Positive and multifaceted perceptions of Israeli kindergarten teachers contribute to developmental gains of children with developmental delay

ABSTRACT

This study examined the links between Israeli teachers’ perceptions of children with developmental delay (DD) in special education kindergartens and changes in children’s development. Participants were 47 teachers and children diagnosed with DD. At Time 1, teachers’ perceptions were assessed via the valence of their narrative about the child and the extent to which the narrative was multifaceted (i.e. addressing more developmental domains). Kindergarten therapists reported twice, at Time 1 and at Time 2 (12 months later), on children's adaptive behavior in terms of communication, socialization, motor, and daily living skills and on children’s peer acceptance. Positive valence of teachers’ narratives was associated with more gains in children’s adaptive behavior from Time 1 to Time 2. Multifaceted narratives were associated with children’s increased peer acceptance from Time 1 to Time 2. Results highlight the importance of supporting teachers’ construction of positive and multifaceted perceptions of children through training and supervision.

KEYWORDS:

• Special education kindergartens
• developmental delay
• Israel
• teachers’ perceptions

Introduction

Children with developmental delay (DD) are slow to reach various milestones, including in motor, cognitive, speech, and language development, as well as in play, and daily living skills (Zablotsky et al., 2019). They also present lower levels of social development (McIntyre, Blacher, & Baker, 2006; Vandesande, Van Keer, Dhondt, & Maes, 2022) and tend to exhibit major and pervasive difficulties in peer interaction skills (Guralnick, Neville, Hammond, & Connor, 2007), such as entering peer groups (Wilson, 1999), maintaining play (Guralnick, Hammond, Connor, & Neville, 2006), and forming friendships (Guralnick et al., 2007). Poorer social skills predict poorer school adaptation, even after controlling for IQ and adaptive behavior (McIntyre et al., 2006). It is therefore important to identify ways to help children with DD attain developmental milestones and improve their social skills.

A supportive teacher–child relationship is thought to foster academic skills and the cognitive and socio-emotional development of typically developing children (Hamre, 2014). A handful of studies have provided support for this notion in the context of children with DD, showing closeness and less conflict in teacher–child relationships predict better learning behaviors (Rhoad-Drogalis, Justice, Sawyer, & O'Connell, 2018), higher social competence, and fewer behavior problems (Blacher, Baker, & Eisenhower, 2009). The present study extended previous work by examining the role of two other aspects of teacher–child relationships: the valence of teachers’ perceptions of children with DD and the extent to which these perceptions are multifaceted.

Research suggests caregivers’ perceptions (i.e. their thoughts, feelings, and expectations) of children under their care guide their behavior toward the children (e.g. Vreeswijk, Maas, & van Bakel, 2012) and may shape the children's development (e.g. Sher-Censor, Khafi, & Yates, 2016). To evaluate caregivers’ perceptions, including those of teachers, researchers have recommended taking a narrative approach and focusing on the valence of caregivers’ narratives about the child, as well as the extent to which their narratives are multifaceted and address varied developmental domains (Sher-Censor, 2015; Spilt, Koomen, & Thijs, 2011). Positive valence in narratives presumably reflects positive expectations and stance. As such, it may relate to a more positive interpretation of children’s behavior during teacher–child interactions. This may result in teachers’ supportive responses and their attuned structuring of children’s behavior and learning, and this, in turn, may foster development. Multifaceted narratives are thought to indicate complex and broad perceptions of children’s cognitive, motor, and socio-emotional needs. Such perceptions may facilitate teachers’ support of these needs (Pianta, Belsky, Vandergrift, Houts, & Morrison, 2008). Teachers whose perceptions are limited and do not include all developmental domains, for example, those who are not aware of the importance of the social domain may not address it in their narratives and may be less likely to support this aspect of children’s development during their interactions (Spilt, Koomen, Thijs, & van der Leij, 2012).

Studies with parents found positive valence of parental narratives was associated with concurrent or subsequent reduced child symptoms (e.g. in the context of children with ASD: Hickey, Bolt, Rodriguez, & Hartley, 2020) and better socioemotional outcomes, including social competence (e.g. in the context of children with typical development: Narayan, Sapienza, Monn, Lingras, & Masten, 2014; in the context of children with ASD: Benson, 2013). Other studies showed that multifaceted narratives of parents, which portrayed varied aspects of the child, predicted better socio-emotional development, including in terms of peer acceptance (Sher-Censor et al., 2016).

Initial support for such links in the context of teacher–child relationships was obtained in two studies assessing teachers’ narratives about children with typical development in their classrooms. The first study found negative valence of teachers’ narratives was associated with expressions of more negative affect toward children in kindergarten and first grade classrooms (Stuhlman & Pianta, 2002). The second study suggested negative valence of teachers’ narratives related to concurrent higher levels of children’s behavior problems in secondary school classrooms (Daley, Renyard, & Sonuga-Barke, 2005). Our study added to the existing literature by examining the valence of teachers’ narratives about children with DD, the extent to which their narratives are multifaceted, and their respective links with children’s developmental outcomes.

The current study

This study focused on special education kindergartens in Israel serving children with DD. Ten percent of Israeli children are diagnosed with DD (Israel Ministry of Health, 2022). Mainstream public kindergartens are characterized by a large class size and a poor teacher/child ratio of 2/35. Only a minority of children with DD receive education within mainstream kindergartens with specialized assistance. The majority attend special education settings tailored to their needs. These settings are designated as kindergartens for ‘developmentally delayed children.’ Children diagnosed with ASD and those with moderate to severe motor disabilities attend different special education kindergartens, namely ‘communication kindergartens’ and ‘rehabilitative kindergartens,’ respectively (Israel Ministry of Education, 2022). Special education kindergartens serve children from age three to age seven and accommodate up to 14 children. The kindergartens’ setting is similar to that of a preschool. The staff includes a special education teacher, assistants, and therapists (i.e. occupational, speech, and emotional therapists). Teachers are the classroom managers. They supervise all learning processes, interact with all children daily, and are expected to combine individual, small group, and whole-group teaching practices. The declared goals of special education kindergartens are to support children’s development, reduce their gaps in socialization, communication, motor, and daily living skills, and increase their social adaptation (Israel Ministry of Education, 2022). Thus, teachers may play a significant role in supporting the development of the children in their classes.

The goal of our study was to examine the contribution of teachers’ perceptions to the development of children with DD. To that end, we used a quantitative repeated measures design. We hypothesized that teachers’ narratives about a child with DD that had a positive valence and teachers’ narratives that addressed more developmental domains would be associated with an improvement over time in the child's functioning. In particular, we expected that both features of teachers’ narratives would be associated with improvement in children’s adaptive behavior (in socialization, communication, motor, and daily living skills) and peer acceptance. To enrich the quantitative analyses, we included elucidations of narratives’ valence, the way teachers’ narratives addressed children’s developmental domains, and the relations of both with changes in children’s functioning.

Methods

Participants

Forty-seven teachers from special education kindergartens designed for children with DD (each from a separate kindergarten) and a randomly chosen child with DD from each class took part in the research. Kindergartens were situated in the northern part of Israel. In all participating kindergartens, children received speech therapy, occupational therapy, and emotional therapy once a week.

Ten of the 47 participating children (21.28%) were female. Their ages ranged from 44 months to 90 months (M = 58.63 months; SD = 9.40). To be included in the study, children had to be diagnosed with DD according to DSM-V criteria (American Psychiatric Association, 2013) and experience delays in more than one developmental domain. Four children had an additional diagnosis; one child was diagnosed with mild cerebral palsy, and three had a genetic disorder: 22q11.2 deletion syndrome, Angelman syndrome, and Moymoya syndrome. Children with these syndromes typically have DD.

Average teachers’ age was 37.64 (SD = 7.30). All were females, and 83% were married or cohabiting. One (2.1%) teacher had only a teaching certificate (in Israel, a bachelor’s degree was not mandatory for kindergarten teachers until 1995); 34 (72.3%) had a teaching certificate and a bachelor’s degree; 12 (25.5%) had a teaching certificate and a master’s degree. Teachers’ average years of experience was 11.00 (SD = 8.11).

Procedure

This study was part of a larger research project on the relationship between teachers and children with DD (Dolev, Sher-Censor, Zach, & Shalem Gan-Or, 2023). Data were collected in the kindergartens in two consecutive years. To allow children’s adaptation to the kindergarten setting and enable a meaningful acquaintance between teachers and children, the data collection process continued from November to July. Letters explaining the study were sent to kindergarten teachers, followed by a recruitment phone call. Parents of a randomly chosen child from each kindergarten were asked to sign a consent form. Eighty-six kindergartens were contacted. In 15 kindergartens, none of the children had a formal DD diagnosis (although they all presented some developmental delays), or the parents of the few children with a formal DD diagnosis did not agree to sign the consent form. Twenty-four teachers declined to participate due to workload (in the broader study, teachers were required to participate in interviews and complete a large number of questionnaires). In 47 kindergartens, the teacher and parents of a randomly chosen child diagnosed with DD agreed to participate. Before data collection, parents and teachers signed consent forms, and children’s informed assent was obtained orally.

At Time 1 (T1), teachers were interviewed using the Five-Minute Speech Sample (FMSS; Gottschalk & Gleser, 1969) in a quiet room with only the interviewer present. Next, an intelligence test, Mullen Scales of Early Learning (Mullen, 1995), was administered to the child participants in a quiet room, and teachers were asked to complete a demographic questionnaire. At T1 and Time 2 (T2), a year later, therapists working with target children (e.g. occupational therapists) were asked to anonymously complete two questionnaires to assess the functioning of the target child: Vineland Adaptive Behavior Scales (Sparrow, Balla, & Cicchetti, 1984) and the peer acceptance scale from the MacArthur Health and Behavior Questionnaire (Armstrong & Goldstein, 2003). The third and fourth authors collected all data at T1. A research assistant blind to information about the participants collected the data at T2. The Human Research Review Board of the participating university approved the study procedure (#347/15), as did the Chief Scientist of the Israel Ministry of Education (#8957).

Measures

Five-minute speech sample (FMSS; Gottschalk & Gleser, 1969)

The FMSS is widely used to assess adult–child relationships (e.g. between a parent and a child or a teacher and a child; for review, see Sher-Censor, 2015). Teachers were asked at T1 to speak for five uninterrupted minutes, giving their thoughts and feelings about the child and the teacher–child relationship, responding to the probe: ‘What is [child’s name] like and how do the two of you get along?’ After the teacher began speaking, the interviewer could offer one prompt during the speech sample if the teacher fell silent for 30 s or longer: ‘Please tell me more about what [child’s name] is like and how the two of you get along for a few more minutes.’ FMSS narratives were audiotaped, transcribed, and coded using the FMSS-Valence and the FMSS-Multidimensionality scales (Dolev, Sher-Censor, & Tal, 2021; Sher-Censor, Nahamias-Zlotolov, & Dolev, 2019), which were adapted from the Life Experiences FMSS Manual (Yates, Sher-Censor, & Grey, 2011).

The 3-point positive FMSS-Valence scale evaluated the affective tone of the narrative, based on its content. A score of ‘1’ reflected teachers’ overall negative view of the child and/or a mostly negative affect, such as distress, sadness, or frustration, with respect to the child and their relationship (e.g. ‘Our relationship, it’s an avoidant relationship. All the kids come to me for a hug, and he avoids me … .It is difficult for me … .You invest a lot in a child, but it leads to nothing, this is what frustrates me’). A score of ‘2’ was assigned when teachers’ narratives had a neutral or mixed affective tone (e.g. ‘In the beginning of the year, I was very stressed before interacting with him, but then I realized I have nothing to be afraid of, he got to know me, and I got to know him’). A score of ‘3’ was given when teachers expressed an overall positive view of the child and/or mostly positive affect, such as happiness, pride, and satisfaction, when describing the child and their relationship (e.g. ‘He has a lot of motivation, which is good … and so he succeeds to fit in the group, which makes us very happy … .I enjoy being with him’). FMSS-Valence was used in two previous studies on teachers’ perceptions of the group of children in their classroom. Interrater reliability in these studies was very good (ICC two-way random model, absolute agreement > .84). The scale was validated by showing that positive valence was associated with observations of higher classroom emotional support (Sher-Censor et al., 2019) and better classroom organization (Dolev et al., 2021).

The 4-point FMSS-Multidimensionality scale (Dolev et al., 2021) evaluated whether teachers portrayed the child in a multifaceted way by addressing four developmental domains: the social domain (e.g. ‘She plays nicely with other kids’), the emotional domain (e.g. ‘He has difficulties in expressing his emotions’), the cognitive/learning domain (e.g. ‘Her cognitive functioning and understanding are poor compared to the group’), and the behavioral domain (e.g. ‘I see that he works nicely with all the therapists’). Each domain referred to by the teacher received one point. Accordingly, scores ranged from ‘0’ (the teacher did not refer to any domain) to ‘4’ (the teacher referred to all four domains). The scale has been previously employed in research investigating how teachers perceive children in their classroom as a group. Interrater reliability in that study was very good (ICC two-way random model, absolute agreement = .97). The scale was validated by showing that addressing more developmental domains in the FMSS was associated with observations of higher emotional and instructional support in the classroom (Dolev et al., 2021).

The first and second authors, who did not take part in data collection and were blind to all other data about the participants, coded the FMSS. Twenty-eight percent were coded by both coders for reliability (ICC two-way random model, absolute agreement FMSS-Valence = .95; ICC two-way random model, absolute agreement FMSS-Multidimensionality = .88).

Vineland adaptive behavior scales – teacher form (VABS; Sparrow et al., 1984)

Therapists completed this well-known self-report at T1 and T2 to assess their experience of children’s adaptive behaviors. The VABS includes the domains of socialization (i.e. functioning in social situations), communication (i.e. how well children listen, understand, and express themselves through speech), daily living skills (i.e. performance of practical, everyday tasks of living that are appropriate in the school setting), and motor skills (i.e. fine and gross motor skills). Items are rated on a 3-point scale ranging from ‘0’ (never) to ‘2’ (usually). The VABS yields standardized scores with a mean of 100 and SD of 15. It has been used extensively in studies on children with DD (e.g. Green, Caplan, & Baker, 2014), including in Israel (e.g. Feniger-Schaal, Oppenheim, & Koren-Karie, 2019; Weisman et al., 2015) and in research of children with other diagnoses, such as cerebral palsy (Brossard-Racine et al., 2012).

Macarthur health and behavior questionnaire (HBQ; Armstrong & Goldstein, 2003)

The HBQ assesses children’s mental and physical health and academic functioning from the ages of four to eight. At T1 and T2, therapists completed the peer acceptance subscale consisting of 13 items rated on a 4-point scale, ranging from ‘1’ (not at all like the child) to ‘4’ (very much like the child’). Items include: ‘The child is liked by other children who seek him/her out for play’. The HBQ has been well validated, including among children with special needs (e.g. Lemery-Chalfant et al., 2007). Cronbach’s alpha in our study was excellent at T1 (alpha = .83) and also at T2 (alpha = .82).

Mullen scales of early learning (MSEL; Mullen, 1995)

The MSEL is an individually administered measure of cognitive functioning for kindergarteners. It assesses five domains: gross motor, visual reception, fine motor, receptive language, and expressive language. Children’s overall level of cognitive functioning at T1 was measured using the Early Learning Composite by converting the raw scores into derived scores based on children’s chronological age. The final IQ score derived from the sum of the standard scores in all five domains has a mean of 100 and a standard deviation of 15. The MSEL manual reports good internal, test–retest, interrater reliabilities, and validity measures (Mullen, 1995). This measurement is widely used in research, including in studies of children with DD (e.g. Feniger-Schaal et al., 2019) and other developmental diagnoses, such as ASD (e.g. Swineford, Guthrie, & Thurm, 2015) and cerebral palsy (e.g. Burns, King, & Spencer, 2013), and in research in Israel (e.g. Yaari et al., 2018).

Results

Preliminary analyses

The distribution of all variables was sufficiently normal to render parametric statistics valid (Afifi, Kotlerman, Ettner, & Cowan, 2007). Percentage of missing data ranged from 2.13% (n = 1) for children’s age at T1 to 23.40% (n = 11) for children's adaptive behavior at T2. Missing data were estimated using the expectation maximization (EM) algorithm in SPSS 20.0, as supported by Little’s (1988) MCAR test, χ2 (66) = 64.27, p = .537. The significance of the findings was the same with and without EM.

Descriptive statistics are shown in Table 1. Teachers’ education level, years of experience, and child age were associated with addressing more developmental domains in the FMSS (r = .31, p = .036, r = .35, p = .015, and r = .35, p = .016 respectively). In addition, child IQ was associated with higher adaptive behavior at T1 (r = .49, p = .001) and T2 (r = .31, p = .037). There were no other significant associations between these background variables and the study variables. Furthermore, duration of teacher–child acquaintance, teachers’ age, teachers’ marital status, and child gender were not significantly associated with any of the study variables (all p’s > .100). Hence, only teachers’ education level, teachers’ years of experience, child age, and child IQ were included as covariates in further analyses.

Table 1. Descriptive statistics (N = 47).

Parental variables	Mean	SD	Range
Teachers’ FMSS-Valence – T1	2.36	.67	1–3
Number of developmental domains addressed in teachers’ FMSS – T1	3.55	.58	2–4
Children’s adaptive behavior – T1	78.76	14.06	27–106
Children’s adaptive behavior – T2	80.64	11.43	51–105
Children’s peer acceptance – T1	3.25	.37	2.38–3.92
Children’s peer acceptance – T2	3.36	.38	2.58–4.00
Children’s IQ – T1	61.98	11.38	49–86

Note: T1 = Time 1; T2 = Time 2.

Valence and the number of developmental domains addressed in teachers’ FMSS and children’s development

As shown in Table 2, more positive valence of teachers’ FMSS was associated with more adaptive behavior of children at T2. There were no significant associations between FMSS-Valence and children’s adaptive behavior at T1 and peer acceptance at T1 and T2. Addressing more developmental domains in the FMSS was associated with higher peer acceptance at T2. There were no significant associations between the number of developmental domains addressed in the FMSS and children’s adaptive behavior at T1 and T2 and peer acceptance at T1.

Table 2. Inter-correlations of study variables (N = 47).

	1	2	3	4	5	6	7
1. Teachers’ FMSS-Valence – T1	–	−.08	.22	.48***	.16	.20	.21
2. Number of developmental domains addressed in teachers’ FMSS – T1		–	.12	.16	−.14	.42**	−.07
3. Children’s adaptive behavior – T1			–	.54***	.04	.37*	.49***
4. Children’s adaptive behavior – T2				–	−.11	.35*	.31*
5. Children’s peer acceptance – T1					–	.40**	−.06
6. Children’s peer acceptance – T2						–	.01
7. Children’s IQ –T1							–

Note: T1 = Time 1; T2 = Time 2.

* p < .05. **p < .01. ***p < .001.

Next, we conducted two regression analyses to examine the contribution of FMSS valence and the number of developmental domains addressed in the FMSS to changes in children’s adaptive behavior from T1 to T2. In the first regression model, children’s adaptive behavior at T2 served as the dependent variable. The first block of the model included children’s IQ, which, as noted above, was associated with children’s adaptive behavior at T2. The first block also included adaptive behavior at T1 to control for differences in child functioning at T2 that were better accounted for by the level of functioning at T1. FMSS-Valence was entered in the second block. As shown in Table 3, teachers’ FMSS valence contributed significantly to the explained variance of adaptive behavior at T2 above and beyond child IQ and adaptive behavior at T1. More positive valence predicted an increase in children’s adaptive behavior from T1 to T2.

Table 3. Regression results of children’s adaptive behavior at T2 on child IQ, children’s adaptive behavior at T1, and teachers’ FMSS-valence (N = 47).

Variable in regression	Children’s adaptive behavior – T2
	β (entry)	β (final)	ΔR^2
Block 1			.29***
Children’s IQ – T1	.06	.01
Children’s adaptive behavior –T1	.51***	.45**
Block 2			.13**
Teachers’ FMSS-Valence – T1		.38**
Total R^2	.42
Final model	F (3, 43) = 10.53***

Note: T1 = Time 1; T2 = Time 2.

** p < .01. *** p < .001.

In the second regression model, children’s peer acceptance at T2 served as the dependent variable. The first block of this model included teachers’ education level and years of experience, as well as children’s age, which, as noted above, were associated with the number of developmental domains addressed in teachers’ FMSS. The first block also included children’s peer acceptance at T1 to control for differences in child functioning at T2 that were better accounted for by the level of functioning at T1. The number of developmental domains in teachers’ FMSS was entered in the second block. As shown in Table 4, the number of developmental domains addressed in the FMSS contributed significantly to the explained variance of children's peer acceptance at T2 above and beyond teachers’ education level, teachers’ years of experience, children’s age, and peer acceptance at T1. Addressing more developmental domain in the FMSS predicted an increase in peer acceptance from T1 to T2.

Table 4. Regression results of children’s peer acceptance at T2 on child IQ, children’s peer acceptance at T1, and number of developmental domains addressed in teachers’ FMSS (N = 47).

Variable in regression	Children’s peer acceptance – T2
	β (entry)	β (final)	ΔR^2
Block 1			.27**
Teachers’ education – T1	.19	.13
Teachers’ years of experience – T1	−.002	−.007
Children’s age – T1	.008	.003
Children’s peer acceptance – T1	.44**	.46**
Block 2			.16**
Number of developmental domains addressed in teachers’ FMSS – T1		.38**
Total R^2	.43
Final model	F (5, 41) = 6.19***

Note: T1 = Time 1; T2 = Time 2.

** p < .01. *** p < .001.

Elucidations of valence and the number of developmental domains addressed in teachers’ FMSS and children’s development

Here we tap into the valence and the number of developmental domains presented in four teachers’ FMSS. As transcripts were long, only excerpts are provided herein. We first present two examples of the FMSS-Valence and relate them to changes in children’s adaptive behavior skills. We then give two examples of how children’s developmental domains were addressed in teachers’ FMSS and relate them to changes in peer acceptance of the children.

Sara talked in her FMSS about Joseph, a boy who was four years and ten months old. Throughout the FMSS, Sara’s descriptions of Joseph and their relationship were characterized by a positive valence:

Joseph is a lovely, adorable child and has a great sense of humor. At the beginning of the year, he was introverted and closed-up. He barely initiated eye contact, but now everything is different … . Now he initiates eye contact, shares his thoughts, he sings. I really love him. I feel connected to him. Every little progress, every little step, makes me so happy. It’s lovely when he comes up to you and smiles or gives you a high-five. I believe that he will be able to have some inclusion hours next year, and I’m so proud of him.

Sara referred to various positive features of Joseph’s behavior and progress. The positive valence was reflected in Sara’s belief in Joseph’s capabilities, and this may have contributed to the improvement Joseph showed in adaptive abilities: from a score of 80 at T1 to 91 at T2.

Rachel, a five-year-old girl, was in Daphne’s kindergarten classroom. Daphne described Rachel through a negative lens:

When I first saw Rachel, I was in a bit of a shock and thought: ‘What am I going to do with such a child? How will we manage?’ She is by far the lowest functioning child in my classroom. She looks weird since she is dysmorphic. Her eyes are small, with glasses, and her face with those teeth … . She is totally dependent on me and can’t do anything without me, and we’re working on that now.

Daphne continued to describe various behaviors and characteristics negatively and finished by saying: ‘She is not open to anything and doesn’t like to try new things’.

There were no positive descriptions in Daphne’s FMSS. Instead, every portrayal had some negative aspect. This negative viewpoint may have hindered Daphne’s ability to form a supportive relationship with Rachel and help her acquire adaptive behavior skills. As reported by the kindergarten therapist, Rachel’s score in adaptive behavior was 62 at T1 and descended to 56 at T2.

The following two examples illustrate how teachers addressed various developmental domains of the children and how this was related to changes in children’s peer acceptance. The first is from Noa, who talked about Dan, a boy aged four years and six months:

Dan has a few difficulties. He has substantial language difficulties but has a lot of motivation, loves to learn, and is an affectionate, cooperative child. Actually, he cooperates too much. We’re helping him to be a bit wilder, to allow himself to expand his own boundaries … . He usually plays by himself, and we’re also working on that … . Still, he is very empathic. If he sees a friend who’s having some difficulties, he goes and helps him or calls me to come and help the child. He also always goes by the rules. He respects the boundaries. As I said, his language level is like a toddler’s, and his speech isn’t very clear, but it has significantly improved this year. I believe that what he needs is more confidence. When he feels confident, he speaks much better.

While narrating Dan’s characteristics, Noa showed a multifaceted perception and referred to all four developmental domains, social (e.g. ‘He usually plays by himself’; ‘If he sees a friend … he goes and helps … ’), emotional (e.g. ‘affectionate, ‘empathic,’ ‘needs more confidence’), cognitive (e.g. ‘He has substantial language difficulties’) and behavioral (e.g. ‘He cooperates too much’; ‘He respects the boundaries’). We suggest Noa’s awareness of the four domains may have enabled her to address and support all of them when interacting with Dan. Such support, in turn, may have contributed to Dan’s social progress, as reported by the therapist. His peer acceptance score increased in the span of one year from 2.5 to 3.8, on a 4-point scale.

The final example is from Rebecca’s teacher, Gil, who focused throughout the FMSS on only one main characteristic, her young age and small features:

Rebecca is very young. She’s also petite, the ‘little girl’, the ‘baby’. She is the youngest child in the group … . She behaves in an infantile manner, even though I think she is a strong child. She is only three years and four months old … . Her appearance reflects her behavior. She’s a bit spoiled and still sucks on her thumb.

Gil addressed only the emotional (e.g. ‘I think she is a strong child’) and behavioral domains (e.g. ‘childish’, ‘spoiled’, ‘sucks her thumb’) and did not refer to Rebecca’s social or cognitive skills. Holding such a narrow view may have hampered Gil’s ability to help Rebecca in these domains, and this, in turn, may have hindered a positive change in peer acceptance, as reflected in Rebecca’s stagnant low score for peer acceptance (1.3 at T1 and 1.15 at T2).

Discussion

The prevalence of DD diagnosis among young children is relatively high. It is estimated to affect about 4% of the children in the US (Zablotsky et al., 2019) and up to 10% of the children in Israel (Israel Ministry of Health, 2022). Yet to the best of our knowledge, research on the relationships these children form with their teachers is scarce (for exceptions, see Blacher, 2009; McIntyre et al., 2006; Rhoad-Drogalis et al., 2018). Our study joins this small body of work and suggests a high quality of teacher–child relationships may contribute to decreased gaps in the adaptive behavior (i.e. in socialization, communication, motor, and daily living skills) of children with DD and to increased peer acceptance.

The study joins prior research suggesting caregivers’ narrated perceptions of children are meaningful for understanding children’s development and well-being. This research has generally focused on the narrated perceptions of parents of children with typical development (e.g. Sher-Censor et al., 2016) or with developmental diagnoses other than DD (e.g. Sher-Censor, Dolev, Said, Baransi, & Amara, 2017) and on the perceptions of teachers of children with typical development (e.g. Dolev et al., 2021). Our results extend previous findings to the context of teachers and children with DD.

We focused on two aspects of teachers’ narratives, their valence, and the number of developmental domains they addressed. These features were not intercorrelated. This suggests that when teachers describe a child across multiple developmental domains, they may not necessarily express a positive tone. The valence of teachers’ narratives and the number of developmental domains they addressed were also not associated with children’s IQ, or with therapists’ reports of children’s adaptive behavior and peer acceptance at T1. Thus, it seems that children’s functioning did not color teachers’ narrated perceptions. Put differently, teachers who narrated about children with lower functioning did not necessary convey less positive valence, nor did they address more developmental domains than teachers narrating about children with better functioning at T1. However, as expected, narrative features were associated with changes in children’s development. Previous research led us to expect that both narrative features would be related to changes in children’s adaptive behavior and peer acceptance. Nonetheless, we discovered specific associations. Positive valence in teachers’ narratives was associated with increased adaptive behavior, while a higher number of developmental domains addressed in narratives was associated with subsequent improvements in peer acceptance.

Positive valence in teachers’ narrations may have indicated more positive views of a child and the teacher–child relationship, and this may have fostered teachers’ involvement in and support of children’s learning. For example, positive perceptions may have contributed to teachers’ more accurate perceptions of children’s strengths and specific needs, and this, in turn, may have facilitated their ability to plan an Individualized Education Program (IEP) that better accommodates the child’s specific needs. Teachers’ positive perceptions may also have fostered children’s sense of security and self-esteem. Previous studies have indicated that when children perceive their teachers as supportive, they develop higher academic self-esteem and become more engaged in learning activities. They are more attentive, responsive to directions, and motivated to invest effort in learning, leading to a greater number of learning opportunities (Guay, Stupnisky, Boivin, Japel, & Dionne, 2019; Losh, Eisenhower, & Blacher, 2022; Susperreguy, Davis-Kean, Duckworth, & Chen, 2018). In the context of children with DD, this positive process may cultivate improved adaptive behaviors. Supporting this interpretation, a recent study of young children with ASD found teachers’ questionnaire-based reports of a more positive relationship with children in their classroom predicted children’s higher levels of engagement in learning activities and better social functioning (Losh & Blacher, 2023).

The second aspect of the narratives we assessed captured the extent to which teachers’ narratives were multifaceted and addressed four domains: social, emotional, cognition, and behavioral. We found that addressing a greater number of developmental domains in the narratives predicted improvement in children’s peer acceptance. A previous study examined the narratives of teachers of typically developing kindergarteners about their experiences with the group of children in their class. The study revealed that teachers who addressed more developmental domains provided higher emotional and instructional support to the children (Dolev et al., 2021). Another study on children with DD found that competence in each developmental domain was associated with enhanced social adjustment (Baurain, Nader-Grosbois, & Dionne, 2013). Our study complements these previous works. Addressing more developmental domains when narrating about the child may indicate teachers’ awareness of the child’s needs across all four domains, thereby facilitating their support for these diverse needs. This, in turn, may have fostered the children’s social adjustment. Future studies with larger samples could explore these potential meditating processes. In particular they may examine the combined profiles of valence and each developmental domain addressed in the narratives, along with their roles in the developmental trajectories of children with DD.

Our study found that kindergarten teachers with higher levels of educations and greater teaching experience tended to address more developmental domains in their narratives. It seems that educators with advanced educational backgrounds may hold a more complex perceptions of children, leading them to reference a wider range of developmental domains in their narratives. We also found that teachers who narrated regarding older children tended to address more developmental domains. It could be that as children approach the age of first grade, teachers tend to consider a wider range of developmental abilities that are necessary for a successful adaptation to school. Given that this was the first study to explore the multifaceted nature of teachers’ narratives within the context of special education kindergartens, future studies should explore these associations further.

The longitudinal design and the use of multi-respondents (teachers, trained coders, kindergarten therapists) and multi-methods (i.e. interviews, standardized tests, self-reports) are strengths of this study. Nevertheless, three caveats warrant consideration and may guide future research. First, the small sample size may have limited the detection of small effects. This could be one of the reasons for the absence of significant associations between the valence of teachers’ narratives and changes in children’s peer acceptance and between the number of developmental domains teachers addressed and changes in children’s adaptive behavior. Further research employing a larger sample size is warranted to validate our findings. Such research may also examine more nuanced changes in specific aspects of children’s adaptive behavior (i.e. socialization, communication, daily living skills, and motor skills) and study differences across children with different types of DD. Moreover, such research may be able to explore the potential mediating and moderating processes in the links between teachers’ narrated perceptions and children’s development. These processes may involve teachers’ support of children’s learning, children’s sense of security, self-esteem, learning engagement, attention span, and challenges in sensory processing and self-regulation.

Second, there was some variation in the treatment plans for children. The small sample size prevented the possibility of including this variability in statistical analyses. Future research should examine the associations between specific children's treatment plans, teachers’ perceptions, and those children's development. Third, teachers’ narratives were collected only at T1 and not at T2. Longitudinal studies employing cross-lagged models are needed to examine the likely bidirectional relations between teachers’ narrated perceptions of children and children’s development.

Implications for special education kindergartens in Israel

Kindergarten is a critical educational setting in the transition from preschool to first grade. For children with DD, it is a window of time which offers an opportunity to decrease developmental gaps and acquire the academic and socio-emotional skills necessary for successful adaptation to school (McIntyre et al., 2006). Despite its limitations, the study contributes to our understanding of the role of the educational setting, specifically, the perceptions teachers form of children with DD, in the development of those children. It is important to note that the study does not suggest teachers and/or educational settings are to be blamed for children’s lack of developmental gains, nor does it posit that teachers’ perceptions singularly shape children's developmental trajectories. Instead, it underscores that teachers’ positive and multifaceted perceptions may represent two contributory factors in fostering children’s development.

Our findings suggest the construction of positive and multifaceted perceptions is attainable in the context of special education kindergartens in Israel. The policy of the Israeli Ministry of Education (2022) may facilitate such perceptions. First, special education kindergartens have a small setting, with a maximum of 14 children; thus, teachers have an opportunity to create a close relationship with each child and learn about that child’s strengths and challenges in varied developmental domains, which may help them perceive children in more positive and multifaceted ways. Second, teachers are expected to complete an Individualized Education Program (IEP) for each child and a Classroom Learning Program and to address all areas of development and learning in both programs; hence, teachers are expected to consider different aspects of child functioning. Third, teachers are given guidelines to promote the participation and integration of children by cultivating their social skills and strengthening their relationships with other children and the community. These guidelines may also encourage the formation of multifaceted perceptions of children.

Yet our study shows that although current policy may promote the formation of positive and multifaceted perceptions, teachers vary in their perceptions, and this variability is linked to changes in the development of children with DD. This research thus offers three important recommendations for policy makers and practitioners. First, the findings suggest forming a high-quality teacher–child relationship, especially constructing positive and multifaceted perceptions of the child, may be a steppingstone to helping children with DD close their developmental gaps. Therefore, pedagogical training and supervision of special education kindergarten teachers in Israel should allocate more time and effort to enhance teachers’ awareness regarding the potential influence of their positive perceptions of children and the importance of considering varied developmental domains. Pedagogical training can help teachers broaden their multifaceted perceptions. Targeted training may emphasize the skills associated with each of the four developmental domains, their development across early childhood among typically developing children, as well as children with DD, and their impact on later child well-being and school success. Helping teachers adopt a positive valence towards children in kindergarten requires ongoing supervision to support teachers’ understanding of the child’s point of view and the motivations that underlie children’s challenging behaviors.

Second, we speculate that one of the ways to support the construction of positive and multifaceted perceptions is to allocate more time in the kindergarten curriculum for one-on-one teacher–child interactions. More frequent and intimate interactions are likely to foster teachers’ positive and broad understanding of children with DD.

Third, our results point to the value of assessing teachers’ narrated perceptions. Our study joins previous research suggesting the FMSS procedure can capture meaningful aspects of teacher–child relationships (e.g. Daley et al., 2005). As the FMSS administration and coding are relatively cost and time efficient, this study supports the utility of the FMSS procedure in teachers’ training and supervision, as well as in future educational research.

Acknowledgment

This article is based in part on master’s theses by Michal Shalem Gan-Or and Esther Zach. We express our sincere gratitude to the teachers and families who participated in this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by Oranim College of Education.

Notes on contributors

Efrat Sher-Censor

Efrat Sher-Censor is an Associate Professor at the Interdisciplinary MA and PhD programs in Child Development, the School of Psychological Sciences, University of Haifa, Israel. Her major research interests include the representations that children, parents and caregivers construct about their relationships. She explores how representations shape parent-child and caregiver-child interactions and predict children's adjustment.

Smadar Dolev

Smadar Dolev is a Senior Lecturer and a tenured faculty member at Oranim College for Education in Israel. She served as the head of the graduate and undergraduate programs for Early Childhood Education at Oranim College. Her primary research interests include parent/teacher-child relationships and the interplay between representations, behavior, and child development, with a focus on populations of early childhood-aged children with special needs.

Michal Shalem Gan-Or

Michal Shalem Gan-Or graduated with a master's degree in developmental psychology. She is a specialized therapist. She works as a psychologist in the public service, consulting kindergartens around issues of classroom climate and children with behavioral and emotional difficulties and developmental challenges. She treats and diagnoses children with communication difficulties, difficulties in the parent-child relationship, and eating and sleeping difficulties.

Esther Zach

Esther Zach holds a B.Ed in Special Education from Oranim College of Education and an MA in Child Development from the University of Haifa. With 14 years of experience managing a special education kindergarten, she is directing a regional support center named “Mati'ah Migido” for the past five years. Esther integrates years of hands-on experience working with special children and parents with her academic research. Her research focuses on emotional availability and its significance in the relationship systems within special education kindergartens.