Play interactions of autistic preschoolers with their mothers and fathers without toys yield more positive interactions than play with toys

Abstract

Observations of parent–child play with toys are often used to assess interactions between parents and non-autistic as well as autistic children, but some research indicates that play without toys may elicit more positive interactions than play with toys. The first goal of the study was to examine whether this is true in the case of autistic preschoolers by comparing their play with their parents with versus without toys. The second goal was to compare mother– and father–child interactions. Seventy-eight autistic preschooler boys were observed interacting in counterbalanced mother– and father–child play with or without toys, and interactions were coded using the Emotional Availability scales. Results revealed that for both mothers and fathers scores on all of the EA scales in the play episode without toys were higher than scores in the episode with toys. Also, almost no differences were found between children’s interactions with their mothers and fathers in both play with or without toys. Both findings broaden our understanding of the early social experiences of autistic children with their parents.

Lay Abstract

Researchers that study the development of young children often observe them playing with their parents with toys. However, a few studies and clinical experience suggest that the interactions between parents and children are often more positive when they play without toys. The first goal of the study was to find out if this is true for autistic preschoolers. We observed 78 children playing with each of their parents, and compared their play with and without toys. The second goal of the study was to compare mother– and father–child interactions to discover if they are similar or different from one another. We measured the emotional quality of parent–child interactions using scales that measure the behavior of both parents and children toward one another. We studied only boys because autism may present differently in girls. We discovered that both parents’ and children’s behavior were more positive when they were playing without toys compared to when they were playing with toys. Also, when we compared the emotional quality of how children played with their mothers and their fathers we found almost no differences, both in how the parents and the children behaved. The findings are important because they can guide researchers, clinicians, and parents to include both play with toys and play without toys in studies, interventions, and daily life. They also emphasize the importance of fathers, because although the fathers in the study were somewhat less involved in everyday caregiving with their children compared to mothers, they had interactions with the children that were as positive as those children had with their mothers. Importantly, we do not know if the findings apply to autistic girls as well.

Keywords

Autism spectrum disorders emotional availability fathers parent–child interaction play

Studies of the interactions between parents and young children often use observations of play episodes in which the parent and the child are invited to play with toys “as they usually do” (Brito et al., 2014). However, most studies do not ask whether this particular context, namely, free-play with toys, provides the optimal conditions for sensitive and attuned parent–child interactions. This is an important question because a few studies with non-autistic children suggest that the presence of toys (vs play without toys) may be less conducive for promoting shared emotional experiences between parent and child (e.g. Sjolseth et al., 2024). The impact of the presence of toys on parent–child interactions may be even more important when autistic children are concerned, because of the difficulties these children show in initiating and maintaining reciprocal and synchronous interactions (Diagnostic and Statistical Manual of Mental Disorders (5th ed., DSM-5); American Psychiatric Association, 2013). Furthermore, at least one study of autistic children (Dolev et al., 2009) showed that mother–child play without toys was associated with more emotionally available interactions compared to mother–child play with toys. The primary goal of the study was therefore to compare the interactions between parents and their autistic preschool-age children when the dyads are engaged in play with toys versus without toys, and to examine whether play without toys yields interactions which are more positive compared to play with toys.

An additional goal of the study was to focus on father–child, in addition to mother–child interactions. Although their importance and distinct contributions for the development of non-autistic children is acknowledged by many (e.g. Bakermans-Kranenburg et al., 2019; Cabrera et al., 2018), fathers are still less frequently included in studies of families with autistic children. This limitation was highlighted in two review papers (Flippin & Crais, 2011; Perzolli et al., 2021) that called for greater involvement of fathers in developmental and intervention studies with autistic children (see also Rankin et al., 2019). The inclusion of fathers in the present study provided an opportunity to examine whether the hypothesized differences between parent–child interactions during play with versus without toys will be the same with both parents. Also, this was an opportunity to examine similarities and differences between mother– and father–child interactions in the case of autistic children.

Parent–child interactions with and without toys

Although the differences between parent–child interactions with or without toys have generally received little research attention, a few studies point to the advantage of play without toys, at least in infancy. With regard to non-autistic children, Maas et al. (2013) found that mothers of 6-month-old infants expressed the most positive regard and the least flatness during face-to-face play with their infants compared to free play with toys (or a diapering interaction), and Frosch et al. (2019) found that ratings of mother–infant emotional connection during face to face interaction predicted fewer behavior problems at the age of 3 years, whereas the same ratings but obtained from mother–infant toy play did not. Along similar lines, Sjolseth et al. (2024) examined how mothers’ bids for interaction with their 6-month-old infants were related to the duration of shared emotional experiences, and found that these experiences were longer following mothers’ efforts to engage their infants in play without toys. The authors suggest that when toys are incorporated into mother–infant play, there may be more disruptions to shared eye-contact and face-to-face moments, whereas without the use of toys, maternal bids are more direct and create a shared connection. Play without toys therefore appears to facilitate a focus on one another of mother and child as relational partners in infancy (Sjolseth et al., 2024). Whether or not the same is true for non-autistic preschoolers (rather than infants) has not been studied, to the best of our knowledge. In fact, it is possible that the advantage of play without toys in the case of non-autistic children is specific to infancy, because face to face interactions are more prominent in this age. Clinical work and a few research findings suggest that when autistic preschoolers are concerned, play without toys may be particularly beneficial, however.

For example, with regard to clinical work, the “Hanen More Than Words” intervention (Carter et al., 2011) stresses that “People games,” namely, social games played with people without the use of toys, are optimal for language development of autistic children when compared to play with toys. This point is similarly emphasized in the Developmental, Individual differences, Relationship-based model (DIR, Greenspan & Wieder, 2006). Perhaps because autistic children often focus more on objects than on people (Akin-Bulbul & Ozdemir, 2024; Chawarska et al., 2016; Vivanti et al., 2017), the absence of toys promotes interactions between the children and their play partners, whereas the presence of toys may distract the autistic child from the play partner.

There is only little empirical support the advantage of play without toys for parent–child interactions in autism, however. In one study of children with “Complex Communication Needs,” Deveney et al. (2016) found that parents were more likely to be directive and responsive during play with, rather than without, toys. An additional study (Dolev et al., 2009) compared mothers’ interactions with their autistic preschool-age children in three contexts: Free play (with toys), Structured play (with toys), and Social play (without toys). Mother–child interactions were coded using the Emotional Availability Scales (EAS, Biringen et al., 2014), and the results revealed that the interactions were consistently more optimal in the social play episode (without toys) compared to the other play observations (with toys). Dolev et al. (2009) suggested that because the social play episode involved no toys, mothers felt free to follow their children’s lead, and matched the level and type of play to what they knew would be most enjoyable for their children and most likely to engage them. This often involved simple sensorimotor games involving heightened positive affect, singing, and movement (e.g. “Itsy Bitsy Spider”), which children seemed to enjoy as evident in their positive affect and requests to continue the games.

Based on Dolev et al. (2009) we hypothesized that parent–child interactions would show higher emotional availability during play without toys compared to play with toys. One limitation of the Dolev et al. (2009) study was that the order of the play episodes was not counterbalanced, and it was therefore not possible to rule out order effects. Specifically, the play without toys episode was last, and a warming-up effect could account for the high scores in this episode. Therefore, in the present study the order of the play episodes was counterbalanced.

Mother– and father–child interactions

As mentioned above, an additional goal of the study was to examine father–child, in addition to mother–child interactions. Flippin and Crais (2011) proposed that because of fathers’ distinct interactional styles with their children, they may have an important influence in facilitating play and language skills of autistic children. For example, Flippin and Watson (2015) found that fathers’ (but not mothers’) verbal responsiveness was associated with children’s language skills after controlling for children’s cognition. With regard to comparisons between mother– and father–child interactions, the focus of the present study, studies reveal a mixed picture. Results of the Flippin and Watson (2015) study mentioned above revealed that children initiated more leads during their play sessions with their mothers than with their fathers, and that mothers were more verbally responsive than fathers. Other studies did not find differences between mother–and father–child interactions, however. Elder et al. (2002) observed mothers and fathers of 22 preschoolers with an autism diagnosis interacting with their children. Observations were coded on several “social reciprocity” parent and child variables, and results showed no differences between mother–child and father–child interactions. Similar findings were obtained by Hirschler-Guttenberg et al. (2015), who studied parent–child interactions of 39 preschoolers with autistic spectrum disorders (ASDs) and found that the mothers and the fathers expressed similar levels of sensitivity, responsiveness, and appropriate limit setting. Finally, Bentenuto et al. (2020) observed 40 children with an ASD diagnosis interacting with each one of their parents. The order of the observations with the parents was counterbalanced, and play observations were coded for their EA (Biringen et al., 2014). Here too no differences between mother–child and father–child interactions on both parental and child EA scales were found. Importantly, although the absence of mother–father differences appears consistent across most of these studies, the samples sizes were modest and possibly too small to detect differences. Therefore, in the present study a larger sample was used.

The current study

The study had two goals. The first was to compare parent–child interactions with versus without toys, with the hypothesis that interactions without toys will be more emotionally available than those with toys. Whether this will be equally true for father–child interactions as for mother–child interactions was difficult to predict because so little past research involved fathers. However, because the studies reviewed above point to the similarity between mother–child and father–child interactions, we expected that the differences between play with versus without toys will be the same with fathers as with mothers.

The second goal was to compare mother–child and father–child interactions. Based on the majority of the studies reviewed above we expected no differences between mother–child and father–child interactions in both types of play observations (with or without toys). The larger sample employed in this study (N = 78), approximately twice the size of the largest previous samples (Bentenuto et al., 2020; Hirschler-Guttenberg et al., 2015), was expected to increase the likelihood to detect differences should they exist. In addition to the comparison of mean level differences between play with or without toys and between mother– and father–child interactions, we also examined the stability of the EA parent and child scales (Bornstein et al., 2006) across contexts (play with vs without toys) and parents (mother–child vs father–child interactions).

Only boys were included in the study. This was done because studying autistic boys and girls as one group, particularly in a study of social behavior, may conceal gender differences. It has been argued that autism may present differently in boys and girls (Hull et al., 2020), and that girls may be more adept at masking (“camouflaging”) their autistic traits (Tubío-Fungueiriño et al., 2021). Ideally separate and large groups of boys and girls should be studied so that each gender would receive dedicated research attention. In this study only boys were included because autism is more prevalent in boys compared to girls (in the United States: Maenner et al., 2020; in Israel: Davidovitch et al., 2013), the gender issues presented above notwithstanding. In addition, and strictly from a statistical power standpoint, combining boys and girls would decrease the size of each group (considering a given sample size) which would decrease the study’s power.

Method

Participants: Seventy-eight Israeli preschooler boys with ASD and their parents participated in the study. Only boys were selected to maximize power as explained above. Power analysis conducted using G*power software for a repeated measures analysis with two levels (toy vs non-toy play) with alpha set at .05, power at 95%, a small to medium effect size (.25) and a correlation of .5 between play with versus without toys, revealed that 54 children were needed. A larger sample was recruited to allow for the additional within subject factor (mother vs father) and to include a covariate (IQ). Children’s ages ranged between 29 and 68 months (M = 49.45, SD = 11.03). Thirty-six of the children (45%) were first born. Mothers’ age ranged between 25 and 47 years (M = 36.62, SD = 5.24) and fathers’ age ranged between 29 and 57 years (M = 39.80, SD = 6.29). Mothers and fathers had a mean of 15.31 (SD = 3.11) and 15.09 (SD = 2.79) years of education, respectively. All families were Jewish. Fathers and mothers worked 46.85 (SD = 12.22) and 33.28 (SD = 14.62) hours per week, respectively, and this difference was significant (t = 6.59, p < .001). Children spent an average of 37.75 h per week in preschool (SD = 14.28). When not in preschool, 80% of the children were cared for by their mothers, 11.7% by a relative or professional caregiver, and 8.3% by the father. Finally, 48.1% of the parents reported dividing caregiving responsibilities equally, 43.2% sharing responsibilities with mothers doing most of the caregiving, 4.9% sharing responsibilities with fathers doing most of the caregiving, and 3.7% the responsibilities were solely on the mothers. Families were recruited through advertisements in treatment centers for children with ASD and through social media. The advertisements included a short description of the study and inclusion criteria. These were the child being male, a known diagnosis of ASD for the child (which was later confirmed in the study), age between 2.5 and 6 years, no known additional medical diagnoses, and two biological and married parents. Families who responded to the advertisements were invited to participate, and the vast majority consented.

Procedure: The study included two laboratory visits: The first laboratory visit included the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000) and an observation of family interactions which is not part of this report (Oppenheim et al., 2023). The second laboratory visit, conducted 1–2 weeks later, included an assessment of children’s cognitive functioning and observations of the child with each parent in two parent–child interactions: Play with toys and play without toys. Parental questionnaires were completed digitally between the two laboratory visits. The study received approval from the University of Haifa IRB (400/16). All parents signed written consent forms. There was no community involvement in this study.

Child diagnostic measures

Autism Diagnostic Observation Schedule (ADOS)

The ADOS is a standardized and widely used assessment for diagnosing individuals with ASD through interactions with a trained professional (Lord et al., 2000). It consists of 4 modules of which one is administered based on the child’s level of expressive language. Thirty-five children were assessed using module 1, 37 using module 2, and 6 using module 3. The ADOS diagnostic algorithm was used (Gotham et al., 2007). Reliability of blind assessors on 17 cases (21% of the sample) was 87.4% agreement (range: 78% to 94%).

Social Communication Questionnaire (SCQ)

The SCQ is a 40-item questionnaire completed by parents regarding children’s autism symptoms (Rutter et al., 2003). The SCQ items are scored as present (1) or absent (0) and summed, with the cut-off for ASD for young children ⩾ 11 (Wiggins et al., 2015). The mean SCQ score was 20.59, SD = 4.97, range = 12–32.

Children’s ASD diagnosis was confirmed based on the ADOS and the SCQ as well as by a licensed psychologist with extensive experience assessing children with ASD. In all cases the diagnoses that the children received in the community were confirmed.

Assessments of children’s cognitive functioning

The Wechsler Preschool and Primary Scale of Intelligence—Third Edition (WPPSI–III; Wechsler, 2002) is a widely used and standardized cognitive assessment for children between the ages of 2:06 and 7:07 years which yields an IQ score and was used for verbal children. Non-verbal children were assessed using the Mullen Scales of Early Learning (MSEL; Mullen, 1995), also a widely used and standardized assessment that yields a “Learning Composite” score. Due to the very low score of many of the children (below 1 percentile) a standardized DQ score for the MSEL scales was calculated for each scale by dividing children’s mental age by their chronological age and multiplying the result by 100. Children’s overall DQ score was calculated by averaging the four DQ scales.

Parent–child interaction

Parent–child interactions were assessed from two 7-min observations with each parent: Play with a set of age appropriate toys and play without toys. The parent and child were instructed to “play together as they usually do at home.” Order of parents and order of type of play were counterbalanced. Parent–child interactions were assessed using the EAS (4th edition; Biringen et al., 2014; Biringen et al., 2022). The EAS consists of six—1 (very low) to 7 (very high)—scales, four focusing on the parent and two focusing on the child. Each of the scales has several sub-scales which assist in assigning the overall scale score. The Parental Hostility scale did not yield meaningful variability and was therefore not used. The scales used were:

Parental sensitivity: The parent’s attunement and adequate responsiveness to the child’s signals, while expressing warmth and emotional connectedness to the child, and emphasizing the appropriateness and authenticity of the parent’s affect. The scale is based on the following sub-scales: balanced and genuine affect; clarity of perceptions and appropriate responsiveness; awareness of timing; creativity and flexibility in the interaction; acceptance; appropriate amount of interaction; and resolution of conflict situations.

Parental structuring: The degree to which the parent appropriately facilitates, stimulates, and helps organize the child’s play by providing the appropriate scaffolding and setting limits without compromising the child’s autonomy. The scale is based on the following sub-scales: use of appropriate guidance and suggestions; success of structuring attempts; amount of structuring; appropriate limit setting; appropriate regulation in case of pressure from the child; balance between verbal and nonverbal structuring; and assumption of a clear adult role.

Parental non-intrusiveness: The parent’s support of the child’s activities and awareness of the best time to enter the interaction without being too challenging or directive. The scale is based on the following sub-scales: following the child’s lead; the use of non-interruptive “ports of entry” into an interaction; limited use of commands and directives; communicative use of adult talking, appropriate use of didactic teaching; avoidance of physical and verbal interferences; and absence of hints in child’s behavior of the adult being intrusive.

Child responsiveness: The child shows clear signs of pleasure in the interaction with the parent and willingness to engage with the parent following the parent’s suggestions. The scale is based on the following sub-scales: positivity of affect and regulation of emotions and behavior; appropriate responsiveness to adult initiations; age-appropriate exploration and autonomy; positive physical contact and positioning; lack of role reversal/over responsiveness; lack of active avoidance; and lack of excessive task-oriented attitude.

Child involvement: The child’s engagement and involvement of the parent in the activity, while maintaining a balance between autonomy and invitations of the parent into play. The scale is based on the following sub-scales: Use of brief, one-time, simple initiatives; the use of elaborative initiatives that lead to an extended period of engagement; appropriate use of the parent for playful and emotional exchanges; lack of involvement through negative behavior; modulated use of eye contact and looking for engagement with the adult; positive body positioning; and the use of verbal involvement.

The interactions were coded by two coders trained by the second author who achieved reliability with the EAS developer (Biringen) in a training workshop. The coders were unaware of the research hypotheses (including, specifically, the expectation that play without toys will yield higher EA scores than play with toys) and not exposed to any other information about the participants. To ensure the independence of the coding of mother–child and father–child interactions, children’s interactions with their mothers and fathers were coded by a different coder. Each coder coded half of the interactions with the mothers and half of the interactions with the fathers so as not to confound coder with the parent being father or mother. The same coder coded the social and free play episodes of each parent–child dyad. Inter-rater reliability was established between the coders and the second author based on 25% of the sample and was good: sensitivity: ICC = .81; structuring: ICC = .72; intrusiveness: ICC = .89; child responsiveness: ICC = .86; child involvement: ICC = .87.

Results

Associations between the EA scales and demographic, child, and parent variables

All variables were normally distributed (i.e. skewness < 1.5, kurtosis < 2; Curran et al., 1996). The associations between the study variables (EA scales in the play episodes with or without toys with mothers and fathers) and parent age, years of education, years married, number of children in the family, and birth order were examined first and none were significant. The associations between the EA scores and children’s IQ were examined next and are presented in Table 1. Children’s IQ was associated with both of the children’s EA scale scores and with most parental EA scale scores so that higher IQ scores were associated with higher EA scores. All of the correlations reported were inspected to detect non-linear associations. None were revealed.

Table 1.

Correlations between study variables and children’s IQ.

EA scale	Play without toys	Play with toys
Maternal Sensitivity	.14	.22
Maternal Structuring	.24*	.35**
Maternal Non-Intrusiveness	.30*	.26*
Child Responsiveness w/mother	.43***	.62***
Child Involvement of mother	.46***	.62***
Paternal Sensitivity	.18	.23*
Paternal Structuring	.26*	.33**
Paternal Non-Intrusiveness	.29**	.21
Child Responsiveness w/father	.35***	.52***
Child Involvement of father	.48***	.56***

p < .05, **p < .01, ***p < .001 (two-tailed).

Comparison of the EA scales between play with or without toys with mothers and fathers

Five 2 × 2 analyses of covariance (ANCOVAs), one for each of the EA scales, were conducted, with the two repeated measures independent variables being type of play (with or without toys) and gender of parent (mother or father), IQ entered as a control variable, and the dependent variables in each ANCOVA being one of the EA scale scores (see Table 2). A main effect of type of play was significant for all of the EA scales with all of the EA scores for play without toys higher than the play with toys scores. No significant interactions between type of play and parent were found.

Table 2.

Comparison of EA scales in play with versus without toys with mothers and fathers—main effects of ANCOVAs with IQ controlled.

	Mother–child interaction				Father–child interaction
	Play w/o toys		Play with toys		Play w/o toys		Play with toys		With vs w/o toys		Mother vs Father
	M	SD	M	SD	M	SD	M	SD	F	η ²	F	η ²
Parent Scales
Sensitivity	5.15	1.17	4.75	.93	5.30	1.06	4.81	.99	17.93***	.19	< 1	.00
Structuring	5.05	1.21	4.66	.94	5.05	1.12	4.65	1.05	13.71***	.15	< 1	.00
Non-Intrusiveness	4.98	1.30	4.64	1.29	5.27	1.10	4.70	1.25	8.21**	.09	< 1	.01
Child Scales
Responsiveness	4.59	1.19	4.32	1.16	4.93	1.09	4.46	1.10	18.43***	.19	4.10*	.05
Involvement	4.08	1.34	3.78	1.24	4.28	1.28	3.71	1.25	19.53***	.20	< 1	.00

df = 1, 76 in all ANCOVAs; η² = partial eta squared.

p < .05, **p < .01, ***p < .001.

Although no interactions between type of play and parent were found, we nonetheless examined directly whether the main effects of type of play applied to both mothers and fathers. This was done with ANCOVAs identical to the ones described above except conducted separately for mothers and fathers (see Table 3). The results revealed that for all of the EA scales the main effect of type of play was also significant when it was examined separately for mothers and fathers (except for non-intrusiveness for mothers which was marginally significant).

Table 3.

ANCOVAs with IQ controlled comparing the EA scales in play with versus without toys separately for mothers and fathers.

EA scales	F(1, 76) with vs w/o toys
EA scales	Mother		Father
	F	η ²	F	η ²
Parent EA Scales
Sensitivity	5.46*	.07	14.93***	.16
Structuring	6.01*	.07	7.30**	.08
Non-Intrusiveness	2.88+	.04	7.83**	.09
Child EA Scales
Responsiveness	6.58*	.08	12.78***	.14
Involvement	6.56*	.08	13.27***	.14

EA = Emotional Availability. η² = partial eta squared.

p < .10, *p < .05, **p < .01, ***p < .001.

Associations between the EA scales across play with or without toys and across parents

The correlations between the EA scale scores across the play with versus without toys episodes are presented in Table 4. All of the correlations were significant, so that higher scores in the play with toys episode were associated with higher scores in the play without toys episode. Because child IQ was associated with many EA scales and could account for these correlations, the correlations were examined again with IQ controlled. All of the correlations remained significant.

Table 4.

Correlations of EA scales between play with versus without toys.

	Mothers		Fathers
	r	Partial r w/ IQ	r	Partial r w/ IQ
Parental Sensitivity	.28*	.26*	.45***	.43***
Parental Structuring	.40***	.34**	.41***	.35***
Parental Non-Intrusiveness	.43***	.38***	.50***	.47***
Child Responsiveness	.46***	.47***	.41***	.28**
Child Involvement	.66***	.53***	.55***	.39***

p < .05, **p < .01, ***p < .001 (two-tailed).

The correlations between the EA scale scores across parents (i.e. mother and father) are presented in Table 5. Children’s Responsiveness and Involvement scale scores were significantly correlated across parents so that higher scores with one parent was associated with higher scores with the other parent. Two of the three parental EA scale scores (Structuring and Non-intrusiveness) were also significantly correlated across parents but only in the play episode with toys. In the play episodes without toys, no associations between the parental scale scores (i.e. maternal and paternal) were found. Controlling for IQ reduced the magnitude of the correlations, in some cases rendering them non-significant (Parental structuring, parental non-intrusiveness, and child responsiveness in play with toys, and child responsiveness in play without toys; Table 5).

Table 5.

Correlations of EA scales between interactions with mothers and fathers.

	Play with toys		Play without toys
	r	Partial r w/IQ	r	Partial r w/IQ
Parental Sensitivity	.12	.07	.04	.02
Parental Structuring	.22*	.11	.07	.00
Parental Non-Intrusiveness	.25*	.20+	.05	.02
Child Responsiveness	.47***	.21+	.28*	.15
Child Involvement	.55***	.31**	.46***	.30**

p < .10, *p < .05, **p < .01, ***p < .001 (two-tailed).

Discussion

The study had two main goals: First, to compare the emotional availability of parents’ interactions with their autistic preschoolers in interactions with versus without toys, and second, to compare mother– and father–child interactions. Regarding the first goal, the hypothesis was that parent–child interactions without toys would be more positive, as evident in more sensitive, optimally structuring, and less intrusive parental behavior, and more responsive and involving child behavior, than parent–child interactions with toys. This hypothesis received strong support: Scores on all five of the EA scales in the observations of play without toys were higher than the scores in the observations of play with toys. This was true for both mother– and father–child interactions. Importantly, order effects (such as in Dolev et al., 2009) were eliminated by counterbalancing the order of the play episodes with versus without toys, and children’s IQ was controlled. This adds additional strength to the findings.

Our thoughts regarding why emotional availability was higher during play without toys follow Dolev et al. (2009). In many instances the toys seemed to distract the children, and they focused on the toys and were less available to interact with their parents. Parents often encouraged the children to play with the toys, directing them more than following their lead. Also, children often played with the toys in non-conventional ways that are characteristic of some autistic children, and parents often responded by trying to lead children to the “expected” play, resulting in less harmonious interactions. Conversely, when no toys were available the children were not distracted by them and the parents could attend more fully to the children’s signals. Parents and children often engaged in sensorimotor play (hugging, tickling, playing finger games, “catch,” singing, etc.) which was usually very enjoyable and elicited positive affect, responsiveness from the children, and requests for “more.” Other dyads engaged in developmentally more advanced play, showing complex and creative play. For example, one child and his father engaged throughout the entire observation in an imaginary “Tennis game” in which they jointly developed rules and elaborated new elements. Such games provided the conditions for more attuned and flowing interactions which were reflected in higher EA scores.

Of importance, alongside the mean-level differences favoring play without toys over play with toys, there was also stability across the play contexts. Thus, while the context of play without toys appeared to have a positive overall effect on the interactions relative to play with toys, the relative ordering of the dyads remained similar, as indicated by the significant correlations between the scales in each of the two play contexts (Table 4). In addition, even when children’s IQ, which was associated with most of the EA scales, was statistically covaried, these correlations remained significant (Table 4). Thus, the associations of the EA scale scores across play contexts seemed to reflect stability in patterns of parent–child interactions across contexts. An important caveat is that the same coder coded both play with and without toys of a particular parent–child dyad, and this could contribute to the associations between the two episodes.

The second goal of the study was to compare mother–child and father–child interactions. As reviewed earlier, most of the previous studies found no differences, and the findings from the current study were the same, with one exception: Children were more responsive to their fathers than to their mothers in the play episode without toys. Importantly, in contrast to the previous studies reviewed above which did not report whether different coders coded mother– and father–child interactions, thus introducing possible coder effects, this was not the case here because different coders coded each child’s interactions with their mother and father. An additional methodological strength was the relatively large sample size, double the size of the samples in the two largest studies conducted to date (Bentenuto et al., 2020; Hirschler-Guttenberg et al., 2015), increasing the confidence in the lack of differences found between mother– and father–child interactions.

Importantly, even though the fathers in this study worked more hours than did mothers, shared caregiving responsibilities with mothers in only 48% of the families, and the children were much less likely to be cared for by them (8.1%) than their mothers (80%) after preschool hours, their potential for establishing emotionally available interactions with their children was not lower than that of mothers. This is not to suggest that more should not be done to support fathers’ involvement with their autistic children, particularly for those fathers who can and wish to do so. Rather, this is just to highlight that fathers’ involvement with their children in this sample, although less than mothers,’ was sufficient to provide the basis for interactions that were as emotionally available as the mothers.’ It is clear, therefore, that not including fathers both in research on autistic children and interventions with this population may mean missing an important contributor to children’s social experiences (Flippin & Crais, 2011; Perzolli et al., 2021; Rankin et al., 2019). An important caveat to the finding of no differences between mother– and father–child interactions is that the study involved families in which both parents were willing and able to devote time to participate in the study. The findings may not be generalizable to other family situations, such as those in which fathers have little opportunity or support to be involved with their children or those in which mothers are single parents.

Alongside the lack of mean level differences between the EA scale scores of mother–child and father–child interactions, there were also significant associations between mother–child and father–child EA scale scores (Table 5). In particular, children’s responsiveness with their mothers and fathers was significantly correlated as was their involvement, and this was true both in observations of play with and without toys. This was expected because both are ratings of the same child, albeit with different partners. Turning to the parental scales: The maternal and paternal EA scales were unrelated to one another in the play episode without toys, but maternal and paternal Structuring and Non-intrusiveness were modestly associated in the episode of play with toys. Controlling for IQ reduced some of the correlations, in some cases rendering them non-significant. Thus, some of the associations between child-mother and child-father interactions were accounted for by children’s IQ which appeared to impact not only the children’s behavior but also their parents.’

Further support for the effect of children’s IQ on parents is evident in the associations between the parental EA scales and children’s IQ (Table 1). Lower IQ was associated with less optimal maternal and paternal structuring and more maternal and paternal intrusiveness in both play with and without toys (except for paternal intrusiveness in play with toys). It is possible that, at least in some instances, high intrusiveness scores and low structuring, which involve highly didactic and directive play with many suggestions, could actually represent parental attempts to guide and teach the children, particularly those with low IQs. Consistent with this, prior studies have shown that parents of children with an ASD diagnosis show more directive play interactions, involving more high intensity behaviors, non-verbal prompts, and high physical proximity, compared with children without an ASD diagnosis (Doussard-Roosevelt et al., 2003, Lemanek et al., 1993, Wan et al., 2012). These findings suggest that interventions with parents of autistic children, particularly those with low cognitive capacities, should work with parents to understand the conditions under which children need more structuring and direction, and those under which children can benefit from less structuring and direction.

Against the associations between IQ and parental structuring and intrusiveness, the lack of association between children’s IQ and mothers’ sensitivity in both play with and without toys and the lack of association between children’s IQ and fathers’ sensitivity (albeit only in the context of play without toys) is noteworthy. Sensitivity, the accurate reading of the child’s signals, appropriate responses, and positive and warm emotional climate, is a central EA scale with roots in Ainsworth’s seminal studies of the antecedents of secure attachment (Ainsworth et al., 1978). Sensitivity has been repeatedly shown to be associated with secure attachment (Madigan et al., 2024) as well as host of additional socioemotional and cognitive outcomes (e.g. Baumwell et al., 1997). The lack of association between this scale and children’s IQ, particularly for mothers, means that the sensitivity of mothers, and to a certain extent fathers, is independent of the degree of children’s cognitive delay. This highlights the capacity of mothers and fathers to adjust their behavior to the specific and unique behaviors and signals of their children in a way that promotes sensitive interactions, i.e. interactions that are harmonious and well-rounded, even with children whose cognitive functioning is relatively low. Such interactions are important for the development of all children, neurotypical or autistic. The lack of association between maternal and paternal sensitivity in both the play episodes (with and without toys) further emphasizes that the level of parental sensitivity is not simply a reaction to intrinsic child factors such as children’s cognitive functioning. If it were, it would lead to associations between fathers and mothers because they are responding to the same child. Rather, it seems that parental sensitivity reflects the specific and unique characteristic of each parent when interacting with a specific child. This points to the possibility and importance of improving parental sensitivity, as indeed is done in several treatment approaches (Green et al., 2010; Poslawsky et al., 2015; Rogers et al., 2012).

In closing, several limitations are of note. First, because only boys were studied the findings cannot be generalized to girls. The issues mentioned earlier about the possible different presentation of autism in girls and boys (Hull et al., 2020) and the phenomenon of “camouflaging” (Tubío-Fungueiriño et al., 2021) caution even more against assuming that the findings would be the same for girls. Also, while the inclusion of fathers was an important contribution, the fathers in this study only interacted with children of the same gender (i.e. boys). Thus, the findings can be related to gender-matching between parent and child. This is another reason why a similar study with fathers, but also including girls, is needed. An additional limitation involves the finding of no differences between mother–child and father–child interactions. It is important to qualify this null finding: It does not necessarily mean that fathers and mothers behaved identically. Because the EA scales are global and capture the overall quality of the interaction, different parental behaviors can receive similar scores. For example, one parent can be more emotionally expressive and warm than the other, but as long that they are similarly accurate in reading the child’s signals and responding to them appropriately they can receive the same sensitivity score.

Clinical wisdom and experience in work with autistic children (Carter et al., 2011; Greenspan & Wieder, 2006) and one previous study (Dolev et al., 2009) have suggested that parents’ play with their autistic preschoolers without toys promotes more harmonious interactions and fosters children’s engagement and responsiveness compared to parent–child play with toys. This study provides empirical support for this clinical hypothesis, and shows that it is true for both mother– and father–child interactions. This does not imply, of course, that parents and autistic children should be discouraged to play with toys. Rather, the findings highlight play without toys as providing optimal conditions for sensitive interactions, and suggest to both researchers and clinicians that combining opportunities for such play together with play with toys and other forms of play is important. The study also points to father–child interactions being equally emotionally available to mother–child interactions. Both findings broaden our understanding of the early social experiences of autistic children with their parents.

Footnotes

Acknowledgements

The authors thank Drs. Yael Maccabi, Lior Hamburger, and Michal Slonim for their help in data collection and Sapir Bar and Shiran Amoosi for their help in coding the observations. The authors also extend their gratitude to the participating families for their time and effort and for sharing their experience with us.

Author contributions

David Oppenheim: Conceptualization; Data curation; Funding acquisition; Methodology; Supervision; Writing—original draft; Writing—review & editing.

Michal Mottes-Peleg: Conceptualization; Data curation; Investigation; Methodology; Project administration; Writing—review & editing.

Smadar Dolev: Conceptualization; Investigation; Writing—review & editing.

Nurit Yirmiya: Funding acquisition; Investigation; Methodology; Resources; Supervision; Writing—review & editing.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by an Israel Science Foundation grant 155/15 and by the Harry and Sadie Lasky Foundation.

Ethical approval and informed consent

This study received ethical approval from the University of Haifa IRB (approval #400/16). All parents provided written informed consent prior to participating.

ORCID iD

David Oppenheim

Data availability statement

The data sets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

Ainsworth

M. D. S.

Blehar

M. C.

Waters

Wall

(1978). Patterns of attachment: A psychological study of the strange situation. Erlbaum.

Akin-Bulbul

Ozdemir

(2024). Evaluation of the Social Attention Hypothesis: Do Children with Autism Prefer to See Objects Rather than People? Journal of Autism and Developmental Disorders, 1–16. https://doi.org/10.1007/s10803-024-06596-9

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425596.

Bakermans-Kranenburg

M. J.

Lotz

Alyousefi -van Dijk

van IJzendoorn

(2019). Birth of a father: Fathering in the first 1,000 days. Child Development Perspectives, 13(4), 247–253.

Baumwell

Tamis-LeMonda

C. S.

Bornstein

M. H.

(1997). Maternal verbal sensitivity and child language comprehension. Infant Behavior and Development, 20(2), 247–258.

Bentenuto

Perzolli

de Falco

Venuti

(2020). The emotional availability in mother-child and father-child interactions in families with children with Autism Spectrum Disorder. Research in Autism Spectrum Disorders, 75, 101569. https://doi.org/10.1016/j.rasd.2020.101569

Biringen

Derscheid

Vliegen

Closson

Easterbrooks

M. A.

(2014). Emotional availability (EA): Theoretical background, empirical research using the EA Scales, and clinical applications. Developmental Review, 34(2), 114–167.

Biringen

Sandoval

Flykt

Joslin

Lincoln

(2022). Emotional availability: Clinical populations and clinical applications. Human Development, 66(3), 192–215.

Bornstein

M. H.

Gini

Putnick

D. L.

Haynes

O. M.

Painter

K. M.

Suwalsky

A. J. T.

(2006). Short-term reliability and continuity of emotional availability in mother–child dyads across contexts of observation. Infancy, 10(1), 1–16.

10.

Brito

Ryan

Barr

(2014). Methods for assessing parent-child interactions in large-scale studies. Handbook of Research Method on Early Childhood Education, 2, 147–189.

11.

Cabrera

N. J.

Volling

B. L.

Barr

(2018). Fathers are parents, too! Widening the lens on parenting for children’s development. Child Development Perspectives, 12(3), 152–157.

12.

Carter

A. S.

Messinger

D. S.

Stone

W. L.

Celimli

Nahmias

A. S.

Yoder

(2011). A randomized controlled trial of Hanen’s “More Than Words” in toddlers with early autism symptoms. Journal of Child Psychology and Psychiatry, 52(7), 741–752.

13.

Chawarska

Shic

Chen

(2016). Multilevel differences in spontaneous social attention in toddlers with Autism Spectrum Disorder. Child Development, 87(2), 543–557.

14.

Curran

P. J.

West

S. G.

Finch

J. F.

(1996). The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1(1), 16.

15.

Davidovitch

Hemo

Manning-Courtney

Fombonne

(2013). Prevalence and incidence of Autism Spectrum Disorder in an Israeli population. Journal of Autism and Developmental Disorders, 43, 785–793.

16.

DeVeney

Cress

C. J.

Lambert

(2016). Parental directiveness and responsivity toward young children with complex communication needs. International Journal of Speech-Language Pathology, 18(1), 53–64.

17.

Dolev

Oppenheim

Koren-Karie

Yirmiya

(2009). Emotional availability in mother-child interaction: The case of children with Autism Spectrum Disorders. Parenting: Science and Practice, 9(3–4), 183–197.

18.

Doussard–Roosevelt

J. A.

Joe

C. M.

Bazhenova

O. V.

Porges

S. W.

(2003). Mother–child interaction in autistic and nonautistic children: Characteristics of maternal approach behaviors and child social responses. Development and Psychopathology, 15(2), 277–295.

19.

Elder

J. H.

Valcante

Groce

Yarandi

Carlton

M. S.

(2002). Social interactions of children with autism in father-child and mother-child play sessions. Pediatric Nursing, 28(6), 573.

20.

Flippin

Crais

E. R.

(2011). The need for more effective father involvement in early autism intervention: A systematic review and recommendations. Journal of Early Intervention, 33(1), 24–50.

21.

Flippin

Watson

L. R.

(2015). Fathers’ and mothers’ verbal responsiveness and the language skills of young children with Autism Spectrum Disorder. American Journal of Speech-Language Pathology, 24(3), 400–410.

22.

Frosch

C. A.

Fagan

M. A.

Lopez

M. A.

Middlemiss

Chang

Hane

A. A.

Welch

M. G.

(2019). Validation study showed that ratings on the Welch Emotional Connection Screen at infant age six months are associated with child behavioural problems at age three years. Acta Paediatrica, 108(5), 889–895.

23.

Gotham

Risi

Pickles

Lord

(2007). The autism diagnostic observation schedule: Revised algorithms for improved diagnostic validity. Journal of Autism and Developmental Disorders, 37, 613–627.

24.

Green

Charman

McConachie

Aldred

Slonims

Howlin

Le Couteur

Leadbitter

Hudry

Byford

Barrett

Temple

Macdonald

Pickles

, & PACT Consortium. (2010). Parent-mediated communication-focused treatment in children with autism (PACT): A randomised controlled trial. Lancet, 375, 2152–2160.

25.

Greenspan

S. I.

Wieder

(2006). Infant and early childhood mental health: A comprehensive development approach to assessment and intervention. American Psychiatric Publishing, Inc.

26.

Hirschler-Guttenberg

Golan

Ostfeld-Etzion

Feldman

(2015). Mothering, fathering, and the regulation of negative and positive emotions in high-functioning preschoolers with Autism Spectrum Disorder. Journal of Child Psychology and Psychiatry, 56(5), 530–539.

27.

Hull

Petrides

K. V.

Mandy

(2020). The female autism phenotype and camouflaging: A narrative review. Review Journal of Autism and Developmental Disorders, 7, 306–317.

28.

Lemanek

K. L.

Stone

W. L.

Fishel

P. T.

(1993). Parent-child interactions in handicapped preschoolers: The relation between parent behaviors and compliance. Journal of Clinical Child Psychology, 22(1), 68–77.

29.

Lord

Risi

Lambrecht

Cook

E. H.

Jr. Leventhal

B. L.

DiLavore

P. C.

Pickles

Rutter

(2000). The autism diagnostic observation schedule-generic: A standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders, 30, 205–223.

30.

Maas

A. J. B.

Vreeswijk

C. M.

van Bakel

H. J.

(2013). Effect of situation on mother–infant interaction. Infant Behavior and Development, 36(1), 42–49.

31.

Madigan

Deneault

A. A.

Duschinsky

Bakermans-Kranenburg

M. J.

Schuengel

van IJzendoorn

M. H.

Fearon

R. M. P.

Eirich

Verhage

M. L.

(2024). Maternal and paternal sensitivity: Key determinants of child attachment security examined through meta-analysis. Psychological Bulletin, 150(7), 839–872.

32.

Maenner

M. J.

Warren

Williams

A. R.

Amoakohene

Bakian

A. V.

Bilder

D. A.

Durkin

M. S.

Fitzgerald

R. T.

Furnier

S. M.

Hughes

M. M.

Ladd-Acosta

C. M.

McArthur

Pas

E. T.

Salinas

Vehorn

Williams

Esler

Grzybowski

. . .Shaw

K. A.

(2020). Prevalence and characteristics of Autism Spectrum Disorder among children aged 8 years—Autism and developmental disabilities monitoring network, 11 sites, United States. Morbidity and Mortality Weekly Report. Surveillance Summaries, 72(2), 1–14. https://dx.doi.org/10.15585/mmwr.ss7202a1

33.

Mullen

E. M.

(1995). Mullen scales of early learning. American Guidance Services.

34.

Oppenheim

Koren-Karie

Hamburger

Maccabi

Slonim

Yirmiya

(2023). Parental insightfulness is associated with mother–father–child interactions among families of preschoolers with an Autism Spectrum Disorder diagnosis. Journal of Child Psychology and Psychiatry, 64(9), 1359–1368.

35.

Perzolli

Bentenuto

Bertamini

de Falco

Venuti

(2021). Father–child interactions in preschool children with ASD: A systematic review. Brain Sciences, 11(9), 1202.

36.

Poslawsky

I. E.

Naber

F. B.

Bakermans-Kranenburg

M. J.

Van Daalen

Van Engeland

Van Ijzendoorn

M. H.

(2015). Video-feedback Intervention to promote Positive Parenting adapted to Autism (VIPP-AUTI): A randomized controlled trial. Autism, 19(5), 588–603.

37.

Rankin

J. A.

Paisley

C. A.

Tomeny

T. S.

Eldred

S. W.

(2019). Fathers of youth with Autism Spectrum Disorder: A systematic review of the impact of fathers’ involvement on youth, families, and intervention. Clinical Child and Family Psychology Review, 22, 458–477.

38.

Rogers

S. J.

Estes

Lord

Vismara

Winter

Fitzpatrick

Guo

Dawson

(2012). Effects of a brief Early Start Denver Model (ESDM)–based parent intervention on toddlers at risk for Autism Spectrum Disorders: A randomized controlled trial. Journal of the American Academy of Child & Adolescent Psychiatry, 51(10), 1052–1065.

39.

Rutter

Bailey

Lord

(2003). The social communication questionnaire. Western Psychological Services.

40.

Sjolseth

S. R.

Frosch

C. A.

Owen

M. T.

Redig

S. L.

(2024). Do toys get in the way? The duration of shared emotional experiences is longer when mothers engage their infants without toys. Infant Mental Health Journal, 45(1), 3–10.

41.

Tubío-Fungueiriño

Cruz

Sampaio

Carracedo

Fernández-Prieto

(2021). Social camouflaging in females with Autism Spectrum Disorder: A systematic review. Journal of Autism and Developmental Disorders, 51, 2190–2199.

42.

Vivanti

Fanning

P. A.

Hocking

D. R.

Sievers

Dissanayake

(2017). Social attention, joint attention and sustained attention in Autism Spectrum Disorder and Williams syndrome: Convergences and divergences. Journal of Autism and Developmental Disorders, 47, 1866–1877.

43.

Wan

M. W.

Green

Elsabbagh

Johnson

Charman

Plummer

, & BASIS Team. (2012). Parent–infant interaction in infant siblings at risk of autism. Research in Developmental Disabilities, 33(3), 924–932.

44.

Wechsler

(2002). WPPSI-III: Technical and interpretative manual. The Psychological Corporation.

45.

Wiggins

L. D.

Reynolds

Rice

C. E.

Moody

E. J.

Bernal

Blaskey

Levy

S. E.

(2015). Using standardized diagnostic instruments to classify children with autism in the study to explore early development. Journal of Autism and Developmental Disorders, 45(5), 1271–1280.