Responding to Joint Attention at 6-months and Outcome

HOME

CLASSES

RESEARCH

VITA

LETTERS

ADVISING

ECEDP

Responding to Joint Attention at 6-months and Outcome

in Infants with Typical Development

As noted in prior papers in this symposium, the capacity to follow direction of gaze, or respond to joint attention, is thought to be a key component of joint attentional skills. In this presentation, I will discuss a study examining the age of onset of the ability to respond to joint attention, and the relations between individual differences in this skill during infancy and early vocabulary acquisition.

Although quite a bit of research has been devoted to examining the development of the ability to follow direction of visual gaze, findings have been somewhat mixed. For example, several studies have found that infants demonstrated the capacity to follow direction of gaze beginning as early as six months of age, while others have found that this ability did not emerge until at least well into the latter half of the first year. One explanation has attributed the discrepancy in findings to differences in the stringency of operational definitions of joint attention used across studies. However, recent work by Morales, Mundy, and Rojas (1998) using a more stringent operational definition of joint attention found infants reliably demonstrated the ability to follow direction of gaze at 6 months of age, and that individual differences in this ability were positively correlated with early vocabulary development. Because of the potential importance of these data for current discussion regarding the age of onset of joint attention, as well as for the possible utility of measures of gaze following as indices of early childhood outcome, the present research was designed to provide a replication of these findings.

To do this, we followed a group of 52 infants across the first year. We conducted laboratory assessments of infant's ability to follow direction of gaze when infants were six months of age, and conducted language assessments when infants were 12 months of age.

The participants were 52 infants (26 boys, 26 girls) and their mothers who were participating in a longitudinal study of social-communication development. The infants came from two-parent middle to upper middle SES families (Hollingshead, 1978). Maternal education ranged from 1 year of college to postgraduate level, with the median education level being four years of college. Fourteen children were Caucasian (non Hispanic), sixteen were Hispanic, five were African American, and 17 were of multi-ethnic background (Hispanic and Anglo).

The measures used in this study included an adapted version of the Early Social Communication Scales Responding to Joint Attention Task, to assess children's ability to follow the direction of gaze, and the MacArthur Communicative Development Inventory, to assess children's early vocabulary development. Let me briefly describe each of these measures.

The assessment of RJA was carried out in a 3 X 3 m sound attenuated laboratory playroom when infants were 6 months of age. The infant was seated in a high chair, and the parent was seated facing her infant. Three wooden targets were mounted on the walls of the laboratory. The first target was positioned 90 degrees to the left of the infant, at infant eye level, the second target was positioned 90 degrees to the right of the infant, at infant eye level, and the third target was positioned 180 degrees behind, and slightly above, the infant. Infants and mothers were videotaped using two wall mounted cameras.

The assessment of RJA was administered within the context of a parent-child interaction. Each session began with a 7-minute face-to-face interaction. Immediately following this interaction, RJA was assessed using the adapted version of the Early Social Communication Scales Responding to Joint Attention Task (Mundy, Hogan, & Doehring, 1996). This procedure consisted of two sets of a "Toy, Tickle, Look" sequence in which RJA was assessed during the Look segment. During the "Look" segment, mothers were instructed to present three consecutive trials to their infant in which they turned and fixated a target that was 90 degrees to the right or left of the infant, or 180 degrees behind the infant. During each head turn trial, parents said their child's name emphatically three times while maintaining their gaze on the referent target throughout the trial. After each trial, mothers returned their gaze to midline before executing the next trial.

We had two independent coders rate videotapes for direction of the first infant gaze or head turn in the horizontal plane to occur during each trial. A gaze or head turn was scored as correct if it was in the same direction of the adult's head turn. A gaze or head turn was scored as incorrect if it was in the opposite direction of the adult's head turn.

The rating procedure yielded four infant response scores; the total number of left and right trials that an infant's first response was correct (range=0-4); the total number of behind trials that an infant's first response was correct (range=0-2); the total number of left and right trials that an infant's first response was incorrect (range=0-4); and the total number of behind trials that an infant's first response was incorrect (range=0-2).

The infant response scores were used to compute a difference score for each infant. The difference score was calculated by subtracting the number of incorrect responses from the number of correct responses. These scores were used to test whether or not infants reliably demonstrated the ability to follow gaze. In order for this skill to be demonstrated, infants had to look in the same direction as the parent significantly more frequently than they looked in the opposite direction.

With respect to our language measure, we used the MacArthur Communicative Development Inventory or MCDI. The MCDI is a parent-report instrument that is designed to assess a wide range of language and language-related skills. We used the MCDI Short form, Level I version for infants to assess both receptive and expressive vocabulary at 12 months. This version of the form contains a 89 word checklist. Parents are given a standard set of instructions in which they are asked to indicate which words their child "understands", and which words their child "understands and says".

Table 1 presents the means, standard deviations, and the range for the number of trials in which infants first response was correct or incorrect for left and right trials combined, and behind trials.

Table 1. Descriptive Statistics for Infant Response Scores

Trial Type

Left/Right Behind

Responses
	M	SD	Range	M	SD	Range
Correct	1.62	1.22	0-4	0.03	0.19	0-1

Incorrect	0.90	0.82	0-4	0.77	0.83	0-2

The average number of trials that infants responded correctly was about 2 for left and right trials combined, and the average number of trials that infants responded incorrectly was around 1 for left and right trials combined and behind trials. Also, approximately 66% of the sample was able to match the adult's response on two or more trials.

To test if infants were reliably demonstrating the capacity to follow the direction of gaze, mean difference scores were calculated for left and right trials combined (M=.73, SD=1.52), and behind trials (M=-0.73, SD=0.87). One- Sample t tests were conducted using mean difference scores. In order for responding to joint attention to be demonstrated, we required that infants show a difference score greater than 0. The results indicated that the mean difference score for left and right trials combined was significantly greater than 0 (t(51)=3.46, p<.001), while the mean difference score for behind trials was significantly less than 0 (t(51)=-6.07, p<.0001).

These findings replicated our 1998 study and suggest that a substantial number of the 6-month-old infants in the current sample reliably demonstrated the capacity to follow adult direction of gaze when parents fixated targets within the infants' visual field. However, as in our previous study, infants were unable to follow direction of gaze when adults fixated targets behind or outside the infants' visual field.

We next conducted Pearson-Product Moment Correlations to examine the associations between individual differences in the capacity to follow direction of gaze and language acquisition. Findings indicated that individual differences in infant's ability to follow direction of gaze, as indexed by difference scores for left and right trials, were positively related to infant's receptive vocabulary, r(52)=.28, p<.05.

In sum, the current study replicated our prior findings with respect to age of onset of joint attentional skills, and with respect to the connections between early gaze following skills and vocabulary acquisition. Thus, we believe that these data, in conjunction with prior research, support the contention that the ability to follow direction of gaze may emerge as early as six months of age.

We recognize that according to some definitions, this type of behavior at six months may not reflect true joint attention skill development, as it is not clear that six month gaze following reflects the social cognitive understanding of others as intentional agents (Tomasello, 1995). However, our observations of an association between individual differences in early gaze following and language, both here and in our earlier study (Morales, et al. 1998) supports the construct validity of six month gaze following as a form of joint attention.

We also recognize that the data on language in this study reflect only parents' perceptions of early language skill which may be influenced by their perception of their child's ability to coordinate attention. However, our earlier observation of an association between 6 month gaze following and 30 month language performance on tester administered assessments mitigates this concern to some degree.

To resolve this issue, it may be useful to return to arguments raised at the beginning of this symposium. Perhaps it is best to think of joint attention skills as a gradually emergent behavioral facility (Moore, 1994). Ultimately, joint attention skill may come to reflect epistemological aspects of social cognition (Tomasello, 1995). However, prior to that, individual differences in joint attention skill development may reflect differences in attention regulation, social applications of basic cognitive processes, and parameters effecting learning and motivation (Corkum & Moore, 1997; Mundy & Gomes, 1997). All of these processes, in combination or perhaps in interaction with the infant's care giving environment, may play a role in the development of subsequent social, cognitive, and communication differences among children.

Moreover, current data would suggest that an important goal for future research may be to examine the utility of measures such as RJA as possible indices of early childhood outcomes, as it is clear from the last two papers that individual differences in RJA skills, perhaps as early as six months of age, may provide unique and important information about different developmental trajectories of young children.