3.13 DIMENSIONAMIENTO DE LOS EQUIPOS SERVIDORES
3.13.1 HARDWARE Y SOFTWARE DE LOS EQUIPOS SERVIDORES
Despite the fact that the number of the modelled acts reproduced by children has been commonly used as the robust measure of testing the hypothesis of imitation, the scoring procedure needs to be refined to take account of both the target and non-target acts that children potentially produce in a scoring period. In Experiment 1, the two- action coding strategy—a refinement of multiple scoring—was developed to examine the methodological issue considered above. In contrast to the dichotomous scoring strategy with which the scorer rated a range of actions a child produced during a scoring period as a whole, the two-action coding strategy considered only the first two distinct actions in the scoring period and coded each of them respectively. In addition, instead of coding the individual action depending on whether it matched the target act, the scorer assigned one of several pre-set scoring categories to each of the first and second actions. The two-action coding strategy was proposed for three reasons.
First, as an alternative to the latency measure, the inunediacy of imitation may be more adequately captured by the order in which the target and other non-target acts were performed. The assumption is that if a child learned to perform the target act by
imitation, she should elicit it more directly with a relative infrequency of exploratory
responses before it was produced. In cases where the target act was not produced directly at the first action, the other non-target acts produced ahead of it may perhaps indicate behavioural strategies different from that of imitation. Thus, coding both the target and non-target actions observed in the scoring period may help us gain more understanding of the behavioural strategies children used to achieve the target acts under different conditions. Therefore, in the present study, the first two actions observed in the recorded time period were independently coded in order to determine whether there might be group differences in the order in which the target and non-target acts were produced.
Second, the latency indicates how quickly, but not how directly, a child produces the target act. The child may spend a relatively short latency bringing about the target act, whereas there was still a likelihood that she elicited other non-target acts in a very short interval before it was produced. In the pilot work of Experiment 1, for example, it was noticed that at times the children spent less than 5 seconds producing a target act, but it was scored at the second action or even after that. On the other hand, under some circumstances it could take the children 15 seconds or more to complete a target act, perhaps due to carefulness or lack of manual dexterity. Nonetheless, the target act should be scored at their first action. It appeared that the order in which the target act was produced could not be ensured by the latency to generate it. Although the two- action scoring strategy developed in the present study did not take account of all the actions the child produced during the scoring period, it is useful to distinguish the case in which the target act was not produced directly at the first action. That is, the child’s production of the target act at the first action was assumed to be characteristic of the immediacy of imitation. In contrast, regardless of whether the child produced other non target acts preceding the target act, the dichotomous coding strategy (e.g., Meltzoff,
1988b, 1995) related the immediacy of imitation to the latency to produce it.
Third, in cases where a child’s responses observed in a scoring period did not meet all the operational descriptions of the target act, the scorer might find it difficult to rate these responses as a whole by simply assigning a yes or no credit. It may be plausible to
broaden the coding categories or measure the degree of match rather than reducing the responses to a dichotomy between the target and non-target action. However, neither the multiple scoring strategy nor the strategy of measuring the degree of match has
considered the number of actions that a child actually produces during a scoring period. Basically the traditional scoring strategies, discussed in this section, rate all the actions that a child produces during the scoring period as a whole. In order to examine this methodological issue, it is important to parse the child’s responses observed in the scoring period so that the actions, which are actually scored, can be recognised in individual segments. Instead of assigning an overall rating to all the actions that the children produced in the response phase, the two-action scoring strategy scored only their first and second actions distinctly observed in the recorded time period. In addition, it was anticipated that adopting a multiple scoring strategy would more adequately deal with the foregoing methodological issues.