AW indicates adult words; CSW, content-specific words; CT, conversational turns; CV, child vocalizations; and RESP, parent responses.
Sosa AV. Association of the Type of Toy Used During Play With the Quantity and Quality of Parent-Infant Communication. JAMA Pediatr. 2016;170(2):132–137. doi:10.1001/jamapediatrics.2015.3753
The early language environment of a child influences language outcome, which in turn affects reading and academic success. It is unknown which types of everyday activities promote the best language environment for children.
To investigate whether the type of toy used during play is associated with the parent-infant communicative interaction.
Design, Setting, and Participants
Controlled experiment in a natural environment of parent-infant communication during play with 3 different toy sets. Participant recruitment and data collection were conducted between February 1, 2013, and June 30, 2014. The volunteer sample included 26 parent-infant (aged 10-16 months) dyads.
Fifteen-minute in-home parent-infant play sessions with electronic toys, traditional toys, and books.
Main Outcomes and Measures
Numbers of adult words, child vocalizations, conversational turns, parent verbal responses to child utterances, and words produced by parents in 3 different semantic categories (content-specific words) per minute during play sessions.
Among the 26 parent-infant dyads, toy type was associated with all outcome measures. During play with electronic toys, there were fewer adult words (mean, 39.62; 95% CI, 33.36-45.65), fewer conversational turns (mean, 1.64; 95% CI, 1.12-2.19), fewer parental responses (mean, 1.31; 95% CI, 0.87-1.77), and fewer productions of content-specific words (mean, 1.89; 95% CI, 1.49-2.35) than during play with traditional toys or books. Children vocalized less during play with electronic toys (mean per minute, 2.9; 95% CI, 2.16-3.69) than during play with books (mean per minute, 3.91; 95% CI, 3.09-4.68). Parents produced fewer words during play with traditional toys (mean per minute, 55.56; 95% CI, 46.49-64.17) than during play with books (mean per minute, 66.89; 95% CI, 59.93-74.19) and use of content-specific words was lower during play with traditional toys (mean per minute, 4.09; 95% CI, 3.26-4.99) than during play with books (mean per minute, 6.96; 95% CI, 6.07-7.97).
Conclusions and Relevance
Play with electronic toys is associated with decreased quantity and quality of language input compared with play with books or traditional toys. To promote early language development, play with electronic toys should be discouraged. Traditional toys may be a valuable alternative for parent-infant play time if book reading is not a preferred activity.
Variation in early language development is in part determined by a child’s language environment.1 The quantity of language input received from caregivers during the first few years is positively associated with a child’s language accomplishments and this early advantage can have long-lasting implications for overall academic success.2,3 Previous evidence suggests that in addition to the quantity of language exposure, measures of interaction quality also influence language development.1 For example, maternal responsiveness and the number of adult-child conversations both positively affect language development.4,5 Beyond family-specific factors, such as socioeconomic status and parental education, little is known about everyday activities that may promote both the quantity and quality of communicative interactions between parents and young children. Television exposure, book reading, and independent and guided play are activities that have been investigated. Television exposure was found to be associated with decreased quantity and quality of parental language input, and media viewing by children younger than 2 years has been shown to negatively affect language development, likely because media use displaces other more beneficial language-promoting interactions.6- 8 On the other hand, book reading and playing together with young children are often recommended as activities that boost language development. Clinic-based programs, such as Reach Out and Read, that distribute books to impoverished families have been shown to improve language ability, and a variety of different play activities, including early symbolic play and block play, have been linked to better language ability in toddlers.9- 12 Based on these and other findings, the American Academy of Pediatrics discourages media use by children younger than 2 years and emphasizes the importance of book reading and other types of parent-child play time.13 However, the reality for many families of young children is that opportunities for direct parent-child play time is limited owing to financial, work, and other familial factors. Thus, optimizing the quality of limited parent-child play time is imperative.
At the same time that parents are being encouraged to read to their young children and engage in direct parent-infant play, they are also bombarded with advertisements for “educational” toys that claim to promote language development in very young children, including infants. These toys are typically battery-operated electronic toys with buttons that produce lights, sounds, music, words, and phrases when activated. Other options parents have when selecting toys for their very young children are traditional nonelectronic toys, including blocks, puzzles, and stacking cups, as well as toys that encourage symbolic play such as dolls and tea sets. Primary care professionals and pediatricians working with families of young children may be called on to provide recommendations regarding selection of toys for infants and toddlers, particularly for those children and parents for whom reading may not be a preferred activity.
The purpose of this controlled experiment was to determine whether the type of toy used during parent-child play time influences the quantity and quality of the communicative interaction in ways that are known to be associated with better language development.
To promote language development, parents are encouraged to read to their infants and spend time playing 1 on 1, but little is known about how different types of play activities affect parent-infant communication.
In the present study, play with books and traditional toys was superior to play with electronic toys in promoting high-quality communication.
Children vocalized less during play with electronic toys than during play with books.
Because play with electronic toys is associated with decreased quantity and quality of language input, it should be discouraged to promote early language development.
My research team and I conducted a controlled experiment using observational measures of a volunteer community sample of parent-infant dyads recruited through posting of flyers in public places frequented by parents of young children in Flagstaff, Arizona, a midsized city in the southwestern United States. Enrollment and data collection occurred over a 16-month period between February 1, 2013, and June 30, 2014. All procedures were approved by the institutional review board at Northern Arizona University. Written informed consent was obtained from parents.
Pairs of parents (mother or father) and infants (aged 10-16 months) were eligible. Parent-infant dyads were excluded if the participating parent did not use English as the primary language with the child. Because race/ethnicity and socioeconomic status may affect parent-child communication, this information was collected via parent report using a written questionnaire with set categories for race/ethnicity and maternal education level.
To maximize the ecological validity of study findings, participants engaged in study procedures in their own homes without being directly observed by researchers. On enrollment in the study, audio recording equipment, 3 sets of toys, and study paperwork were delivered to participants. The recording equipment used was the LENA Pro System (Language Environment Analysis; LENA Foundation). The system includes a small digital recording device called a digital language processor that is placed in a pocket in a vest worn by the child. The processor records up to 16 hours of recorded sound and is worn continuously by the child for at least 10 hours. The accompanying LENA software conducts automatic analyses of the recordings and generates estimates of the amount of speech produced by adults in the child’s environment, the number of child vocalizations, the number of adult-child conversational turns, and the amount of exposure to electronic noise (eg, television). The algorithms used to generate these automatic analyses are designed to calculate estimates over an extended recording period rather than specific counts during a short period. For that reason, the automatic analyses were used to obtain general information about the language environment of each child but were not used in data analysis for the main research questions.
Each dyad engaged in 2 15-minute play sessions per toy set over a 3-day period, resulting in 30 minutes of play per day. Participants were given a daily log indicating which toy set to play with each day and in which order. To minimize potential order effects, 6 different daily logs were generated, each representing a different presentation order. As participants were enrolled, they were assigned a daily log-in sequential order (from 1 to 6 and then starting again at 1). All but 1 participant followed the order presented on their daily log; this participant did only 1 play session on the first day and 3 play sessions on the second day. Participants were instructed to engage in play sessions at their convenience during the day and play sessions did not need to be back to back. Parents were instructed to play as they usually do with their child and were not explicitly told to minimize distractions (eg, presence of siblings, pets, and television). Start and stop times for each play session were recorded by the parent on the daily log so that they could be found easily on the audio recordings for coding and analysis.
All toys were selected based on their potential to elicit speech centered on 3 semantic themes: animal names, colors, and shapes. The electronic toy set consisted of 3 battery-operated toys with buttons and switches that can be manipulated to produce lights, words, phrases, and songs. The electronic toys included a baby laptop, a talking farm, and a baby cell phone. These 3 toys were selected because they are marketed as educational toys that promote language development for children in this age range and are advertised as teaching animal names, colors, and shapes. The traditional toy set consisted of 3 nonelectronic toys that also have the potential to teach animal names, colors, and shapes. These traditional toys included a farm animal chunky wooden puzzle, a shape-sorter with 10 pieces representing 5 different shapes and 5 different colors, and a set of 10 small multicolored rubber blocks with pictures of animals and common objects on each side. The books toy set consisted of 5 board books: 2 with a farm animal theme, 2 with a shape theme, and 1 with a color theme. Three of the books included lift-the-flap opportunities.
Each play session was transcribed by a research assistant and coded for the variables of interest: (1) number of adult words, (2) number of content-specific words, (3) number of child vocalizations, (4) number of conversational turns, and (5) number of parent responses per minute. These measures were operationally defined as follows. An adult word was any word produced by the parent during the play session. Words produced by other adults who may have been present in the interaction were not included in the adult words count. A content-specific word was a word produced by the parent that was in 1 of the 3 semantic categories that the toys were selected to elicit (ie, animal names, colors, and shapes). Child vocalization was a speechlike utterance consisting of, at minimum, a voiced vowel. Non-speechlike utterances, such as cries, grunts, and laughs, were not coded as child vocalizations. A conversational turn was a speechlike utterance by either the parent or the infant that occurred within 5 seconds of an immediately preceding utterance by the other speaker. Parent response was a speechlike response by the parent that occurred within 5 seconds of an immediately preceding child vocalization. A parent utterance was only counted as a parent response if it was determined to be a direct response to the preceding child utterance; a parent utterance that occurred within the 5-second window but was judged to be unrelated to the previous child vocalization and not prompted by the child vocalization (eg, speaking to a pet or other adult) was not counted as a parent response. The conversational turn measure is a purely quantitative measure of back-and-forth vocal interactions between infant and parent while parent response measures direct parental verbal responsiveness to a child’s vocalization. Per-minute counts for each outcome variable were averaged across the 2 play sessions with each toy set; this average per-minute count was used for all analyses. Reliability coding was conducted by a second research assistant for all of the play sessions for 15 of the participants.
Pearson-product moment correlations were calculated between coder 1 and coder 2 results for each outcome variable in each toy condition. Intercoder agreement was very high for all outcome variables in all conditions (overall r = 0.981; range, 0.925-0.999). Coders could not be blinded to the condition because the toy set was identifiable by listening to the play session.
My research team and I analyzed the data using repeated measures analysis of variance for each of the 5 outcome variables, with the toy set as the within-participants factor. This design allows each dyad to serve as its own control, thereby eliminating the influence of dyad-specific factors that may influence the general quality of the parent-infant communication interaction (eg, age of the child, developmental level of the child, and socioeconomic status of the family). We used post hoc analyses to explore differences in outcome measures between individual toy sets.
Thirty-seven parent-infant dyads were enrolled in the study. Seven participants were not home at the scheduled material delivery time and 4 did not complete all study procedures, resulting in useable data from 26 parent-infant dyads. Twenty-five parent participants were biological mothers and 1 parent participant was the biological father. Two children were born prematurely and were identified with moderate developmental delay; parents of the other children reported no developmental concerns. The sex of the child, age of the child, race/ethnicity of the child, and maternal education are presented in Table 1. Demographic data from the 7 families who were not home at the delivery time are not available. Of the 4 families who did not complete all study procedures, 2 were Hispanic and 2 were non-Hispanic white, suggesting overrepresentation of Hispanic families in the nonparticipant group.
Average per-minute counts with 95% CIs for each outcome variable by condition are displayed in the Figure and reported in Table 2 along with repeated-measures analysis of variance P values. Main effects of condition were statistically significant for all outcome variables. My research team and I conducted post hoc analyses for each outcome variable to examine significant differences for the 3 possible toy set contrasts (ie, electronic vs traditional, traditional vs books, and books vs electronic); Bonferroni-adjusted P values for each 2-way contrast are reported in Table 3.
During play with electronic toys, there were fewer adult words (mean, 39.62; 95% CI, 33.36-45.65), fewer conversational turns (mean, 1.64; 95% CI, 1.12-2.19), fewer parental responses (mean, 1.31; 95% CI, 0.87-1.77), and fewer productions of content-specific words (mean, 1.89; 95% CI, 1.49-2.35) than during play with traditional toys or books (Table 2). Children vocalized less during play with electronic toys (mean per minute, 2.9; 95% CI, 2.16-3.69) than during play with books (mean per minute, 3.91; 95% CI, 3.09-4.68). Parents produced fewer words during play with traditional toys (mean per minute, 55.56; 95% CI, 46.49-64.17) than during play with books (mean per minute, 66.89; 95% CI, 59.93-74.19) and use of content-specific words was lower during play with traditional toys (mean per minute, 4.09; 95% CI, 3.26-4.99) than during play with books (mean per minute, 6.96; 95% CI, 6.07-7.97).
Results showed that the largest and most consistent differences were between electronic toys and books (with greater values for all measures with books), followed by electronic toys and traditional toys (with larger values in the traditional toy condition for all measures except child vocalization), and the least consistent and smallest differences were between books and traditional toys (with larger values in the book condition for adult words and content-specific words only).
In this controlled experiment in a natural environment of parent-infant communicative interaction during play, my research team and I found that the type of toy used during the play session was significantly associated with both the quantity and quality of language use. For all outcome measures, play with books provided a better communication interaction than play with electronic toys. The effect was most pronounced for the quantitative measures of parental language use, suggesting that parents tend to let the toys do the talking for them when their child is interacting with electronic toys. This is particularly worrisome given that there is no evidence that children this young are able to learn vocabulary from media or other nonhuman interactions.14,15 For the 2 qualitative measures of parent-infant interaction, conversational turns and parental responses, play with traditional toys was equivalent to book reading, yet both were superior to play with electronic toys. While play with books may provide a script that encourages parents to talk more and produce more words associated with the themes in the books, it was not superior to play with traditional toys in promoting parent-child conversations and parental responsiveness. For parents who may not be inclined to read to their preverbal infants or for parents whose children do not prefer book reading activities, play with traditional toys, such as blocks and shape sorters, may be an equally valuable use of limited parent-infant play time. This is consistent with and may help explain results of previous work that found that distribution of blocks to families of young children resulted in better child language ability.12
While the results of this study are robust given that the play sessions were conducted in the participants’ homes rather than in a controlled laboratory setting, there are important limitations that should be taken into account. The most notable limitations were the small sample size and the relative homogeneity of participants by race/ethnicity and socioeconomic status. Given the robustness of the results despite the small sample size, however, it is likely that the findings would be replicated with a larger group of participants. The question of whether results would differ for families from more diverse demographic backgrounds is an important one that remains to be answered. Given that cultural and socioeconomic factors are known to influence both the amount of time spent engaged in parent-child play and the nature of the play interaction, it is important to acknowledge that the effect of toy type on parent-infant communication may be different for parents from different cultural and economic backgrounds.16- 18 Furthermore, the findings of this study are applicable to children in a very limited age range of prelinguistic and very early language development and effects may be different for older toddlers and preschoolers. Another limitation is that investigators did not have complete control over the timing of the play sessions and it is possible that participants may have chosen to play with certain toy sets when either the parent or child was more tired, thereby impacting the communicative interaction.
Despite these limitations, the results provide an empirical basis for primary care professionals and pediatricians for making recommendations regarding the types of play activities that promote rich language interactions.
These results provide a basis for discouraging the purchase of electronic toys that are promoted as educational and are often quite expensive. These results also add to the large body of evidence supporting the potential benefits of book reading with very young children. They also expand on this by demonstrating that play with traditional toys may result in communicative interactions that are as rich as those that occur during book reading. I do not claim that book reading and play with traditional toys are developmentally equivalent activities; book reading provides numerous benefits that cannot be replicated in other activities (eg, literacy socialization and exposure to novel vocabulary and concepts) and play with toys, such as blocks and puzzles, provides developmental and cognitive benefits beyond the language domain.19- 22 However, if the emphasis is on activities that promote a rich communicative interaction between parents and infants, both play with traditional toys and book reading can be promoted as language-facilitating activities while play with electronic toys should be discouraged.
Corresponding Author: Anna V. Sosa, PhD, Department of Communication Sciences and Disorders, Northern Arizona University, 208 E Pine Knoll Dr, 15045, Flagstaff, AZ 86011 (email@example.com).
Accepted for Publication: October 18, 2015.
Published Online: December 23, 2015. doi:10.1001/jamapediatrics.2015.3753.
Author Contributions: Dr Sosa had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded by a research grant from the American Speech-Language-Hearing Foundation.
Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The following individuals participated in data coding while graduate students at Northern Arizona University and received hourly financial compensation as student research assistants, providing statistical consultation and analysis assistance for this project: Lauren Crane, MS, Brooke Santos, MS, Heather Wiest, MS, and Luke Plonsky, PhD (Northern Arizona University).