Characterizing talker variability in everyday life of urban, high SES 4- to 12-month-olds
Most research on the relationship between input and infants' language development focuses on the main caregiver, despite emerging reports that all input, be it from adults or children, counts for language development. The present study seeks to characterize the variability in the number of people that regularly talk to the infant and thus provide language input in a homogeneous population. In particular, the role of daycare format and infant age were examined.
To quantify potential differences in the number of input talkers, we developed a detailed talker variability questionnaire and administered it to children from urban high-SES families in France, aged between 4 and 12 months (n=216), divided into three age groups (4-, 6-, and 12-month-olds, n=84, 75, 57 respectively). In this questionnaire, parents noted who was talking to their child for more than 20 minutes during a typical week, with days being divided into "morning", "afternoon", and "evening". For each talker, parents also noted their sex, age, and level of education. In a revised version, the questionnaire contained accent information. All materials are openly available (link removed for abstract anonymization).
Concretely, the questionnaires were analyzed to answer the following questions:
Is there substantial variation in the number of talkers children are receiving their input from?
Does the choice of daycare format affect how many talkers are providing input?
Is there an effect of age that is separable from daycare?
The answer to the first question is yes, we find that children hear between 2 and 30 adults (counting everyone aged 13 and older) in a typical week. When counting children between 2 and 12 years of age as well, this maximum increases to 33. The median number of talkers is 7 (counting adults and children), and the distribution of talkers is biased towards lower numbers with a long tail. Further investigation revealed that 48% of infants heard children regularly, be it siblings or visitors.
To address the second question, namely whether and how daycare format affects talker variability, the data was split into whether children are at home with a parent, taken care of by a nanny (at least 3 weekdays), or visit group-based daycare.There is a significant difference in the number of talkers for each of the three daycare models, both when considering only adults and every person over 2 years of age. Interestingly, the lowest number of talkers is found for infants who are with a nanny (n=61, median=6 talkers), which is even less than observed for stay-at-home parents (n=96, median=7 talkers). As expected, children in formal daycare have the highest number of input talkers (n=59, median=10 talkers).
The last analyses concerned the third question, that is whether the number of talkers changes as a function of infant age. For example, parents might be more willing to engage in social events with an older child, a factor that would be independent from daycare. To first assess whether age and daycare are correlated, the data were split accordingly. Indeed, care models differed as infants grew older, with only 7% of the 12-month-olds having a stay-at-home parent, compared to 57% of children aged 4 to 7 months. There is no main main effect of age when predicting the number of talkers in infants' input; a significant interaction between age and daycare seems driven by the few oldest children with a stay-at-home parent who are exposed to more talkers than their younger peers with the same daycare model.
These results show that even within the same culture and SES group, considerable variability can be found in children's everyday experience, and this variability is partially driven by differences in daycare arrangements. Future research will address how this variability translates to differences in input, for example in terms of amount of infant-directed speech, word counts, or syntactic variability.
