Developing Mental Models: A Child's Conception of a Computer System Between the Ages of Nine and Fourteen

Denham, Pearl Patricia (1990) Developing Mental Models: A Child's Conception of a Computer System Between the Ages of Nine and Fourteen. PhD thesis, University of Glasgow.

Abstract

The purpose of this research was to determine whether a child novice-user of a computer system possessed a mental model of that system. If such evidence existed, to establish its nature, causally or terminologically related to the model. A necessary element of the research was to devise a method of data capture appropriate to the nature of the data to be collected and to the youth of the population. Four studies established the effectiveness of subject drawings as a method of data collection. Initial studies on the internal appearance of the computer identified a working model with a common set of five components. Three of them were fundamental functions; Communication-links, Memory and Input/output. More of the common set of five components are found to be included in the model as age increases, although the 'preferred' or ranked order of component choice was reliably maintained across three age-bands. Communication, a pre-requisite for an interactive system, was perceived by virtually all the subjects. 'Communication-links' were identified in three modes. The most important mode, reflecting component inter-dependence, was identified in the drawings of a substantial and increasing number of subjects across the three year age-range. Whilst it was shown that a significant variation of choice pattern existed within each age-band, age itself was a significant influence on mode choice. A subsequent study of 122 subjects, to identify modes of data transmission between keyboard and visual display unit, revealed four forms. The most advanced mode proposed a conduit together with evidence of data transmission using keyboard characters, electrical current and/or impulses. It was found that as the subject became older he is more likely to explain effectively the hidden system processes between input and output by including an intermediate step. Furthermore, the intermediate step is more likely to be a memory component and the subject is more likely to illustrate the explanation with visible transmission of data. The conceptual model of memory was found to be weak with no sense of its fundamental role in accommodating system and user needs of data retrieval. Throughout the study memory was generally recognised by the novice only when the use was explicit, implying that it is 'empty' most of the time and, because it is perceived as a formless space, no structures existed for subsequent data retrieval. In both cases it was shown that the perception could be modified by experience. Recognition of memory as an indespensible component was found to be age and/or experience related and, at post-test only, that a relationship existed between year-band and the model of memory storage adopted. There were also reliable indications that different age groups had different beliefs about how data was stored in memory, particularly in terms of addressed memory locations. There were also indications that intermediate experience modified the subject responses positively and that the combinatorial effect of age and interim experience had a positive effect on the results. Finally, a relationship between subject age and the use of equations when allocating memory space to arithmetic results was found. Memory, as the hidden process, was contrasted with an observable process, the visual display unit. Here it was found that experience and the combined effects of experience and age-group had a positive effect on whether subjects believed that visual display screens were 'blank' during processing, that is, blank screens were predicted less often. Perceptions of the hidden process between keyboard and visual display unit fell into three categories. One group included an intermediate step using 'memory', sorting/ordering and 'others', the incidence increasing between years 1 and 3. It was predicted that the model type was not unrelated to the year of the subject and also, that where an intermediate process was shown, the choice of process model was related to the yearband. Language, is believed to play a role in mental model development and when early work with subjects revealed confusion between the everyday and esoteric roles of computer terminology, it was explored through several studies. When asked to classify a list of computer related items according to their perceived system role no group observation was as expected, whether the items were given as individual words or as a sentence. It was also shown that subjects from the different age-bands acted similarly in the way they classified the task roles of the items with clear direction to the 'Command' category. However, the single Memory category revealed that the average number of allocations to the memory category is not unaffected by the year group of the subject. The results generally indicated a growing awareness of system memory requirements. The subjects were also asked to classify computer related words as 'picture', 'word' or 'nothing' evoking and the rankings across the age-range 1st to L4 for 181 subjects, were found to be not unrelated. Finally, a study examined whether the novice possessed an adequate and helpful model which would enable the correct analysis and predicted outcome for a 'bugged' computer program. A high percentage of the subjects anticipated a correctly running program and the predicted program effect for the three age-groups was not dissimilar.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Developmental psychology
• Date of Award: 1990
• Depositing User: Enlighten Team
• Unique ID: glathesis:1990-78425
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 30 Jan 2020 15:28
• Last Modified: 30 Jan 2020 15:28
• URI: https://theses.gla.ac.uk/id/eprint/78425