Parent Behaviors in Relation to Characteristics of Young Children With Autism Spectrum Disorders

Emotional self-regulation (ESR) challenges are well documented in the developmental profiles of children with Autism Spectrum Disorder (ASD); however, less is known about the development of ESR and the role of parents in ESR development for this population. This study explored the associations between diagnostically significant characteristics of children with ASD (i.e., social communication profile and sensory processing abilities) and parent behaviors associated with development of a child’s ESR. Participants were 37 children previously diagnosed with ASD and their parents recruited throughout Southern New England. The children’s ages ranged between 3048 months. The study was a cross sectional design involving observations of parentchild dyads in their home environment during naturalistic routines: free play, social communication assessment, and snack. Observations were video recorded for subsequent coding using combined event/ time sampling procedures. Parent behaviors included physical engaging and helping, language-based engaging and helping, redirection/distraction, vocal comfort, physical comfort, language-based comfort, emotional following, and active ignoring. Children’s social communication abilities were assessed using the Communication Symbolic Behavior ScalesDevelopmental Profile (Wetherby & Prizant, 2002) and their sensory processing abilities were assessed using the Sensory Processing Measure – Preschool Home form (Ecker & Parham, 2010). A demographic questionnaire was also completed by the parent.

Results indicate that parents engaged in all of the behaviors hypothesized to be associated with ESR development during the observations but to varying degrees.
Parents most frequently used physical engaging and helping, and language engaging and helping behaviors, while behaviors associated with comfort were infrequent and active ignoring was rare. Children's social communicative abilities were associated with parent engagement such that children with high levels of social communication had parents who used more parent language engaging and helping and less physical engaging and helping, redirection/distraction, and physical comfort. Overall, no parent behaviors were associated with child sensory processing abilities. These findings were fairly consistent across activities studied in the research protocol although some variations were noted.
Future research should explore additional aspects of the interactions by examining the transactions between child emotional state and parent behavior response. Delineation of specific parent behaviors could also serve to further the understanding of the particular qualities of parent behaviors that are most supportive of child ESR development. Understanding the interactive processes between parents and children with ASD has implications for the development of targeted parent-based interventions that increase child ESR capabilities and in turn decrease the secondary burdens and long term difficulties posed by ESR challenges for this population.    ( Konstantareas & Stewart, 2006;Nader-Grosbois & Mazzone, 2014;Samson et al., 2014). ESR refers to a developmental capacity which enables an individual's ability to monitor, evaluate, and modify their emotional state and arousal level to maintain engagement and accomplish objectives (Cole, Martin, & Dennis, 2004;Fox, 1994;Gulsrud, Jahromi, & Kasari, 2010;Kopp, 1982). These skills are essential for participating in daily activities, forming relationships, and engaging in positive social interactions (National Research Council and Institute of Medicine, 2000). In typical development, ESR abilities become increasingly complex as children's cognitive, language, and attentional skills develop, and as children experience responsive interactions with parents (Hubley & Trevarthen, 1979;Kochanska, Murray, & Harlan, 2000;Kopp, 1982;National Research Council and Institute of Medicine, 2000;Tronick, 2002). Characteristics consistent with challenges in ESR, such as difficulties managing emotions, inhibiting reactions, delaying gratification, and tolerating transitions, are frequently associated with ASD (American Psychiatric Association, 2013; Baron, Groden, Groden, & Lipsitt, 2006;DeGangi, 2000;National Research Council, 2001) Parental behavior has been associated with ESR development for typically developing children (Morales, Mundy, Crowson, Neal, & Delgado, 2005;Saarni, 1998;Sanders & Mazzucchelli, 2013). For example, parents who respond to their child's emotional dysregulation by validating their child's emotional experience, labeling their emotional expression, and/or remediating frustrating circumstances have children who utilize more sophisticated ESR (Eisenberg & Sulik, 2012;Spinrad, Eisenberg, & Gaertner, 2007;Tronick, 2002). Parents of typically developing children differentially engage in behaviors associated with supporting child ESR (e.g., helping, redirection, verbal comfort, and physical comfort) during interactions based upon their child's age and related cognitive abilities. For example, as typically developing toddlers age, parents' use of physical behaviors to support engagement has been found to decrease while their use of verbal behaviors increases. Likewise, parents' use of active strategies, such as hand over hand assistance, decreases as they begin to provide more time and opportunity for their children to problem solve challenging situations independently (Grolnick, Kurowski, McMenamy, Rivkin, & Bridges, 1998).
In addition to age, children's behavioral and developmental characteristics influence parental behaviors associated with supporting ESR. For example, parents of children who are described as having difficult temperaments, displaying frequent, intense emotional distress, have mothers who engage in more redirection of attention and provide more reassurance than parents of children who demonstrate less distress (Grolnick et al., 1998). Similar associations have been found among children diagnosed with ASD. In a study of children 21-36 months of age diagnosed with ASD, the association between externalizing problem behaviors, such as aggression, and parental regulatory strategies was examined. During play, parents of ASD children who demonstrated higher levels of externalizing behaviors, utilized more prompting and redirection, and physical comfort as compared to parents of ASD children who demonstrated low levels of externalizing behaviors . The associations between behavioral characteristics of children diagnosed with ASD such as social communication and sensory processing abilities and parent behaviors associated with supporting ESR have not been previously studied.

Statement of the Problem
The development of a child's ESR is influenced by the behaviors parents engage in during daily interactions. Children diagnosed with ASD have significant social communicative deficits and sensory processing differences which have the potential to impact parent behaviors. Therefore, exploring the relationship between ASD specific child behavioral characteristics and parents' engagement in behaviors associated with supporting ESR is an important area of inquiry.

Significance of the Study
To date, associations examining social communication and sensory processing abilities, and parent regulatory behaviors have not been explored. Child social communication and sensory processing abilities have been critically linked to child ESR capabilities and are likely to influence behaviors parents engage in during daily interaction (Norona & Baker, 2014;Sameroff & Fiese, 1990;Zimmerman, 2000).
Given the importance of the parental role in supporting the development of ESR (Kopp, 1989;Sameroff & Fiese, 1990;Tronick, 2002), understanding parental behaviors in relation to child characteristics is an important area of inquiry in the study of ASD. Additional knowledge linking child characteristics and parent behaviors that are theoretically supportive of the development of ESR could help to further our understanding of the ESR challenges seen in children diagnosed with ASD and factors influencing the development of these challenges. Likewise, findings illustrating the relationship between child characteristics and parent behaviors could help to inform future family-mediated interventions targeting the development of ESR for young children diagnosed with ASD.

Purpose of the Study
The purpose of the study was to examine the association between behavioral characteristics of young children diagnosed with ASD and their parents' engagement in behaviors which have been previously associated with the development of a child's ESR.

Research question 1.
Are the social communication abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR?

Research question 2.
Are the sensory processing abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR?

CHAPTER 2
Review of the Literature

Overview of Emotional Self-Regulation (ESR)
Emotional self-regulation (ESR) refers to a variety of developmental abilities and intentional behaviors which serve to help an individual shift physiological arousal level, modulate emotional state, and modify attentional focus, all in an effort to meet social expectations, maintain engagement, and accomplish objectives (Cole et al., 2004;Eisenberg & Spinrad, 2004;Fox, 1994;Grolnick, Bridges, & Connell, 1996;Kopp, 1982;Thompson, 1994). Emotional regulation is a developmental construct, which implies that a child's skills and capacities are influenced by the child's cognitive and physical development and maturation, as well as scaffolded through modeling and direct teaching by others (Thompson, 1994;Tronick, 2002;Zimmerman, 2000). As a child's ESR abilities increase in breadth and depth so does their ability to navigate challenges independently. (Grolnick, Kurowski, McMenamy, Rivkin, & Bridges, 1998;National Research Council, 2000).
The principle function of ESR is attainment and maintenance of emotional and arousal states that support engagement and learning (Fox, 1994). ESR enables individuals to use regulatory strategies to shift emotion or arousal states to meet the demands of social and physical environments. The match between internal physiological states and environmental demands is often referred to as a well-regulated state. An individual experiences emotion dysregulation when this adaptive shift of emotion and arousal level does not occur, resulting in an arousal level that is either too high or too low to engage in the social or physical environment (Eisenberg et al., 2003;Rothbart & Bates, 1998).

Development of ESR
ESR abilities develop rapidly in early childhood and continue to mature into adulthood (Eisenberg et al., 2010;Eisenberg & Sulik, 2012). It is understood that burgeoning ESR abilities are influenced by development and maturation, as well as interactions with the social and physical environment (Thompson, 1994;Tronick, 2002).
Children are born with a bio-behavioral drive for homeostasis and regulation (Fox, 1994). However, infants have very few strategies available to them to help regulate their arousal level, emotional state, and attention. In general, infants are born with reflexive abilities that serve a regulatory function. These include gaze aversion, a non-nutritive suck, and sleep. The utility of these strategies is limited in terms of supporting engagement; therefore, infants require frequent assistance with regulation from parents and caregivers (Fox, 1994). In an effort to support infants' regulation, parents typically respond to signals of dysregulation (e.g., crying and vocalizing) by employing respondent mutual regulatory strategies (e.g., physical help, redirection, comfort, etc.) (Tronick, 2002). These strategies provide infants opportunities to experience new regulatory means in supportive interactive relationships. Repeated interactions with parents scaffolding regulatory development paired with increasing developmental abilities (e.g., motor skills, communicative abilities, and social cognitive awareness) support a child's ability to learn, integrate, and utilize new, more sophisticated and socially conventional ESR strategies in his or her behavioral profile which can be employed to soothe, distract, self-comfort, delay gratification, and problem solve (Grolnick et al., 1996;Zimmerman, 2000). This cycle of interactive support, child development, and acquisition of new ESR strategies continues throughout childhood, eventually leading to a child's ability to initiate intentional requests for regulatory assistance and to continue refinement of socially appropriate ESR abilities based on feedback given by the child's social partners.
Children typically exhibit the ability to utilize their expanding ESR abilities for the function of intentional behavioral control starting at 9-12 months of age, to guide interactive responses with internalized behavioral expectations by 24 months, and to meet situational demands with considerable flexibility and expanded effortful control by 36 months (Eisenberg et al., 2010;C. Kopp, 1982;Sanders & Mazzucchelli, 2013).

ESR Among ASD
ASD is a neuro-developmental disorder characterized by impairments in social communication and the presence of restricted repetitive and stereotyped patterns of behavior, interests, and activities (American Psychiatric Association, 2013). In addition to these diagnostically significant behavioral patterns, it is widely accepted that individuals with ASD frequently exhibit additional associated challenges. For example, it is generally acknowledged that children with ASD display characteristics that are consistent with challenges in ESR . Difficulties regulating sleep-wake cycles, managing emotions, focusing attention, inhibiting reactions, delaying gratification, tolerating transitions, and seeking comfort in conventional ways during times of stress are all challenges frequently associated with the behavioral profile of individuals with ASD (American Psychiatric Association, 2013; Baron et al., 2006;DeGangi, 2000;National Research Council, 2001).
Likewise, individuals diagnosed with ASD have also been reported to demonstrate high rates of internalizing and externalizing behavioral disorders which are linked to deficits in ESR (Mazefsky, 2015;Richey et al., 2015;Swain, Scarpa, White, & Laugeson, 2015;Wilson, Berg, Zurawski, & King, 2013). Collectively these ESR related behavioral presentations are a primary reason of referral for treatment in ASD Samson, Hardan, Lee, Phillips, & Gross, 2015).
Several recent studies documented the existence of ESR deficits in a sample of children with ASD compared to age-matched typically developing controls and/or to other developmentally disabled populations. For example, Konstantareas and Stewart (2006) found evidence of impaired affect regulation and reduced numbers of effective ESR strategies in the behavioral profile of children with ASD when compared to those of typically developing controls. Glaser and Shaw (2011)  When ESR strategies were employed by the ASD sample they were characterized as physical and verbal venting, less goal directed, and less socially oriented when compared with typical controls. Additional studies provide further evidence for a maladaptive pattern of ineffective self-regulatory response and suggest that this ESR profile may be linked to high levels of co-morbid psychopathology diagnosed in individuals with ASD (Gerstein et al., 2011), increased negative emotional experiences (Samson et al., 2015), and internalizing and externalizing behavioral disorders .
Additional studies have examined ESR in ASD as a predictor of engagement and adaptive functioning. One longitudinal study looked at the self-regulation profiles of children with ASD as a predictor of peer and school engagement, finding that differences in ESR were associated with adaptive functioning. Study results suggest that students with ASD who had relatively greater ESR abilities demonstrated more prosocial peer engagement a year later as compared to those with lower ESR abilities (Jahromi et al., 2013). Similarly, Gray and colleagues (2014) reported that older individuals diagnosed with ASD who demonstrated greater emotional and behavioral difficulties associated with ESR challenges also demonstrated decreased ability to independently engage in activities of daily living and required more extensive support systems and services.
Given the increasing body of literature supporting the existence, prevalence, and implications of ESR challenges in individuals with ASD's behavioral profiles, there is mounting interest in understanding how these challenges develop and the factors influencing their expression.

The Development of ESR in ASD
The study of the development of ESR among individuals diagnosed with ASD is challenging since many of the behaviors associated with ESR are diagnostically related to ASD. In fact some have argued that the development of ASD is inextricably intertwined with altered development of ESR, and that the presence of early self-regulation difficulties in a child's behavioral profile may be an early indicator of a potential diagnosis of ASD (Gomez & Baird, 2005;Loveland, 2005;Whitman, 2004). Still others have suggested that the degree of ESR impairment displayed by a child with ASD may account for at least a portion of the heterogeneity in long term outcomes and adaptive functioning exhibited by children on the spectrum-with the most dysregulated children demonstrating the poorest outcomes (Jahromi et al., 2013).
To date there has not been a longitudinal study looking at the normative development of regulatory abilities in children with ASD. Research in this area is just beginning to emerge (Mazefsky, 2015). However, several areas of impairment including social communication deficits and sensory processing challenges have been directly associated with ESR differences and emotional dysregulation for children diagnosed on the spectrum (Samson et al., 2014). Challenges in these areas have the potential to influence parent child interactions and the transactional nature of ESR development Nader-Grosbois & Mazzone, 2014;Prizant et al., 2006b). Both of these developmental domains will be reviewed in relation to ESR and the current ASD literature.

Social Communication Abilities in ASD
Social communication abilities refer to a wide range of developmental skills.
Conceptually, this developmental domain can be divided into two primary capacities: 1) social abilities (e.g., communicative functions) and 2) communicative abilities (e.g., communicative means) (Prizant, Wetherby, Rubin, & Laurent, 2003). Both of these capacities have been critically linked to social competences and a child's ability to acquire more social conventional and effective regulatory strategies through a transactional process of engaging with parents (Bachevalier & Loveland, 2006;Morales et al., 2005;Vaughan Van Hecke et al., 2012). A core capacity related to a child's social abilities is joint attention. Joint attention allows children to see another person as separate from themselves, share attention, share emotions, express intentions, and develop an awareness of another person's attentional focus, knowledge and preferences (Prizant et al., 2003;Stern, 1985). Joint attention correlates with developmental capacities such as being able to respond consistently to the sound of their mother's voice, follow gaze, referentially look, socially reference, and understand the communicative perspective of another person. Deficits in each of these capacities are considered diagnostically significant for ASD and limit the individual's Typically developing children who engage more frequently in joint attention with parents demonstrate more sophisticated ESR as well as a greater ability to modify emotional reactions and regulatory strategies in response to feedback of others (Bachevalier & Loveland, 2006;Konstantareas & Stewart, 2006;Morales et al., 2005;Raver, 1996). Implications of these finding extend to ASD and suggest that individuals with ASD who demonstrate less joint attention exhibit greater ESR challenges  and greater emotional dysregulation (Samson et al., 2014) than children with ASD who demonstrate more joint attention.
Communicative development typically refers to the process of acquiring and utilizing sophisticated, conventional communication processes. In infancy this process is concerned with the acquisition of nonverbal communicative means that have shared meaning; while in early childhood it is largely focused on the development of symbolic communication (e.g., vocabulary and syntactic structures). Diagnostically, individuals with ASD exhibit quantitatively and qualitatively impaired receptive and expressive language (e.g., nonverbal and symbolic means)(American Psychiatric Association, 2013). Delays in receptive language impact a child's ability to understand emotions, respond to language or parental support, and to use languagebased ESR strategies to guide behavior (Kopp, 1982;Prizant et al., 2006b;Vygotsky, 1962). In addition, delayed expressive skills are associated with a child's limited ability to request assistance and communicate emotional states using words and nonverbal means (Wetherby, Watt, Morgan, & Shumway, 2007). Collectively, these expressive and reception communication deficits have been shown to be positively associated with higher levels of emotional dysregulation for children with ASD (Samson et al., 2014).

Sensory Processing Abilities Among ASD
Children with ASD demonstrate sensory processing deficits (American Psychiatric Association, 2013). These manifest in many ways including hyporeactivity (e.g., decreased sensitivity and response to environmental stimuli such as touch and sounds) and hyper-reactivity (e.g., increased sensitivity to sensory stimuli such as sounds and movement). Greater challenges responding adaptively to sensory stimuli and/or displaying unusual interests in sensory aspects of the environment are associated with greater emotional dysregulation in children with ASD (Samson et al., 2014). Sensory processing deficits are also associated with increased risk for the development of anxiety, depression, and maladaptive behaviors all of which have been linked to deficits in ESR (Brindle, Moulding, Bakker, & Nedeljkovic, 2015;Loveland, 2005;. Additional research highlights temperamental challenges which have been associated with sensory processing deficits for this population (Samson et al., 2014).
Children diagnosed with ASD often demonstrate temperaments that are characterized as difficult, and demonstrate behaviors consistent with ESR challenges such as irritability, poor impulse control, difficulty soothing, and unfocused attention (Clifford, Hudry, Elsabbagh, Charman, & Johnson, 2013;Jahromi et al., 2012;Konstantareas & Stewart, 2006;Loveland, 2005). These temperamental differences frequently translate clinically to large scale emotional reactions and relatively high rates of distress during normal play interactions with mothers .

Parent Behavior Associated with Facilitating Child ESR Among ASD
Parents play a critical role in the development of ESR for all children. The interdependent relationship between the young child and their social context is considered the foundation of emotional regulation development (Hubley & Trevarthen, 1979;Kopp, 1982;Sameroff & Fiese, 1990;Tronick, 2002). ESR is a transactional process that flourishes within environments that are supportive of socialcognitive and social-emotional learning (Sanders & Mazzucchelli, 2013;Zimmerman, 2000). For example, the quantity and quality of ESR behaviors utilized by typically developing children is shaped by parental emotional expressivity (Eisenberg et al., 2001(Eisenberg et al., , 2003 and parental use of supportive behaviors (e.g., redirection of attention, reassurance, physical comfort, etc.) (Grolnick et al., 1998).
Research suggests that parents use a variety of strategies during interactions with their children that are associated with the children's ESR development (Morales, Mundy, Crowson, Neal, & Delgado, 2005;Saarni, 1998;Sanders & Mazzucchelli, 2013). These include, but are not limited to, engaging, helping, redirection of attention, verbal comfort, physical comfort, and emotional following. While evidence suggests that each of these individual parent behaviors plays a role in the development of ESR, previous studies have grouped them into theoretical categories based on their intended function (i.e., engaging, comforting, or providing opportunity for child to independently employ ESR) and their symbolic qualities (i.e., physical or language based) (Grolnick et al., 1998;Gulsrud et al., 2010).
These theoretical categories conceptually reflect the responsive qualities of parent behavior but also developmental sophistication. For example, parents use of language strategies (e.g., verbal problem solving, reframing of emotions) is regarded as a higher order or more complex when compared to parent use of physical strategies such as hand over hand assistance while manipulating a toy or holding during distress (Zimmerman, 2000). With regard to the responsive quality or the function of the parent strategies, actively engaging behaviors are those that shift attention away from arousing situations, as well as goal directed behaviors such as helping and problem solving; comforting strategies are those that provide soothing and reassurance; while passive strategies focus on providing the child opportunity to self-regulate often referred to as active ignoring .
Several child characteristics have the potential to impact aspects of regulatory interactions between parents and young children diagnosed with ASD. To date studies have focused on the associations between child developmental age, externalizing problem behaviors,  and cognitive abilities (Hirschler-guttenberg, Feldman, Ostfeld-etzion, Laor, & Golan, 2015) and the behaviors employed by parents to support child ESR.
One study examining the relationship between a child diagnosed with ASD's cognitive abilities and his/her parent's differential employment of regulatory strategies found that fathers' use of physical comfort, holding and refocusing of attention during play was inversely associated with IQ during stress conditions . This finding is interesting to consider in the context of an additional longitudinal study that examined the transactional relationship between parenting and emotion regulation in children with and without cognitive developmental delays (Norona & Baker, 2014). Parents of children diagnosed with developmental delays exhibited less scaffolding of ESR (e.g., less activity demonstration, less praising, less refocusing, less sensitivity toward emotional state, and less emotional following) during interactions with their young children than parents of typically developing children. Finally, Gulsrud and colleagues (2010) reported that parents of cognitively delayed toddlers diagnosed with ASD engaged in high levels of physical prompting, assisting, and comforting during play with their child in contrast to more developmentally sophisticated behaviors (e.g., verbal problem solving, emotional labeling, etc.) frequently employed by parents of typically developing children at similar chronological ages.
These findings suggest that parents of children with ASD differentially employ behaviors associated with supporting ESR in response to their child's characteristics much like parents of typically developing children. For example, parents of typically developing toddlers use regulatory focused behaviors differentially based upon their child's age and cognitive abilities. One study found that parents of 32-month-old toddlers used fewer active engagement strategies (e.g., helping, redirection, and physical comfort) during times of stress than did the parents of younger children (Grolnick et al., 1998). The authors of this study also reported that parents of children between 18-32 months provided consistent opportunities for their children to selfregulate by engaging in more passive behaviors such as active ignoring (Grolnick et al., 1998) which was not seen with younger children.
In addition to age and cognitive abilities, other factors such as social communication delays and sensory processing differences may also influence parent engagement in behaviors when interacting with their young children diagnosed with ASD. In typical development, as children's age, cognitive and social communicative abilities increase parents use of language strategies for problem solving, maintaining engagement, emotional labeling, and cognitive reframing also increases (Grolnick et al., 1998;Saarni, 1998;Zimmerman, 2000). Parents use of these types of strategies has been related to child executive functioning and adaptive problem solving in longitudinal studies of ESR development (Spinrad et al., 2006). This reduction in scaffolding has been associated with qualitative differences in social interactions between parents and children. Current research suggests that parents of toddlers with ASD use more helping and physical behaviors and less verbal support strategies such as emotional expression or previewing events during stressful interactions. . A recent study comparing preschoolers diagnosed with ASD to typically developing preschoolers found that parents of children diagnosed with ASD used fewer complex strategies (e.g., language based strategies) and relied on more simple strategies such as physical comfort and physical engagement during fear and joy paradigms than parents of typically developing children . Likewise, clinic-based studies of the development of play and language have documented that parents of children with ASD's use of physical prompts and directive interactive styles is negatively correlated with their children's social communicative abilities (Kasari, Sigman, Mundy, & Yirmiya, 1988;Konstantareas, Zajdeman, Homatidis, & McCabe, 1988). Despite these related findings, the association between social communicative difficulties for children with ASD and parent behaviors associated with ESR development has not been examined directly.
The association between sensory processing differences exhibited by children with ASD (e.g., sensory processing differences) and parent regulatory strategies also remains unstudied. However, research suggests that parents of typically developing children who have difficult temperaments which are often associated with overreactivity to sensory stimuli, primarily utilize physical strategies to support their children's emotional regulation (Cole et al., 2013(Cole et al., , 2004Sallquist et al., 2009).

Present Research
The purpose of this study was to examine the associations between social communication and sensory processing abilities of children with ASD and parent behaviors associated with development of a child's ESR in the naturalistic setting of the child's home. The presence of ESR challenges in the developmental profiles of children with ASD has been well established; however, less is known about the role of parents in ESR development for this population, underscoring the need for further study in this area. Additionally, the majority of research on ESR in children with ASD has been completed in unfamiliar laboratory contexts Konstantareas & Stewart, 2006), so the examination of parent behaviors within natural environments is needed. This study was conducted in family homes in an effort to capture parent behaviors during naturalistic routines.
It was anticipated that parents of young children who exhibit greater social communication impairment and sensory processing challenges will engage in behaviors that are physical in nature (e.g., helping activate toys, providing hand over hand assistance, providing postural support, etc.), comfort oriented (e.g., hugging, vocal soothing, reassuring, etc.), and focused on active engagement (e.g., orienting to toy, redirecting attention from distraction, etc.) and use fewer behaviors that are language-based (e.g., verbal problem solving, language-based reassurance, etc.) and are associated with allowing the child time to independently engage in ESR (i.e., active ignoring).

Research Hypotheses
Research question 1. Are the social communication abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development the development of child ESR?

Hypothesis 1.
Higher levels of child social communication abilities will be associated with lower levels of parent physical engaging and helping, physical comfort, redirection/distraction, vocal comfort, and emotional following.

Hypothesis 2.
Higher levels of child social communicative abilities will be associated with higher levels of parent language-based engaging and helping, language-based comfort, and active ignoring.
diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR? Hypothesis 1. Higher levels of child sensory processing abilities will be associated with lower levels of parent physical engaging and helping, physical comfort, redirection/distraction, vocal comfort, and emotional following.

Hypothesis 2.
Higher levels of child sensory processing abilities will be associated with higher levels of parent language-based engaging and helping, language-based comfort, and active ignoring.

Participants
Participants were 37 primary caregiver-child dyads who were recruited throughout Southern New England. Children aged 30-48 months who had previously been diagnosed with ASD were included in the study with their primary caregiver.
Primary caregiver was defined as the family-identified parent who spends the most time caring for and interacting with the child throughout his or her normal daily routine. One primary caregiver identified as their child's legal guardian, all others identified as mothers or fathers. For the purposes of this study we refer to all primary caregivers as parents.
Children were required to have an ASD diagnosis validated by a positive Children were excluded if they had a history of seizures, blindness, deafness, physical developmental disabilities that significantly impaired mobility, and/or they were medicated for regulation-related challenges. Participants were required to speak English.
Power analysis for bivariate correlation coefficients (power set at .80, r= .44, α=.05) and regression analyses with three predictors (power set at .80, α=.05, r 2 =.261) indicated that a sample size of 35 would be adequate. During the recruitment process, 52 families agreed to be contacted by the researchers to discuss the study and 37 enrolled in and completed in the study.
The study was approved by the Institutional Review Board (IRB) at the University of Rhode Island in December 2014 (IRB #HU1415-082).

Demographics.
Parent participants completed a demographic questionnaire including information on family composition/size, parental age, education, occupation, race, and child sex. Additional questions included the type(s) and focus of educational and therapeutic interventions in which the child participates. Data from the questionnaire were utilized in a descriptive analysis of the sample and to identify potential covariates. The complete demographic questionnaire is included in Appendix A.

Child social communication abilities.
Children's social communication abilities were assessed using the

Communication and Symbolic Behavior Scales -Developmental Profile Behavior
Sample (CSBS-DP; . The CSBS-DP is a standardized measure designed to assess the social communicative behavior of children between the developmental ages of 6 months and 24 months . However, it has been used as a clinical and research tool for children who demonstrate The CSBS DP evaluates the presence of eight social behaviors (i.e., gaze shifts, shared positive affect, gaze/point following, behavior regulation, social interaction, joint attention, and conventional gestures), as well as the frequency of a child's social behaviors during six play-based activities (i.e., wind-up toys, balloons, bubbles, jar with preferred object enclosed, book sharing, and pretend play). The child's use of speech (e.g., different speech sounds, words, and word combinations) is also assessed during each of these six play-based activities. In addition, during the pretend play activity the child's symbolic abilities were assessed with respect to language comprehension and the sophistication of play (e.g., types of play schemes, inclusion of self or others in play, and constructive play).
Based upon the scoring protocol, points were given when a child independently exhibits the behavior in accordance with the assessment criteria (e.g., gaze shifts-child shifts gaze between toy and parent during interactions; gestureschild uses a wave to draw parent's attention to toy; words-child uses spoken word to label object during play; understanding-child responds to the question "Where's Initially, our intent was to examine social abilities (i.e., Social Composite) and expressive language abilities (i.e., Speech Composite) separately in the data analysis; however, these were found to be highly correlated with one another (r=.811, p<.01).
Therefore, the Total Composite was used in analyses as a measure of the child's overall social communicative ability.
Video data were coded by trained research assistants. Inter-rater reliability using Kendall's Tau-b was calculated for 20 % of the data. Kendall's Tau-b was used to determine inter-rater reliability as it is a statistical method that can be used with data representing ongoing judgements about behaviors in interactions that are not simply reflective of behavioral performance on discrete trials or during discrete time intervals. Kendall's tau-b for the total score was 1.0. (Appendix C) Child sensory processing ability.
Child sensory processing abilities were assessed utilizing the The Sensory Processing Measure-Preschool-(SPM-P) Home Form Miller Kuhaneck, Ecker, Parham, Henry, & Glennon, 2010). The SPM-P is completed by the parent and is designed to report a child's response to sensory experiences in the context of daily activities (e.g., how often does your child seem bothered by the sound of a vacuum cleaner; how often does your child enjoy watching objects spin or move; how often does your child gag or vomit in response to certain foods or textures). The form includes seventy-five items covering eight functional areas (i.e., social participation, vision, hearing, touch, body awareness, balance and motion, planning and ideas, and total sensory systems) that parents rate on a four-point frequency scale (1-never to 4-always). The scoring protocol specifies that raw scores from individual functional areas are summed to yield the Total Sensory score which is converted to a tscore. The Total Sensory t-score is an indicator of overall sensory processing, with higher scores indicating greater impairment/abnormality. This measure is included in Appendix D.

Parent behaviors.
Eight categories of parent behaviors previously identified as representing behaviors important for supporting the ESR of children (Grolnick et al., 1998;Gulsrud et al., 2010) were assessed during the observation. These categories included: 1) physical engaging/helping, 2) language-based engaging/helping, 3) redirection/distraction, 4) emotional following, 5) physical comfort, 6) vocal comfort/intonation, 7) language-based comfort/reassurance, and 8) active ignoring. Interrater reliability for each of the eight parent behaviors, as well as an additional "none of the above" category was calculated using percent agreement and Cohen's Kappa coefficients for 15 % of the study data and yielded an average score at or above 91 % for each of the parent behaviors (range = 91-100) and Kappa coefficients (κ range = 0.807-1.000), with the exception of Active Ignoring Based on previous research which examined parental behaviors in terms of their function and symbolic quality , the eight individual parent behaviors were combined to form five composites: 1) active engagement strategies, 2) comfort strategies, 3) passive strategies, 4) physical strategies and 2) language-based strategies. See Figure 1 for composite structure. Test of internal consistency (Cronbach's α) of each of these composites revealed poor internal consistency (α < .40) ; and therefore, individual parent behaviors were retained for analyses.

Procedures
Participants were recruited through medical and educational agencies. Letters introducing the investigators and describing this study (Appendix G) were sent to local early intervention (EI) providers, school districts, community support agencies, and local medical doctors/psychologists, as well as the Rhode Island Consortium for Autism Research and Treatment (RI-CART). Additional recruitment occurred at Autism awareness events where researchers were on site to discuss the project and to provide consent to contact forms. Follow up phone calls were made to all of these recruiting sources to further describe the study. Sources who agreed to share information with potentially eligible families were provided with informational flyers describing the study and consent to contact forms (Appendix H). Flyers contained the phone number of the researchers so that interested families could directly contact the researchers. Signatures on consent to contact forms were also obtained by referral sources and then returned to the study investigators.
Families were contacted to further describe the study and to determine child/parent eligibility once a consent to contact form was received and to describe parent roles (e.g., play partner, provider of snack, etc.), and to answer any questions about the research. A home visit was scheduled at a mutually convenient time when parents verbally consented to the study phone. The demographic questionnaire and the SPM-P  were mailed to the family for completion prior to the home visit (Appendix A and D).
At the beginning of each home visit, the researcher reviewed the study protocol with the parent and answered any questions. Upon confirmation of the child's eligibility to participate in the study, written consent for the parent and parental permission for the child (Appendix I) was obtained. Once informed consent was obtained, the demographic questionnaire and the SPM-P Home Form  were collected (Appendix A and D) and the home-based research protocol commenced. Three families had not completed their questionnaires prior to the home visit and were provided with self-addressed stamped envelopes and instructed to mail the forms back.
Two researchers attended every home visit, the investigator and a trained research assistant who video recorded the observation using a small monopod mounted camera. The structure of the home visit included a free play period, administration of the CSBS-DP , and a snack time. Video recording began with the start of free play and continued though snack time.
Each home visit began with a parent/child free play session in a location within the house of the parent's choosing. During this time the child was free to move about the room and interact with his/her parent and the researcher. Toys for free play included those of the child as well as novel toys provided by the researchers (e.g., Gazoobo shape sorter, Hoberman sphere, pop up toy). Free play concluded after each novel toy had been introduced and the the researcher and parent agreed that the child had had adequate time to explore all of the activities. Following the completion of free play, the child, parent, and researcher moved to a table of the parent's choosing within the child's home for the administration of the CSBS DP. Researchers attempted in all instances to position the parent and child with the parent and child seated next to one another and across from the researcher .
However, at times this seating arrangement varied due to space constraints within the house or the child having difficulty maintaining a seated position without physical support from his/her parent. In accordance with the CSBS DP protocol, activities were presented in a predetermined sequence. Initially, minimal direction was given to the child when toys introduced, then parents were instructed to interact naturally with their child.
At the completion of the CSBS DP assessment, the parent was asked to engage the child in a natural snack time routine. Parents were given the instruction to offer snack to the child as they typically do. If the parent chose to offer snack to the child in a different room or in a space that allowed for the child to move about freely, the camera was repositioned to capture the child and parent in the frame. Throughout the entire observation, if either the child or the parent moved out of the frame, attempts were made to maintain the focus of the camera on the parent. If the child and parent remained out of proximity to one another (e.g., greater than 60 seconds) the camera was moved to capture the child and parent in an effort to ensure adequate data capture.

Parent behavior.
Parent behavior was coded for each activity: free play, assessment and snack.
First, data were examined for codability. Time intervals in which the activity or the parent/child interaction was interrupted were deemed uncodable and excluded from the data. For example, time intervals during which another child demanding the parent's attention, the parent answered a phone call, or the parent talked to the researcher to the exclusion of the child, were excluded from further coding and analysis. On average over 80 % of 10 second intervals (M=81.63%, SD=10.31%) were codable across the combined observation for the sample.
Next, data were examined for variability and consistency across each activity.
Initial review of the video data for the snack activity indicated a wide range in the length of time from 1.25 minutes to 27.5 minutes, as well as significant variability in the qualities of the snack activity in which dyads were engaged (e.g., parent/child dancing, preparing valentines, placing food on a table, making a meal together). This wide variability in quantity and quality limited our ability to analyze the snack segments as a cohesive representation of a single activity. Therefore, the snack activity was dropped from further analyses.
The free play and the CSBS DP assessment data, were also examined for variability and consistency. Free play ranged from 8 minutes to 33 minutes in length (M=16.34 min, SD=5.74 min). Coding began for each observation with the introduction of the first novel toy by the researcher free play and continued for "up to 16" minutes to standardize the protocol, and reduce the wide variability in duration.
Mean duration of the coded segments of free play was 13.97 min (SD=2.46 min).
The administration of the CSBS DP included three consecutive "play based" tasks during which parents were directly encouraged to actively participate. Parents were asked to read a book to the child, engage in pretend play (e.g., feeding a stuffed animal), and assist their child with building a tower of blocks.

Data Analysis
Data were entered into Microsoft Excel and uploaded into SPSS 23.0 for analyses.
All variables were checked for accuracy and reviewed for trends, central tendency, variability and distributions. When assumptions of normality were not met, both square root and Log 10 transformations were attempted. Transformations resulted in fewer interpretable scores, and therefore; non-transformed variables were retained for analyses.
Parent behavior composites were created; internal consistency measures were confirmed and based upon the results parent behavior variables were either retained in their individual form or as composites. Data were examined for mean differences between protocol conditions (free play vs CSBS DP assessment) using t-tests; to determine whether parent behavior differed significantly by condition. Data were analyzed separately based on condition, as well as across the entire observation.
The association between demographic variables (e.g., race/ethnicity, socioeconomic status, age, etc.) and the independent (i.e., child social communication and child sensory abilities) and dependent variables (e.g., parent physical engaging and helping, parent redirection/distraction, etc.) were assessed in two ways. The association between categorical variables and the independent and dependent variables were examined using t-tests, and the continuous variables were examined using correlational analyses. Demographic variables that were significantly associated with the dependent and independent variables were entered as covariates in the final analyses.
Pearson product-moment correlations examined the association between the independent and dependent variables. Initial research questions and hypotheses were conceptualized around different domains of parent behaviors (e.g., active, passive, comfort, etc.). However, research questions were reframed to focus on individual parent behaviors because none of the composite scores achieved acceptable reliability.
Regression analyses were conducted to test the modified research questions for those parent behaviors that were significantly associated with the independent variables, controlling for covariates.

Research questions analyses.
Research question 1-Are the social communication abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR?
Research question 2 -Are the sensory processing abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR?
Each dependent variable was regressed separately on the independent variables: child social communication and child sensory processing. Child age and race/ethnicity (i.e., White/ non-white) were entered as covariates, as were interaction terms (i.e. social communication by race/ethnicity or sensory processing by race/ethnicity) as appropriate.
Analyses were also run separately by condition (i.e., free play and assessment) and for the combined observation. All results are reported for the combined observation. Additional results are reported for individual conditions when they differed from the combined condition results.

Descriptive Analyses
Demographics.
Participants were 37 parent/child dyads. Families in this study were racially, ethnically, and economically diverse ( At the time of the study, all child participants were engaged in some form of intervention services. The average duration of intervention programming was 17.42 hours/week, (SD=10.55 hours). Participants reported receiving services that included speech language therapy (92%), occupational therapy (83%), educational services (62%) and Applied Behavioral Analysis therapy (57%). Parents reported that high levels of intervention priorities and goals of services included verbal communication (78%), social interaction (68%), and attention and focus (70%) with less emphasis on coping and soothing (43%) and on augmentative communication (24%). See Appendix K for complete information related to educational programming and intervention services for the child participant.

Child characteristics.
Children's social communication profiles and sensory processing ability scores are reported in Table 3

Parent behaviors.
Proportion scores of parent behaviors, for the combined observation and by condition are presented in Table 4. Overall, during the combined observation parents engaged in relatively high amounts of physical (M= 49.83,SD=16.41) and languagebased (M=52.67, SD= 17.28) engaging and helping behaviors. Parent use of redirection and emotional following was less frequent while comfort behaviors (i.e., physical, vocal, and verbal) and active ignoring were rare. On average, parents were not engaged in any of the previously identified behaviors associated with supporting child ESR more than 20% of the coded time. Similar parent behavior patterns were observed during the free play and CSBS DP assessment conditions. See Appendix L for a summary of parent behaviors coding.

Differences in parent behavior by condition.
Paired Sample t-tests were conducted to determine whether the mean of coded responses of parent behaviors was statistically significant by condition. Results revealed significant differences in several parental behaviors, with higher frequency scores generally noted in free play conditions as compared to the assessment. Parents engaged in more physical engaging and helping (t=4.29, p=0.000) and verbal engaging and helping (t=2.64, p=0.012), and more emotional following (t=2.93, p=0.006) during free play than during the CSBS DP assessment (Table 4). All subsequent analyses were performed by condition as well as for the combined observation.

Association between Demographic Variables, Independent and Dependent Variables
The association between child characteristics, parent behavior, and categorical demographic variables (i.e., child sex, White/children of color, income) were examined using t-tests and ANOVA. Results indicated significant differences on parent and child characteristics based on race and ethnicity of the family. As compared to children of color, White children scored higher on the social communication assessment and lower on the sensory processing measure indicating overall fewer deficits (Table 5). Additionally, income was significantly associated with SPM-P scores (ANOVA, F(5, 31)=3.709, p <.01). Post-hoc analyses were unable to be performed due to groups with fewer than two cases. See Appendix M for further details related to SPM-P and family income.
In terms of parent behaviors, parents of color used redirection/distraction significantly more often than White parents during the combined observation (t(35)=-2.547, p<.05) and the CSBS DP assessment condition (t(35)=-2.382, p<.05). In contrast, parents of color used language and helping behavior significantly less often than White parents during free play (t(35)=2.425, p<.05). No other differences were observed for parent behaviors based on race/ethnicity, income, or child sex.
The association between child characteristics, parent behavior, and continuous demographic variables (age and hours engaged in intervention) were examined using Pearson product moment correlations. Correlations between continuous demographic variables (age and hours engaged in intervention) and child social communication abilities and sensory processing abilities revealed no significant associations (Table 6).
During the combined observation, child age was significantly correlated with the parent physical engaging and helping (r(35)=-0.403, p=.013), with younger children receiving more parental physical engaging and helping. Child age was not significantly correlated with any other independent or dependent variables (Table 7). Similar relationships between continuous demographic variables and parent behaviors were reported for free play and the CSBS DP assessment. A significant correlation between child age and parent physical engaging and helping were reported for the free play condition (r(35)=-.419, p=.010), but not for the CSBS DP assessment condition. No other significant correlations were reported for individual conditions (Table 7).

Inter-correlations Among Parent Behaviors
The inter-correlations among parent behaviors are reported in Table 8 for the combined observation. Two variables that represent parent engaging or helping (i.e., physical engaging and helping and language engaging and helping) were positively associated (r(35)=0.382, p=.020). Likewise, three variables assessing parent comfort (i.e., vocal comfort, physical comfort, and language-based comfort) were positively associated with one another. Redirection/distraction was inversely related to parent language engaging and helping (r(35)=-.352, p=.033) and positively associated with more emotion following (r(35) =.398, p=.015). Finally, active ignoring was positively associated with more language and vocal comfort (r(35)=.463, p=.004).

Associations between Child Characteristics and Parent Behaviors
The correlations between child characteristics and parent behaviors are reported in Table 9, for the combined observation and by condition. In general, children who scored higher on social communication had parents who used more language based behaviors (r(35)=.389, p=.009) and fewer physical engaging and helping (r(35)=-.367, p=.013), fewer redirecting/distracting (r(35)=-.548, p=.000) and less physical comforting (r(35)=-.373, p=.012). Child sensory processing scores were not associated with parent behaviors at statistically significant levels during the combined observation.
Patterns of associations within each condition showed similar results with only minor differences. During the free play condition, child sensory processing ability was inversely associated with parental engaging (e.g., children with more deficits had parents who engaged in less physical engaging and helping) (r(35)=-.301, p=.035) and the association between social communication and physical engagement no longer reached conventional levels of statistical significance (r(35) = -.259, p = .06). In addition, child social communication and parent physical comfort were not statistically associated during the free play condition. Finally, during the CSBS DP assessment, parental language was not statistically significant associated with child communication abilities (r(35)=.203, p=.115).

Research Questions
Research question 1. Are the social communication abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR? Hypothesis 1. Higher levels of child social communication abilities will be associated with lower levels of parent physical engaging and helping, physical comfort, redirection/distraction, vocal comfort, and emotional following.

Hypothesis 2.
Higher levels of child social communicative abilities will be associated with higher levels of parent language-based engaging and helping, language-based comfort, and active ignoring.
The first set of hypotheses were tested using regression analyses to assess the association between child social communication and parent behaviors: language engaging and helping, physical engaging and helping, redirection/distraction and physical comfort behaviors. Additional parent behaviors (i.e., language-based comfort, vocal comfort, emotional following, and active ignoring) were not included due to low p=.056); however, the model examining physical comfort did not.

CSBS DP assessment.
During the CSBS DP assessment condition, only the model examining child social communication abilities in relationship to redirection/distraction parent behaviors was significant (Table 12). Child social communication scores were significantly inversely associated with redirection/distraction after controlling for age and ethnicity (β=-.557, p=.001). Child age was also a significant predictor in this model (β=.338, p=.027.
Race/ethnicity was not significantly associated with language engaging and helping in this model.

Research question 2.
Are the sensory processing abilities of young children diagnosed with ASD associated with parent behaviors that are supportive of the development of child ESR? Hypothesis 1. Higher levels of child sensory processing abilities will be associated with lower levels of parent physical engaging and helping, physical comfort, redirection/distraction, vocal comfort, and emotional following.

Hypothesis 2.
Higher levels of child sensory processing abilities will be associated with higher levels of parent language-based engaging and helping, language-based comfort, and active ignoring.
Similar regression analyses were used to assess the association between child sensory processing scores and high frequency parent behaviors: physical engagement and helping and language-based engagement and helping. Additional parent behaviors (i.e., physical comfort, language-based comfort, vocal comfort, emotional following, redirection/distraction, and active ignoring) were not included in analyses due to low frequency of occurrence and lack of bivariate association with the child characteristics. All analyses controlled for child age, race/ethnicity and an interaction variable representing the relationship between child sensory processing abilities and race/ethnicity. Results examining these relationships for the combined observation yielded one significant model (Table 13) for parent physical engaging and helping.
Child sensory processing abilities were significantly associated with parent physical engaging and helping (β=-1.255., p=.040) after controlling for age and ethnicity, as well as the sensory x race/ethnicity interaction variable. Child age, race/ethnicity, and the interaction variable representing sensory processing abilities and race/ethnicity were not associated with physical engaging and helping in this model. The model for language engaging and helping approached, but failed to meet conventional levels of statistical significance (F(4,32)=2.656, p=.051) (Table 13).

Analyses by condition.
Similar analyses to assess the association between child sensory processing abilities and parent behaviors were undertaken by condition. These analyses, controlling for child age, race/ethnicity, and interaction terms yielded several differences.

Free play.
Analyses assessing the relationship between child sensory processing scores and physical engaging and helping and language-based engaging and helping yielded two significant models for the free play condition (Table 14). All analyses controlled for child age, race/ethnicity and an interaction variable representing the relationship between child sensory processing abilities and race/ethnicity. Child sensory processing abilities were significantly associated with parent language engaging and helping after controlling for age and race/ethnicity , as well as the sensory x race/ethnicity interaction variable (β=-1.360, p=.028). Race/ethnicity (β=-5.04., p=.024) and the interaction variable representing sensory processing abilities and race/ethnicity (β=5.337, p=.034) were associated with language engaging and helping in this model.
Child age was not associated with language engaging and helping in this model. Despite a significant model, sensory processing abilities were not associated with parental physical engagement at conventional levels of statistical significance (β-.990, p=.105) in this condition. Child age was significantly inversely associated with parent physical engaging and helping in this model (β=-.328, p=.049).

CSBS DP assessment.
During the CSBS DP assessment condition, none of the models examining child sensory processing abilities in relationship to physical and language-based parent behaviors were significant (Table 15).

Summary of Findings
The present study examined the association between behavioral characteristics of young children diagnosed with ASD and their parent's engagement in behaviors associated with supporting the development of ESR in children. To date, relationships examining child social communication and sensory processing abilities and parent behaviors have not been explored. Given the importance of the parental role in supporting the development of ESR (Kopp, 1989;Sameroff & Fiese, 1990;Tronick, 2002), understanding parental behaviors in relation to their child's diagnostically significant characteristics is an important area of inquiry in understanding ASD.
The current study yielded three main findings related to the research questions and hypotheses of interest. First, in support of our first set of hypotheses, several parent behaviors were associated with child social communication abilities. Parent's use of physical engaging and helping, redirection, and physical comfort were associated with lower levels of child social communication abilities, while parent use of language engaging and helping was associated with higher levels of child social communication. Second, we found limited support for our hypothesized associations between child sensory processing abilities and parent behaviors associated with supporting ESR. Of the eight parent behaviors assessed only one, physical engaging and helping, was associated with child sensory processing. Third, based upon the literature we expected parent behaviors to be grouped in theoretical composites according to function and symbolic quality. However, we did not find support for these previously identified theoretical composites of parent behaviors which we had intended to include in our analyses. Therefore, only individual parent behaviors were retained for our analyses. Each of these key findings, as well as, several additional findings will be discussed in the context of the current literature.

Associations between child social communication and parent behaviors
As expected, child social communication was associated with a number of parent behaviors. Specifically, as high levels of child social communication abilities were associated with lower amounts of parental physical engaging and helping, redirection/distraction, and physical comfort decreased. In addition, we found that parent physical engaging and helping was inversely associated with child age in our sample. Collectively, these findings suggest parents of children with lower social communication abilities engage in higher amounts of physical engaging and helping, as well as redirection/distraction. A finding that is in agreement with the normative development literature related to chronologically younger children. Previous studies have reported that parents of young typically developing children (<24 months) frequently utilize physical engaging behaviors and redirection/distraction when interacting with their children (Eisenberg & Spinrad, 2004;Grolnick et al., 1998).
Given that the children in our study were older (30-48 months) than those in studies of typically developing children, it is interesting to consider our findings related to parents' use of physical engaging and helping in relation to studies in the ASD literature. Several studies have suggested that parents modify interaction styles based upon the developmental age of the child as opposed to keeping with chronological age norms seen in studies of typical development. Kasari and colleagues (1988) reported that parents of 4 year olds diagnosed with ASD engage in high levels of physical support and assistance when scaffolding their children's play when compared to parents of typically developing children. The fact that parents used more physical comfort with children with social communication delays also mirrors previous literature that suggests that parents of developmentally younger children adapt their interactive style and utilize more physical engaging behaviors (Eisenberg & Spinrad, 2004;Grolnick et al., 1998).
In addition to these inverse associations between child social communication abilities and parent behaviors, we also found that parents' use of language-based engaging and helping increased as child social communication increased. This finding also aligns with previous research related to the development of ESR in typical populations. Parents of older, more developmentally-advanced, typically developing children frequently utilize language-based strategies to help maintain child engagement (Grolnick et al., 1998;Sameroff & Fiese, 1990). This finding is consistent with additional studies in the ASD literature that have found that parents modify their interactions to the developmental level of their child (Hirschlerguttenberg, Golan, et al., 2015;Kasari et al., 1988;Kasari, Gulsrud, Wong, Kwon, & Locke, 2010). In particular, Kasari and colleagues (1988) reported that parents' use of language when supporting play was associated with more advanced child social communicative abilities. And, Hirschler-guttenberg and colleagues (2015) have reported parents' differential use of parent use of language-based strategies with autistic preschoolers based upon their cognitive abilities (e.g., higher child IQ associated with more parent language).
Collectively, our findings related to physical engaging and helping and language-based engaging and helping, as well as redirection and distraction suggest sensitive parenting practices, which involve modifying interactive style based on child developmental level. While directions of associations cannot be inferred from our data, these parent behaviors are associated with the sophistication of their child's social communication abilities.

Associations between child sensory processing and parent behaviors
In contrast to our significant findings linking child social communication and parent behaviors, we found minimal support for our second set of hypotheses that child sensory processing would be associated with parent behaviors. Of the eight parent behaviors, only physical engaging and helping was positively related to child sensory processing abilities. Parents of children who had better sensory processing used greater amounts of physical engaging and helping. This finding runs counter to our hypothesized relationship that higher sensory processing abilities would be associated with lower amounts of physical engaging and helping. It also stands in contrast to previous work which relates to sensory sensitivity and parent use of physical strategies in typically developing children (Cole et al., 2013). This finding may in part be due to the nature of the SPM-P. This tool looks broadly at sensory processing differences (e.g., social participation, vision, hearing, body awareness, etc.), whereas previous research in typical development has focused on measures of sensory over-reactivity and related temperamental differences. Alternatively, it might be a unique finding that parents may be more "hands off" with children who demonstrate increased sensory processing impairments in an effort not to complicate the child's sensory environment/experience. Future research is needed to further examine and better understand these associations.

General Discussion
It should be noted that a number of parent behaviors were not associated with either child social communication or child sensory processing (e.g., vocal comfort, language-based comfort, emotional following, active ignoring, etc.). In contrast to high frequency behaviors (e.g., physical engaging and helping, language-based engaging and helping, and redirection/distraction), these behaviors were observed only rarely. Therefore, it is possible that the lack of association was due to the low frequency of occurrence.
Several possible explanations exist for the low frequency of behaviors. While all of the behaviors included in the study are common parent behaviors, some of the behaviors may be more likely to be used/observed under specific conditions. For example, parent use of comfort behaviors may be generally more likely observed during times of child distress. While not formally measured, anecdotally, parents appeared to use comfort more often when their children were demonstrating more stress. As a group, the children in this study were observed to be fairly well regulated, and as a result, may have elicited less need to respond with comfort behaviors. And while episodes of distress, intense emotional expression, and arousal changes were observed during the visits, they were not the prevailing child behavioral presentation.
Children's infrequent distress may have been in part due to the nature of the home visit. Despite the inclusion of a variety of activities in the home observation protocol which were designed to capture a range of parent child interactions, we did not intend to induce stress. It may be that being in their home environment, interacting with their parents, and having access to familiar activities and objects, helped to reduce child stress and, in turn, the need for parental comfort. Therefore, both the nature of the activities and the child's emotional state may have contributed to the low frequency of several parent behaviors.
It should be noted that the children's level of organization and engagement anecdotally reported in our study stands in contrast to the behavioral presentation of young children with ASD described in many lab-based studies where children experienced frequent distress. Given this difference in child presentation, it is possible that our findings more accurately reflect daily interactions between parents and young children diagnosed with ASD and also more accurately represent a wide range of effective ESR child abilities. Further study of the interactions in the home and aspects of the home environment may provide important information related to the ESR capacities of young children with ASD. Additionally, more fully understanding the nature and context of supportive environments reinforces the need to continue to conduct research within families' natural environments.
The literature on emotional regulation has often discussed parent behaviors associated with ESR in relation to their function and/or their symbolic quality. Our coding protocol was originally designed to reflect these theoretically-based categories of behavior. However, we found no empirical support for these constructs. While theoretically meaningful, previous studies which have conceptualized (Grolnick et al., 1998) or reported on similar composites  have offered minimal empirical support for the validity of the constructs. Therefore, it is possible that our failure to find internal consistency is not an issue of replication, but rather a reflection of the difference between theoretically driven composites and those that achieve statistical relevance. In addition, it is possible that our small N and the relative low frequency of several of the parent behaviors included in the composites contributed to our lack of anticipated findings of cohesive composites.
Despite the lack of composite structure, the associations among parent behaviors support the validity of our measures. We found relevant and meaningful associations between behaviors associated with functions of behavior. Specifically, engaging and helping variables were associated independent of whether they were physical or verbal. This suggests some cohesion in terms of function which was hypothesized; however, also unique from the original "active" behavior composite which also included redirection/distraction. Likewise, comfort behaviors (e.g., vocal, physical, and language-based) also hung together indicating a functional relationship.
Here again, an additional variable included in the proposed "comfort" behavior composite (i.e., emotion following) was not related. Collectively, these results seem to indicate a need for refinement and greater specificity in the originally proposed composites representing functions of parent behavior. Additionally, several other associations between parent behaviors were found that warrant consideration when conceptualizing functional composites of parent behaviors and or considering patterns of parent response. For example, high amounts of parental redirection/distraction were associated with high levels of emotion following and low levels of engaging and helping. These relationships may reflect established developmental patterns associated with normative development. Parents of developmentally young children redirect and distract based upon the child's emotional experience and expression; however, as the child's developmental skill levels (e.g., ESR, language abilities, cognitive capacities, etc.) increase with age and maturation, parents shift to more language-based behavior (Grolnick et al., 1998). These findings indicate that parents were engaging in behaviors in response to their child's behavior presentation and their developmental abilities. In an effort to better understand parent behaviors in relation to theoretical constructs of function, future research should focus on these associations.
As previously stated, our observation protocol was designed to capture parent behaviors across a variety of activities in naturally occurring settings and was intended to be reflective of typical parent child interactions. During our initial scanning of the data, parent behavior seemed to vary based on activity. Therefore, we decided to examine the data according to different conditions (i.e., free play and assessment) and also as a combined observation. Our analyses revealed several patterns of parent behavior which were consistent across all three conditions. In general parents tended to engage more frequently in physical engaging and helping, language-based engaging and helping, and redirection/distraction than in behaviors associated with comfort (i.e., physical comfort, vocal comfort, language-based comfort or emotional following). Parent use of active ignoring was rare.
Minimal variations in the amount of behaviors emerged by condition. In general, parents used more engaging and helping behaviors during free play than during the CSBS DP assessment; and, they engaged in more redirection and distraction during the CSBS DP assessment than during free play. These differences in amount of behaviors appear to be related to the unique qualities in the nature of the activities. During free play, parents were able to engage their children in activities of their choosing without expectations and often followed their children's focus of attention. In contrast, during the CSBS DP assessment, predetermined toys were introduced in a structured protocol for the purpose of assessing the child's skills. In this instance parents may have engaged in greater amounts of redirection/distraction in an effort to encourage their children to "do their best." Despite these slight differences in amount of behaviors reported by condition, as previously stated the distribution of behaviors was consistent. Therefore, it appears that parents are consistent in the types of behaviors they use when interacting with their children regardless of the activity they are engaged in.
Overall, parents demonstrated high levels of engagement with their children.
Eighty percent of the time parents were engaged in at least one of the parent behaviors included in our protocol. While it is not possible to comment on what was happening during the additional 20% of the time without further inquiry, this overall finding reflects that parents in our study were attentive to and interactive with their children for the overwhelming majority of time during a variety of naturally occurring activities in the home environment.
The parents and children in our study were a unique sample who welcomed us into their homes. The children participating in our study were between 30-48 months of age (m=40 months). As a group, they were not observed to be generally stressed during the home visit; however, individually they did exhibit a wide range of emotional states and arousal levels throughout the visits. Removal of preferred toys and transitions between activities appeared to be the most frequent causes of distress.
As a group they were relatively engaged in the home visit activities and took a particular interest in novel toys (i.e., Hoberman sphere) introduced by the researchers. also demonstrated considerable sensory processing challenges according to parent report (e.g., constant seeking of movement, sensitivity to sounds, withdraw from busy environments, etc.). However, here again, individual abilities ranged widely from functioning considered typical for their age group to significant challenges that could impact most aspects of daily life. In the current body of ASD literature, there is no comparable group with regards to age, diagnosis, and child characteristics reported to compare our sample of child participants to. Therefore, we do not know how representative our sample is of the ASD population at this range. However, we feel confident given our review of child characteristic scores that our sample represents the broad spectrum of abilities seen in Autism diagnoses.
Participants in this study were racially and ethnically diverse. In contrast to much of the research of children with ASD at this age, 32.4% of our sample identified as families of color. These data is consistent with racial and ethnic diversity reported in the national census (US Census Bureau, 2014). Historically, young children of color are under-represented in the ASD literature. This is in part due to the average age of diagnosis for children of color is 5 years of age, in comparison to White children who are diagnosed on average at 2.5 years of age (ADDM, 2012). Several factors are often cited in relation to these diagnostic age differences. They include families of color's lack of connection to services, as well as differing cultural expectations of behavior for young children.
The children of color in our sample demonstrated more significant delays in social communication and sensory processing than the White children in our study.
These significant delays may have accounted for their diagnosis earlier than the national average for their race and ethnicity, and for our ability to recruit their participation. Additionally, our ability to recruit a diverse sample may have been related to families of color often being underserved by educational and therapeutic systems. Several parents of color commented on how grateful they were to have someone coming into their home to talk with them and see the realities of daily life.
At the conclusion of our home visits, many of these families asked questions related to how to access statewide services and/or how to advocate for school based interventions, which suggests lack of family support mechanisms despite a clear need.
The diversity of our sample allowed us to explore group differences between parent behaviors used by parents of color and those used by White parents. Very few differences emerged. Only two warrant mentioning. Parents of color utilized significantly more redirection and distraction than White parents during the CSBS DP assessment. They also used less language-based engaging and helping during free play. However, after controlling for child social communication ability these differences were no longer significant. As such, any interpretation of these findings should be undertaken with caution, as the group sizes in this study were small. Further research is needed to explore the associations between parent behaviors and child characteristics for racial and ethnically diverse groups.

Strengths of the Present Study
One of the major strengths of this study was that the data were collected during home visits in the child's natural environment with their parents. Most previous research on ASD has been conducted in clinic-based settings and has focused on times of child stress rather than in natural environments and being inclusive of all child experiences. This study was designed to address these shortcomings in the literature in an effort to gain a broader understanding of behaviors parents in the context of daily activities within the natural environment. In addition, the observational methods employed enabled us to capture and quantify data related to parent behaviors in naturalistic interactions rather than relying on parent report related to their interactive style. Observations across a number of activities allowed for a more detailed analysis and our ability to consider the influence of activity/context on parent/child interactions.
Participants in this study had a confirmed diagnosis of ASD and were between the ages of 30-48 months, an important age for exploring ESR development.
Participants in ASD research often represent large age ranges spanning many years which make it difficult to interpret findings. Our relatively narrow age range for child participants in this study was purposeful to address this shortcoming in the literature.
Likewise, the considerable diversity in the sample with respect to race/ethnicity as previously discussed is considered to be a strength as families of color are often not well represented in ASD research despite their representation in the ASD population at large.

Limitations of the Present Study
A limitation of this study is that the findings reported here are cross sectional in nature and do not allow for causal conclusions about child characteristics and parent behaviors, nor do they allow us to make inferences related to change over time or compare to typically developing peers group parent/child interactions. The data provide only a brief snap shot of parents' overall interactive styles. While parents reported their children on a whole behaved as they typically would and that their interactions with them were natural, the data may not be representative. Parents were aware that researchers were interested in how they supported their children. This may have served as a prime for parents. Therefore, social desirability may have influenced their behavior and/or answers on parent report measures.
Additionally, given that this was an observational study the presence of the two researchers in the home, as well as the camera, may have been a factor influencing both parent and child behavior. As previously mentioned, despite variation in emotional expression and arousal levels, collectively child participants were fairly well regulated throughout the observation. This overall presentation may in part have been related to the study being conducted in the comfort and familiarity of their home environment, which is supportive of regulation. In turn, this may have decreased our ability to assess behaviors engage in when supporting their children during stressful interactions, activities, and environments.
And finally, our coding scheme of parent behavior included categories of behavior that were relatively large and did not account for specific qualities of the behavior. For example, the parent behavior language-based engaging and helping consisted of behaviors ranging from giving the child a direction to commenting on how a toy worked. Therefore, much remains to be investigated with relation to the specific qualities of parent behaviors and their potential impact upon the development of child ESR.

Implications of Study Findings
Our results suggest that parents engage in a variety of behaviors that have been previously associated with supporting child ESR during interactions with their children, and that their engagement in these behaviors is associated with the social communicative abilities of their child. This is considered to be initial evidence related to the relationship of parent behaviors theoretically supportive of ESR and diagnostically significant ASD child characteristics. These initial findings are a critical first step towards greater understanding of the factors impacting ESR development among children with ASD. They may also serve to inform the design of parent based intervention approaches. It is generally acknowledged that ESR challenges emerge early in development for this population and persist, negatively impacting engagement in daily activity and quality of life. Gray and colleagues (2014) highlighted the long term implications of behavioral challenges related to ESR difficulty for adults with ASD citing greater needs for community support and reduction in employment. Therefore, understanding risk and protective factors related to the development of ESR is essential in the design of effective interventions targeted at minimizing developmental challenges.
The findings of the study have additional implications for consideration related to race/ethnicity for families of young children diagnosed with ASD. We found greater impairments in social communication and sensory processing for young children of color than for White children, lending evidence to the assertion that children of color who demonstrate greater skill in these areas are often diagnosed later and do not have access to early intervention services which are thought to be critical for positive long term outcomes for individuals diagnosed with ASD (ADDM, 2012).

Future Directions
Based on study findings and limitations, future research should seek to clarify a number of questions raised by the current study. The focus of this study was on parent behavior in relation to child characteristics. However, in an effort to further understand the complexities inherent in the transactional nature of ESR development, coding for child emotional state in addition to parent behaviors is considered to be a logical next step of inquiry. This additional data will allow for the exploration of relationships between parent behaviors and particular child emotional states. Previous work has suggested parents of children with ASD use less frequent language based behaviors in times of stress .
In addition, coding for ESR behavior used by children during these same observations will allow for examination of the relationship between parent behaviors supportive of ESR and child ESR strategies. Ideally, this work would be conducted in a longitudinal manner and in natural environments which would allow for gathering of information related to growth and development of skill over time in natural environments. Likewise, the utilization of qualitative methods to further explore of the current video data set for themes related to parental experience supporting ESR and parental impressions of their child's ESR abilities is considered to be important for providing additional context for the quantitative findings. Collectively, all of these relationships will be important in helping to provide a more comprehensive understanding of parenting practices related to supporting young children with ASD's ESR and the implications of those practices on child ESR abilities. Such knowledge could then be used to inform the development of targeted parent based interventions that may decrease the secondary burdens and long term challenges posed by ESR challenges.
In addition, although this sample is relatively large by ASD research standards, recruitment of a larger sample could be helpful to further understand the associations examined in our study and also potentially illuminate associations with lower frequency behaviors. This may be particularly relevant to the further examination of associations between sensory processing and parent behaviors. Furthermore, future research within culturally diverse populations is warranted to further understand racial/ethnic group differences that emerged in our findings in relation to both child characteristics and parent behavior.
Finally, parent behaviors associated with supporting ESR in the current study are defined broadly. Previous work exploring the relationship between parent language models and child language acquisition have examined specific qualities of language used in order to establish more specific relationships. Refining the categories of the parent behaviors included in our study for future research projects could help to provide additional, more specific information which may be useful in the design of targeted interventions. Additional consideration should also be given to the theoretical constructs for categorizing parent behaviors in an effort to help refine our understanding of parenting strategies and also potentially to reconsider how we measure them.