jurnal psikologi sosial

REPRESENTATIONAL FLEXIBILITY IN THE SECOND YEAR OF LIFE : AN EXPLORATORY STUDY OF THE EYE TOY

Abstract

The ability to comprehend symbolic representations is deemed as an imperative property of human cognition in the midst of this symbol-flooded world. This study aimed to investigate the early development of this ability and its relation to self-recognition in the second year of life. I presented children with a novel video game (SONY PlayStation 2 Eye Toy or ET), in which child’s own image is incorporated in the game’s display. Using detailed observation of children’s responses to the experimental and control conditions, I investigated if they recognized this unusual symbol-referent relation. I also surreptitiously marked each child on their forehead to study if they recognized themselves in the video image (VSR).
The underlining assumption is that in order to pass the self recognition task, children should grasp basic correspondence of image-referent in the video reflection. Thus, those who passed the visual self-recognition task were hypothesized to also demonstrate better performance on the Eye Toy game than those who did not pass. To examine the consistency of this trend in the second year of life, I compared two groups of children from early and later part of this particular time span.
Participants consisted of 8,24-month-olds and 12,30-month-olds. Results revealed that there was no difference between older and younger children performance in both ET and VSR tasks. Although children did display some insight to the novel symbol- referent relation in the game, their performance in ET and VSR task was not related. Implications and suggestion for future research in this area are discussed in light of the new findings.

Introduction
Symbols play a prominent role in human modern environments. They have been a pervasive part of our everyday lives in various forms, ranging from words, pictures, gestures, number, etc (Park & Clements, 1995). The use of symbols has greatly benefited human society in terms of communication and learning (DeLoache & Smith, 1999; Park & Clements, 1995). Through symbols in literature, for example, we can acquire information about historical events and places without direct experience, thus expanding our learning opportunities.
A symbol can be defined as “something that someone intends to represent something other than itself” (DeLoache, 2004, p.66). In order to comprehend the meaning of particular symbol, we need to recognize the relationship between the symbol and what it represents, which is known as the referent (Park & Clements, 1995).
So how does this ability to understand symbols come about ? obviously children are not born into this world with an immediate ability to interpret and use symbols around them. The discrepancy between younger and older children’s understanding of symbol-referent relation implies that such ability develops though the lifespan
The child was presented with a novel video game, known as the Eye Toy (ET), on a PlayStation2. This game involves a live camera incorporating the child’s own image into the game’s display. Instead of a conventional joystick, the child’s body movements act as the input device to interact with the computer-generated stimuli (e.g. balloons, bubbles, sparkles, etc.) on the screen. Unlike touch-screen technology, the child does not manipulate the depicted stimuli by directly touching the screen. Instead, interaction occurs between the child’s depicted image and the stimuli on the screen. The trick of the game is that a child needs to use his or her own video image to guide his or her interaction with the computer-generated stimuli in the game.
The uniqueness of the ET game is reflected through the existence of computer-generated stimuli (virtual symbol) in the screen. These virtual symbols (e.g., balloons) co-exist with the child’s (live) image in the screen. While the child’s image is representing the child (self) in the current reality, the virtual symbols only exist in the screen and have no direct reference to current reality. In addition, the child through his or her reflection can also interact with these virtual symbols. This type of symbol-referent relation is considered novel because in previous experience with such pictorial representations (e.g., photos, pictures, television, videos), the current reality could not directly affect the symbols (Pierroutsakos & Troseth, 2003; Troseth, 2003; Troseth & DeLoache, 1996).
In order to play with this novel game, children are required to have some understanding about their depicted self and also about the virtual symbol. There are two ways in which a child may perform as if he or she understands a symbol. The child could either be trained or conditioned to respond, or the child has insight into the symbol-referent relation. Therefore, the ability to play with a novel game carrying an unusual symbol-referent relation, such the Eye Toy, can only be done, at least initially by having a representational insight.

Research Question
The present study aims to assess children’s understanding of symbol or external representation by applying a new experimental approach. A unique and novel symbol-referent relation in ET was presented to the child. We know that adults and older children had no difficulty to understand and play with such technology. However, we are interested in how younger children will respond to a novel symbol-referent relation with which they had no prior experience.
In addition to having an insight to the novel symbol-referent relations in the game, children also need to use their image as a guide or proxy for the interaction. This requires children to correctly infer the identity of reflection, an ability which is labeled as self-recognition (Gallup, 1997).
Hence, we will first use external representations to assess children’s representational ability and then we will explore the extent to which their performance in the representational task (ET) reflects their self-recognition ability as a measured by video self-recognition task (VSR).

Visual Self-recognition
The self-recognition test involves unobtrusive application of rouge or mark to facial features visually inaccessible without a mirror (Gallup, 1979). In order to pass this test, children need to use the external representation of self, which is expressed in forms of mirror reflection or video image, to gain ‘new’ information about the referent (the presence of a mark or sticker) and guide the retrieval behaviour. If this is the case, we would expect that the so called ‘self recognition’ ability will be dependent on children’s representational capacity.
The main difference from the traditional version is the use of video reflection instead of mirror reflection. This medium alteration is seen as crucial considering the direct comparison that will be made with the following representational task (ET) which also uses video technology.
Based on previous findings (Amsterdam, 1972; Johnson, 1983; Gallup, 1977), children’s reaction towards reflections can be categorized into two clusters of behaviors: mirror directed behavior and self-directed behavior. Most children under two years of age treat their reflection as if it was another person (Johnson, 1983). They reach toward the mirror, look behind it, and try to manipulate their own reflection. These behaviors are categorized as mirror-directed behavior, which according to Perner (1991) reflects child’s inability to detect the correct relation between symbol and referent.
The second category of behavior, which normally appears during second year of life, is known as (mirror-mediated) self-directed behavior (Johnson, 1983). This behavior is characterized by children’s use of mirror to guide behaviors directed to self. In particular, investigation of the mark on their own indicates that the child considers the reflection as being about him or herself (Perner, 1991; Zelazo et al., 1999).

Connecting Representational Ability and Visual Self-Recognition
This corresponding developmental trend between performances in the two tasks seems to support the Perner’s (1991) theory. Firstly, children’s performance in the search task and the self-recognition task could be attributed to the same basic mechanism shared by the two tasks, which is to identify the symbol-referent relation (Perner, 1991; Suddendorf, 1999). Secondly, the emergence of representational insight and self-recognition in the second year of life are expressions of the same capacity; secondary representation (Perner, 1991; Suddendorf & Whiten, 2001).

ET task and VSR task
In regards to similarity of symbol-referent mapping inside the task, the ET task is directly comparable to live video self-recognition (VSR) task. By formulating VSR in terms of symbol referent relation, it was evident that ET task and VSR task share the same conceptualization of video representation as a symbol and self as a referent. The fundamental difference between these two tasks, however, is their principal criterion. In the VSR, the sticker or mark exists in the children’s current reality. If children understand that the video image is referring to the self, then they can aim to retrieve the sticker by reaching to their own head (screen mediated self-directed behavior). On the contrary, the target stimuli (balloons and sparkles) of the ET task are nonexistent in the immediate reality. They exist only as symbols on the screen. Therefore, to manipulate these stimuli children need to monitor their screen reflection as a guide for their movement towards the stimuli (screen mediated object-directed behavior).
In this study, the ET task was always administered prior to VSR task to avoid any transfer effect to the ET performance. While there have been a lot of studies providing the data for children’s VSR performance, children’s performance on the ET task is investigated for the first time in this study. Thus, with such sequence arrangement, carry-over effect from VSR task can be controlled.
The symbol-referent relation in the ET task can be considered as bidirectional in nature. A symbol would normally contain information about the referent (e.g. child) in the real world, hence allowing the child to act upon the information (DeLoache, 2004). But with the ET the direction of relationship also goes the other way. The referent (self) can also affect the symbols in the virtual world (virtual stimuli) via its video representation. To fully grasp this notion bi-directional symbol-referent relation, children will need to demonstrate flexibility in understanding external representation.
This study presented two different tasks to children in their second year of life: the VSR and ET task. Three main hypotheses were tested. Firstly, we want to test whether children as young as 30 months old and 24 months old could understand the nature of symbol-referent relations behind the ET task. If 24 and 30 month-olds could link their actions to the virtual symbols in this ET task, then we hypothesized the older children would perform better on the ET task than the younger ones.
Secondly it was also predicted that older children would outperform younger children on the VSR task. Previous research suggested (Simcoc et al., 2004) that a majority of children pass this task by 30 months of age. Thirdly, we hypothesized that children who passed the VSR task would perform better in the ET task than those who failed the VSR task.

ET task
The eye toy test consisted of 8 consecutive 30s sessions forming a double-ABBA design. There was a 10s break between each session. These eight sessions were divided into two subtest based on the type of stimuli used; balloons and sparkles. In the balloons subtest, different color balloons are descending from the top to the bottom of the screen and children can interact with these balloons simply by touching them. Children also received the sparkles game, in which sparkling bright color appeared on the screen as a result of their movements.
Each subtest (balloons test or sparkles test) consisted of four sessions that alternated experimental condition and control condition in either BAAB or ABBA configurations. Within the control condition, the virtual stimuli are non-interactive, hence child’s movement will not inflict any effect to the stimuli.

VSR task
For the live video self-recognition test, the toddler was engaged in a distracting game so that the experimenter could surreptitiously affix a bright sticker to his or her forehead (randomly to the left or right fringe). After several trials (distracting game) the experimenter used praise as an opportunity to affix the sticker on the child’s forehead.
Performance in VSR test was rated as pass or fail. Similar to passing criteria in previous self-recognition study, participant passed the test by retrieving the sticker or reaching within 2 cm from the sticker during a 30 second time period when they are exposed to the live video feedback.

Discussion
Taken together, these results suggested that both younger and older children could discriminate between experimental and control condition in both sub-tests (balloons and sparkles). Not only that, but this difference in activity level across conditions also implied that children as young as 24 months of age could display an insight to the ET game.
There are several possible explanations to the VSR finding. Firstly, in this study we used the ET video camera, which has a lower picture resolution than normal digital or analogue video cameras used in another studies. Hence, children were exposed to a video representation of self in relatively poor quality, which may have increased the difficulty level to recognize the sticker.
Secondly, the VSR task was always administered after the ET task to avoid any transfer effect to the ET performance, which is our main interest. However, in turn there might have been a carry-over effect from the ET task to VSR task. In this sequence, children had to switch from a self-affecting-video (ET task) to a video-affecting-self (VSR task) relationship. Recognition might have been more difficult. Children with a flexible representation should be able to distinguish the different symbol-referent relations contained in each tasks. However, for those whose representational ability is still fragile, the sticker could be mistaken for another virtual stimulus that does not exist in reality. Consequently, they would display screen-directed behavior in the VSR task.
In contrast to our hypothesis, there was no evidence of relationship between ET and VSR task in the second year of life. Again, the predicted direction of relationship emerged although it was not significant. Two parsimonious explanations were provided. First, this result could be attributed to the carry-over effect in the VSR task. Secondly, in light if Suddendorf’s (1999) study which found significant relationship between surprise mark and surprise object ( a type of representational problem-solving task) task in 3 and 4-year-olds, it is possible that such relationship is not distinctive in the second year of life and only becomes distinct in the third and fourth year of life.
Finally, after evaluating the current results, a more conservative approach in theoretical level should be taken into consideration. Observing the simultaneous or nearly simultaneous emergence of several capacities in development should not necessarily be taken as strong evidence for the actual causal relations to some underlying process or to each other. The simultaneous emergence two accomplishments may well be a coincidence or be related to something much more general in development. In this context too much was made of the apparent correspondence between the development of VSR and understanding of symbol-referent relations. A problem exacerbated by emphasis being placed on the observation that as children get older they do better on measure of these abilities.

Conclusion
In conclusion, the present findings bring some implications to the research of VSR and representational insight. Interactive-computer technology (ET) was first introduced to examine the representational ability of children in the second year of life. The results revealed that there was no difference between 24-month-olds and 30-month-olds performance in the ET task. In addition, the difference in activity level across ET conditions was also observed, suggesting an evidence of insight. However, while it is clear the children were able to discriminate between experimental and control condition in ET task, it is not really clear whether representational insight fully accounts for this trend. The overall findings, however, suggested that there was no evidence of a relationship between self-recognition and representational ability, as measure by ET and VSR task, in the second year of life. Finally, the ability of ET task in this study to investigate children representational ability is far from convincing, hence future research is needed.