2.2 ALMACENAMIENTO DE INFORMACIÓN
2.3.1 INTERFAZ DE COMUNICACIÓN POR MEDIO FÍSICO O CABLE
Unfinished target acts
As none of the children’s second acts fell into this category, only the
proportion of unfinished target acts produced at the first action was analysed. There was a significant condition effect in the proportion of unfinished target acts produced at the first act, F (3, 36) = 3.52, p< 0.03. The infants in the Full-Demonstration condition produced unfinished target acts more frequently than they did in the Adult- Manipulation control (p < 0.03), and that was the only pair of group difference revealed by follow-up Tukey HSD tests.
Other acts produced at first act
The analyses of other acts produced as the children’s first acts served as a supplement to the analyses of their performance on target acts. The question asked was whether the infants’ behavioural tendency to produce the target acts as their first acts had a relation to various types of demonstrated acts they had seen. When the infants produced the target acts at the first act, they could have learned about these acts directly without trial and error. That is, they should have produced many fewer irrelevant other acts as their first acts. The results showed that there was a significant difference in the proportion of other acts produced at the first act as a function of condition, F (3, 36) = 8.94, p < 0.001. Pairwise comparisons using a follow-up Tukey HSD test showed that the infants in the Full-Demonstration condition produced fewer other acts falling within the category of Other Act at the first act, compared to what the infants did in the Failed-Attempt, Emulation-Leaming, and Adult-Manipulation conditions, p < 0.03, p < 0.005, and p < 0.0001, respectively. No group differences were found among these latter three conditions.
Using a finger to activate the beeper in the object set o f box and stick
Cases where the infants used the finger to push the button to activate the beeping noise in the object set of box and stick were coded independently in a sub category. Although the infants did not watch the experimenter directly manipulate the object set in the Emulation-Leaming condition, they might anticipate the button-
Table 2.2. Number of 19-month-old children using a finger or stick to activate
Action
Condition First action First and second actions combined
Stick Finger Stick Finger
Full-Demonstration Failed-Attempt Adult-Manipulation
7 2 7 2
Emulation-Leaming 1 3 1 6
pushing action upon hearing the beeping (18-month-old infants being acquainted with buttons that are “pushable”) and thus used a finger to push the button. In contrast, the infants in the Full-Demonstration, Failed-Attempt, and Emulation- Leaming conditions had witnessed the experimenter’s manual contact with the stick, and that perhaps would cue them to use the stick to push the button. Table 2.2 provides the number of children using the stick and a finger to activate the beeper in the Emulation-Leaming condition and the other three conditions combined. Chi- Square tests using a Fisher’s exact test on the number of children using their finger to activate the beeper at the first action did not reveal a reliable difference between the Emulation-Leaming condition and the other three conditions combined.
However, the difference was significant when counting the first and second acts combined, p < 0.05 (Fisher’s exact test). Overall, the infants who successfully
activated the beeper in the Emulation-Leaming condition more frequently pushed the button with their finger, and those from the other three conditions combined were more likely to use the stick.
Object parts infants first touched
Whereas the infants in the Emulation-Leaming control, like those in the other three conditions, did see the experimenter put the objects down and lift them up at the beginning of the demonstration, they did not watch the experimenter’s manual contact with the object sets in relation to their transformation. Instead, the infants in the other three conditions watched the experimenter always start by acting on certain parts of the objects (i.e., cube, stick, loop, beads, and square). The infants might orient themselves to the object sets by observing the experimenter’s first contact with the objects, and that could lead them to produce the target acts more efficiently.
Table 2.3. Means of the proportion of parts of objects 19-month-old children Consistent with Different from the Touching more
Touched part the demonstration demonstration than one part No response
Condition Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Full-Demonstration .80 (.21) .16 (.16) .04 (.08) .00 (.00)
Failed-Attempt .82 (.18) .06 (.10) .08 (.14) .04 (.08)
Emulation-Leaming .60 (.23) .34 (.27) .06 (.13) .00 (.00)
Adult-Manipulation .64 (.20) .22 (.26) .12 (.14) .03 (.08)
Thus, the parts of the object sets that the infants first touched in the scoring period were coded in several ways. Table 2.3 displays the means of the proportion of object parts the infants in the four conditions first touched, according to whether the touched parts were consistent with, or different from what the experimenter first handled, or whether they started by touching more than one part, or did not touch the object set at all. One-way ANOVAs were performed on the data with the exception of the measure of No Response because cases where the infants did not touch the object sets at all rarely happened
There was a significant difference in the proportion of first-touched object parts which were consistent with those the experimenter had first touched, F (3, 36) = 2.98, p < 0.05. Pairwise comparisons using a Tukey HSD test, however, did not show a reliable difference between any pair of groups. However, a significant effect of condition was indicated in the proportion of first-touched object parts which were different from those the experimenter had first touched, F (3, 36) = 3.19, p < 0.04. Follow-up Tukey HSD tests revealed a difference in only one pair of groups: the infants in the Emulation-Leaming condition more often first touched the different parts of the objects from the experimenter than the infants in the Failed-Attempt condition, p < 0.03. No significant effect of condition was found as regards occasions when the infants touched more than one part when making their first contact with the object sets.
Latency to produce target acts^
The time the infants spent producing the target acts was analysed to study how quick they were to bring about such acts after seeing various types of demonstration. The means of the latency were calculated for the target acts produced as the
children’s (1) first acts, (2) first and second acts combined, and (3) overall
performance in the 20-second response period. As pointed out in Section 2.2.5, there has been no conclusive evidence that infants after seeing the full demonstration of the target acts would produce the target acts more quickly than they did after seeing other types of demonstration. However, previous studies did not specify the order in which the target acts were produced (e.g., Devouche, 1998; Meltzoff, 1988 a, b, 1995). According to this account. Experiment 1 presents an attempt to analyse the latency to produce target acts according to the act at which children’s performance on target acts took place (Table 2.4).
As in the foregoing analyses, a one-way ANOVA were performed on each of the three measures and that was followed by Tukey HSD tests. Counting the
children’s first acts only, no significant difference was found among the four
conditions in the latency to produce target acts, F (3, 28) = 1.26, p > 0.3. In contrast, there was a significant difference as a function of condition when the children’s first and second acts combined were scored, F (3, 30) = 4.62, p < 0.009. Pairwise
comparisons showed that the infants in the Failed-Attempt condition took an overall shorter latency to produce the target acts as their first and second acts combined than the infants in the Emulation-Leaming and Adult-Manipulation conditions, p < 0.03, in both. There was no difference among the Full-Demonstration, Emulation-
Leaming, and Adult-Manipulation conditions. Also, neither was any difference found between the Full-Demonstration and Failed-Attempt conditions conceming the latency to produce target acts as the first and second acts combined.
^ The latency to produce target acts was not reported in Bellagamba et al.’s study. In Meltzoff’s study, only the latency to produce target acts as the children’s overall performance in the 20-second response period was reported. Similar to Meltzoff s findings, the author found that the Full-Demonstration and Failed-Attempt groups were similar in both the number of target acts produced in the 20-second response period and the lateiicy to produce them. The two groups did not differ in the latency to produce the target acts at the first act, but the Full-Demonstration group produced more target acts at the first act than the Failed-Attempt group. See Section 2.4 for discussion concerning scoring issues.
Table 2.4. Means of the latency for target acts pr first and second actions combined (TA-1®^ + 2" ),
reduced at first action (TA-1^% and in 20 seconds of the
Latency (sec) Condition TA-L‘ Mean (SD) TA-L'4-2"'^ Mean (SD) TA-20 sec Mean (SD) 5.04 (3.66) 5.12(3.07) 5.41 (2.61) n = 10 n = 10 n = 10 3.65(1.41) 3.68(1.41) 5.91 (4.66) n = 9 n = 9 n = 9 6.05 (3.71) 7.85 (3.33) 8.03 (3.12) n = 9 n = 10 n = 10 6.65 (2.81) 8.90 (4.62) 10.91 (5.06) n = 4 n = 5 n = 7
Ful 1-Demon strati on Failed-Attempt Emulation-Leaming Adult-Manipulation
Note. ”n” is the valid number of children.
Finally, latency to produce target acts was computed for the children’s overall performance in the 20-second response period. A significant effect of condition was found, F (3, 32) = 4.64, p < 0.008. Pairwise comparisons showed that the infants in the Adult-Manipulation condition took a longer latency to produce target acts as their overall performance in the 20-second response period compared to what they did in the Full-Demonstration and Failed-Attempt conditions, p < 0.009 and p < 0.03, respectively. No difference was found among the Full-Demonstration, Failed-
Attempt, and Emulation-Leaming conditions. Also, no difference was found between the Emulation-Leaming and Adult-Manipulation conditions.