a 0005 Information Processing Architectures : Fundamental Issues

abspara0010 This chapter starts with a brief history of information processing architectures, emphasizing a traditional experimental paradigm, the additive factors method, and the classic problem of model mimicry. Several solid approaches to identifying mental architecture are introduced with a discussion of the necessary assumptions for these tests. Other fundamental issues attached to information processing systems are also presented. A brief discussion with regard to challenges for the future concludes the chapter. p0010 In the late nineteenth century, budding experimental psychology began to investigate perceptual and mental operations that would be considered part of the contemporary purview of cognition. Such interests reappeared in the 1960s in the emerging information processing approach to elementary cognition. Over the last several decades of the twentieth century, considerable progress has been made in this tradition through the combination of mathematical modeling and experimental design, an approach becoming known as metamodeling. This section reviews a number of selected findings within this approach, which is relatively novel in the social sciences. One benefit of the strategy is that potential dilemmas involving model mimicking, where models based on strikingly different principles can imitate each other’s predictions, are brought to light. In particular, problems and experimental solutions associated with testing among parallel (simultaneous) versus serial (one-at-a-time) processing, stopping rules (logical basis for cessation of processing), limited versus unlimited versus Supercapacity (effects of increasing workload on processing speed), and dependence (issue of stochastic independence of channels or events) relationships are described. The foregoing issues are fundamental in almost all information processing situations since virtually any operating system must take them into account. The section concludes by pointing to future challenges within this approach. p0015 The term ‘information processing architecture’ refers broadly to the arrangement of mental subsystems that are hypothesized to be active in the performance of one or more psychological tasks. For instance, the simplest, most prototypical, and opposed types of architectures are serial (one-at-a-time) versus parallel (simultaneous) arrangement of two or more separate subsystems or processes. More complex arrangements are mentioned below (see Additive Factor Models). Further, there are a number of other aspects of perceptual and cognitive processing that are often included under the ‘architecture’ label including the questions of the basis on which cognitive processes will cease, various kinds of independence and dependence, and processing capacity. These will be described and discussed. p0020 An early researcher especially pertinent to this review was Donders, a nineteenth-century Dutch physiologist. Donders believed that he could uncover the durations taken by various thought processes through his method of subtraction. The method of subtraction was based on the idea that complicated mental activities were compounded in a simple sequential fashion from less complex parts. Let mean response time be written as RT and response time in general as simply RT. Then, for instance, the scientist might engage the subject in a task requiring both perception and decision and compare RT from that condition with RT from a task requiring only perception. The difference in RTs was interpreted as the mean duration of the decision operation. p0025 The issues selected for review here seem elemental in the following sense. The construction of almost any system intended to carry out the processing (e.g., detection, search, comparison, recognition, recall, analysis of various kinds, and so on) of a finite number of tasks or items would have to make decisions on each of the studied issues. In addition, they are somewhat unique in having been subjected to quite general theoretical, quantitative, and methodological analysis, perhaps more than any other such concepts in the field. The issues will be introduced in the context of a popular experimental paradigm and then discussed in more detail. s0010 An Experimental Paradigm and the Major Issues p0030 The concepts to be defined below have figured prominently in studies on short-termmemory and visual display search and we shall employ that type of paradigm for illustration. Considerable impetus was given to the information processing approach by several pioneering studies in the 1960s focusing on shortterm memory and visual processing, using RT as the dependent variable ( bib2 Atkinson et al., 1969; bib10 Egeth, 1966; bib28 Sternberg, 1966). In the roughly 45 years of the interim, scores and perhaps hundreds of studies in memory and visual search have been carried out. We focus on bib28 Sternberg’s (1966) short-term memory search paradigm for illustration. Short-term memory is specified operationally by the tasks requiring the retention of a small number of items for anywhere from a few seconds to several minutes. In this paradigm, a varying number (less than or equal to the number that can be maintained in shortterm memory without error) of items, for instance, randomly arranged letters, is presented to the experimental participant. This is called the memory set. Then, a few seconds later, that participant is presented a so-called probe item and the task is International Encyclopedia of the Social & Behavioral Sciences, 2nd edition http://dx.doi.org/10.1016/B978-0-08-097086-8.43103-X 1 ISB2 43103 To protect the rights of the author(s) and publisher we inform you that this PDF is an uncorrected proof for internal business use only by the author(s), editor(s), reviewer(s), Elsevier and typesetter TNQ Books and Journals Pvt Ltd. It is not allowed to publish this proof online or in print. This proof copy is the copyright property of the publisher and is confidential until formal publication.