Table of Contents | Previous | Next |
RELATIONS BETWEEN SPECIFIC AND GLOBAL FEATURES OF MOTHER-CHILD INTERACTIONS AND LANGUAGE
Catherine S. Tamis-Lemonda, Elizabeth Spier And Mark Spellmann New York University
Barbara Alexander Pan And Meredith Rowe Harvard University
The quality of parent-child interactions is one of the most powerful predictors of children’s emerging cognitive competencies, especially that of language. It is no wonder that researchers, practitioners, educators and parents alike have ubiquitously been concerned about the features of parenting that are most relevant to positive outcomes for children, as well as the best ways to capture and evaluate those features in research and practice settings. Indeed, numerous approaches to the coding of parent-child interactions are available, and decisions about which to use are guided by both theoretical orientation as well as practical constraints.
In the Early Head Start Consortium, although local and National research teams shared a conceptual interest in measures of parental stimulation, cognitive support, and sensitivity, they adopted different coding strategies to assess such parenting constructs. For example, National measures of caregiver-child interactions of parent-child free play (referred to as the “three-bag-task” given the placement of toys in three bags) are based on global ratings of six dimensions of behavior in mothers (i.e., sensitivity, intrusiveness, stimulation, positive regard, negative regard, detachment) and three in children (i.e., engagement, attention, negativity). For such ratings, coders assigned mothers (and children) a score of 1 to 5 on each construct after one or two “passes” or viewings of the interaction. Such global ratings are frequently relied upon in large-scale studies due to the time-efficiency of coding as compared to other labor-intensive approaches (such as the transcription of parent-child interactions, which can take up to 10 hours per 10 minutes of interaction). In contrast, researchers at many local sites, including Harvard and New York University, placed resources into describing and capturing specific aspects of parent-child engagements, by transcribing the full array of verbal and gestural exchanges between mothers and children during the 3-bag task. Such an approach is more frequently characteristic of small-scale, single-site investigations. Both “macro” and “micro” approaches to assessing parenting have merits, and both are fraught with limitations. Yet little is known about whether and how data obtained from the two relate to one another. Unfortunately, methodological integration, although empirically valuable, is rare.
Here, we explore associations between transcriptions of mothers’ and children’s language (at a local level) and global ratings of mother-child interactions (at the National level) as a first step toward understanding the interface between general and specific features of (and approaches to) dyadic engagements. We expected aspects of mothers’ language to relate to global measures of maternal sensitivity and stimulation, as mothers use language as a primary mode of engagement with children. Mothers who verbally respond to their children’s initiatives, provide language-rich environments, and ask questions of their children are likely to be those who are viewed as more sensitive and cognitively stimulating at a National level. Reciprocally, children begin to use language around the second year as a principal means of communication and as a way to maintain involvement in interactions with their caregivers. Thus, coders are likely to consider children’s verbal expressions as an index of their engagement, especially at this time.
Methods
To this end, research teams at Harvard Graduate School of Education and at New York University Graduate School of Education longitudinally examined mother-child discourse in a total of 146 dyads (balanced for child gender) during the three-bag task at 14 and 24 months. Mothers from the two sites ranged in age from 14 to 43 years at the time of their children’s birth. The sample was ethnically diverse: 47 percent identified themselves as White, 25 percent as African American, 17 percent as Latina, and 11 percent as other (e.g., West Indian, mixed ethnicity).
Maternal language samples were obtained through transcription of the three-bag, semi-structured play task used in the national protocol. With the assistance of Child language Data Exchange System (CHILDES) (MacWhinney & Snow, 1985), a computer program that uses electronic files of verbal transcripts to analyze various aspects of maternal and/or child speech, a count was obtained of the number of different words used by each mother and each child (i.e., word types); the total number of words used by mother and child (i.e., tokens), and the number of “wh” questions (e.g., “What is that?,” “Where is the blue block?”) used by each mother during the 14- and 24-month sessions. Global ratings of mother-child interactions were those coded at the National level from the three-bag task.
Results
As expected, mothers varied considerably in the amount of language (tokens) and diversity of language (types) they expressed toward their children at both ages. Similarly, children varied in word types and tokens at both ages, with variation increasing substantially by 24 months in line with children’s emergent productive language (see Table 1). Global measures of mothers’ and children’s behaviors varied at both ages as well.
Maternal language was strongly related to global ratings from the three-bag task at 14 and 24 months (see Table 2). Mothers’ total words, word types, and “wh” questions were positively associated with ratings of sensitivity, stimulation, and positive regard, and negatively associated with detachment (r range from .19 to .66, p < .05 to .0001). Mothers’ use of “wh” questions was negatively associated with negative regard and intrusiveness, although associations were small (r range from .19 to .21, p < .05). In general, findings were consistently robust across the three major ethnic groups. As an example, correlations between mothers’ language types and global ratings of stimulation in African-American, White, and Latina mothers were .67, .64, and .70 at 14 months, and .66, .45, and .76 at 24 months (all p < .001).
| 14-MONTH ASSESSMENT | 24-MONTH ASSESSMENT | |||||
|---|---|---|---|---|---|---|
| M | SD | RANGE | M | SD | RANGE | |
| MOTHER | ||||||
| Word Tokens | 508.2 | 266 | 30-1244 | 631.2 | 249.2 | 55-1294 |
| Word Types | 124.7 | 46.2 | 14 -221 | 160 | 49.1 | 29-320 |
| Wh- Questions | 10.4 | 9.63 | 0 - 22 | 12.3 | 11.5 | 0-83 |
| CHILD | ||||||
| Word Tokens | 6.58 | 10.46 | 0-63 | 95.8 | 72.3 | 0-333 |
| Word Types | 3.02 | 4.14 | 0-22 | 39.2 | 24 | 0-99 |
| Mother Language 14 Months | Mother Language 24 Months | |||||
|---|---|---|---|---|---|---|
| Types | Tokens | "Wh" Questions |
Types | Tokens | "Wh" Questions |
|
| Mothers' Global Ratings | ||||||
| Sensitivity | .36*** | .30*** | .43*** | .48*** | .46*** | .45*** |
| Intrusiveness | 0.05 | 0.1 | -0.15 | -0.06 | 0.02 | -.21* |
| Stimulation | .66*** | .61*** | .33*** | .57*** | .55*** | .34*** |
| Positive Regard | .54*** | .56*** | .46*** | .47*** | .44*** | .42*** |
| Negative Regard | -0.06 | -0.05 | -.20* | 0.08 | 0.11 | -.19* |
| Detachment | -.41*** | -.42*** | -.19* | -.48*** | -.48*** | -.24** |
| Children's Language 14 Months | Children's Language 24 Months | |||||
| Types | Tokens | Types | Tokens | |||
| Children's Global Ratings | ||||||
| Engagement | .20* | .19* | .51*** | .42*** | ||
| Attention | .17* | 0.16 | .48*** | .33*** | ||
| Negativity | -0.04 | -0.01 | -0.12 | -0.01 | ||
| * p < .05, ** p < .01, *** p < .0001 |
The robust associations identified between aspects of mothers’ language and the global measures of sensitivity, stimulation, and positive regard accords with the finding that these three measures covaried strongly at the National level, leading to the creation of a composite index of “supportiveness” (a composite score created by summing individuals’ ratings on the three items). Consequently, we tested the joint contributions of mothers’ language types, tokens and “wh” questions to the composite score of “supportiveness.” At both ages, maternal language types and “wh” questions (but not tokens) contributed unique variance to the composite measure of supportiveness, together accounting for 40 percent and 42 percent of the variance in “supportiveness,” at 14 and 24 months respectively.
In contrast to the consistently strong associations between mothers’ language and global ratings, the magnitude of associations between children’s language and ratings of their engagement, attention, and negativity varied with age. At 14 months, children’s word types and tokens were weakly associated with global measures of child engagement and attention (significant r range from .17 to .20, p < .05); by 24 months, however, associations among these same measures were moderate to strong (r range from .33 to .51, p < .001).
Discussion
One of the fundamental benefits of the National-Local partnership structure of the EHS consortium is the ability of investigators to integrate site-specific and National data, thereby shedding light onto the nature, meaning and ecological validity of both local and National findings. An area of inquiry in which this synergistic partnership is exemplified is in the merging of local and National measures of parenting, as illustrated in this investigation.
In general, results support the validity of National measures of parent-child interactions by demonstrating their strong associations to independently coded, in-depth measures of mother and child language at two local sites. It appears that the team of national coders were especially aware of the verbal exchanges between mothers and children when evaluating mothers’ sensitivity, stimulation, positive regard, and detachment and children’s engagement and attention. The fact that mothers’ language strongly related to global ratings of their interactions at both ages, whereas associations for children changed over age dovetails with developments in children’s language across the one-year period. At 14 months, children are at the dawn of productive language, and their verbal expression of “words” is limited. Consequently, coders likely rely on non-verbal aspects of children’s behaviors in their assessment of children’s engagement and attention. By 24 months, however, children’s verbal expressions become aptly central to coders’ evaluations of children’s engagement and attention, in line with the remarkable gains in language that occur at this age. These sensitivities in coders, and the fact that more cost-effective global ratings dovetail with findings at a micro-level, lends further support to the validity of the National findings on parenting. Importantly, these findings also bear on the training of program staff, who should be sensitized to the importance of mothers’ and children’s language interactions as key expressions and indicators of mutual sensitivity and cognitively rewarding dyadic interactions.Reference
MacWhinney, B. & Snow, C. E. (1985). The Child Language Data Exchange System. Journal of Child Language, 12, 271-296.
| Table of Contents | Previous | Next |