human Development
Depositing User: Mr Andrew Welshman
Date Deposited: 12 Jun 2014 14:35
Last Modified: 23 Mar 2015 11:25
View Record in Library Catalogue: http://ioe.sirsidynix.net.uk/uhtbin/cgisirsi/x/0/0/5/3?searchdata1=a128583{CKEY}
View Item View Item