Teachers’ relational competence: perceptions of teachers and students with and without ADHD and ASD

ABSTRACT

This study examined whether teachers’ professional development of their relational competence with students with attention deficit hyperactivity disorder (ADHD) and/or autism spectrum disorder (ASD) modifies teachers’ and students’ perceptions of their teacher-student relationships (TSR). Participants comprised teachers (n = 33) and students (n = 232) from two elementary schools: one intervention school (InS) and one control school (CoS). InS teachers reported significant TSR improvements, regardless of student group or gender (p = .03). Among InS students, significant results were driven by female neurodiverse (ND) students and neurotypical (NT) male students (p = .03). Nevertheless, positive effects were solely observed among ND female students, while NT male students, conversely, reported decreased TSR during follow-up tests. No significant effects were found at the CoS irrespective of teacher or student ratings. The findings suggest that enhancing teachers’ understanding of relational competence concerning ND students will not only improve their own perceptions of their TSR but also those of ND female students. Nonetheless, directing teachers’ focus towards one student group (ND students) risks diminishing teachers’ attention towards other student groups, potentially explaining the poorer follow-up results among NT boys. The finding warrants further investigation, as it indicates a challenge for teachers to establish sufficient relational engagement with all students.

KEYWORDS:

• ADHD
• autism
• neurodevelopmental disorders
• relational competence
• teacher-student relationship

Like many other countries of the Organization for Economic Co-operation and Development (OECD), Sweden has formally introduced inclusive educational settings for neurodivergent (ND) students. Children with average to high intellectual abilities diagnosed with attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD), and other neurodevelopmental conditions (American Psychiatric Association 2013) are almost exclusively taught in regular mainstream classrooms by regular teachers. However, difficulties in educational settings are common for ND students (Bolourian, Stavropoulos, and Blacher 2019; Lasko 2020), who exhibit high rates of school absenteeism (Fleming et al. 2017; May et al. 2021), often experiencing mental health problems (Kalb, Stuart, and Vasa 2019; Lawrence et al. 2019), and experience multiple challenges in other important areas of functioning (Bölte et al. 2018, 2019).

Teacher-student relationship

Comprehensive research on the nature of social relationships between teachers and students indicates that high-quality relationships are a success factor for enabling students’ social learning and academic outcomes in school settings (e.g. Allen et al. 2013; Ansari, Hofkens, and Pianta 2020; Nurmi 2012; Pianta 1999). Archambault et al. (2013) and Allen et al. (2013) demonstrated that high-quality teacher-student relationships (TSR), where teachers sensitively respond to students’ needs, are both associated with student engagement in school and a positive classroom climate. Allen et al. (2013) also reported that the TSR quality is associated with classroom size, with classes containing fewer students often exhibiting better TSRs. In addition, teachers’ relationships with female students are generally of higher quality compared to relationships between teachers and their male students (Goldie and O’Connor 2021; Tennant et al. 2015). However, there is usually more social interaction between teachers and male students as male students tend to both draw and receive more attention from teachers. These relationships are often characterised by high mutual attention and are generally less positively loaded than relationships between teachers and their female students (Eliasson, Sørensen, and Karlsson 2016).

Relationships between teachers and neurodivergent students

Despite comprehensive research confirming positive associations between high-quality TSR and student outcomes among neurotypical (NT) students without any diagnosis, the literature on TSR regarding neurodiverse (ND) students (which, in this article, refers to students with ADHD and/or ASD) is still limited (Eisenhower, Blacher, and Bush 2015; Ewe, 2019). This is unfortunate, as good TSR can be assumed to be of greater importance for ND students compared to NT students (Blacher et al. 2014; Portilla et al. 2014).

Studies have consistently found that relationships between teachers and students with ADHD are usually significantly more strained and less emotionally close compared to TSR in general (e.g. Ewe, 2019). Furthermore, teachers seem to perceive their relationships with students with ADHD as more conflictual and less cooperative (Cook and Cameron 2010; Rogers et al. 2015). Students’ externalising behaviour tends to be a primary promoting factor for troubled TSR (Blacher et al. 2014; Caplan et al. 2016; Zendarski et al. 2020), as it negatively affects teachers’ emotions, and in the long run, risk hampering their interaction with these students (Nucifora and Walker 2021). Santos et al. (2016) confirmed this by stating that teachers’ interactions with students with ADHD and externalising behaviour comprise higher amounts of reprimands, which is concerning, as formal and informal labelling risks exacerbate externalising behaviour and thereby hamper the TSR even more (Gwernan-Jones et al. 2016). Results from studies conducted by Roorda et al. (2021) and Rushton et al. (2020) revealed a connection between high levels of TSR conflict and decreased school engagement among these students, further complicating the situation. However, at the same time, Mejia and Höglund’s (2016) research reveal strained TSR between teachers and introverted students with ADHD. Thus, being diagnosed with ADHD appears to present an increased risk of strained TSR, regardless of extroverted or introverted behaviour. This result is supported by Eisenhower et al. (2015) and Blacher et al. (2014), who indicated more conflict in relationships between teachers and students with ASD compared to relationships between teachers and NT students.

Eisenhower et al. (2015) state that existing research on relationships between teachers and students with ASD tends to focus on teachers’, rather than students’, perceptions of their TSR. Students’ perspective can thus be considered a gap in need of further examination. Filling this gap can be considered particularly important as existing research focusing on relationships between teachers and NT students shows that the coherence between teachers’ and students’ perceptions of their TSR tends to be moderate at best (Koomen and Jellesma 2015). More importantly, the existing scarce literature has predominantly focused on describing the quality of the relationships between teachers and ND students, rather than providing the means to improve them, which would increase students’ perceptions of secure and trustful TSR (Eisenhower, Blacher, and Bush 2015; Ewe, 2019). Roorda et al. (2021) and Feldman et al. (2019) highlight the importance of targeted interventions to improve the TSR between teachers and ND students. Therefore, there is an urgent need to investigate and develop measures promoting positive TSR that are significant for ND students, not only reported by teachers but also by students themselves.

Gender differences

Neurodevelopmental symptoms tend to present differently when viewed from a gendered perspective (Bauermeister et al. 2007; Moyse and Porter 2015). Girls with ASD often hide their needs and therefore risk receiving inadequate teacher support (Moyse and Porter 2015; Solomon et al. 2012). The situation seems to be identical for girls with ADHD, and Fugate and Gentry (2016) refer to this as the ‘masking effect’ (p. 83). Moreover, a systematic review conducted by Eccleston et al. (2019) demonstrates that girls with ADHD tend to strive to internalise expectations from others to a greater extent compared to boys with the same diagnosis.

A study by Cook and Cameron (2010) indicated that teachers perceived relationships with boys with ADHD as containing higher levels of conflict compared to their relationships with girls with the same diagnosis. However, results from Lawrence et al. (2017) showed teachers accepting the behaviour of boys with ADHD to a greater extent compared to the behaviour of girls with the same diagnosis. Rather, the behaviour of girls seemed to be problematised to a greater extent than those of boys (Lawrence et al. (2017).

A study conducted by Roorda et al. (2021) revealed that male students with ASD placed in special education classes perceived their TSR as more conflictual and less engaging compared to boys with the same diagnosis placed in general school settings. These high feelings of conflict seem to be consistent with the perceptions of students with ADHD. However, some gender differences are identified. While girls with ADHD seem to perceive higher levels of teacher rejection in comparison with their NT female peers, boys with ADHD seem to perceive their TSR in the same way as boys in general do (Rogers et al. 2015).

Teachers’ professional learning

In a previous controlled pilot study, the first author examined whether teachers’ relational competence regarding students in general, could be increased through a professional development program (Ewe, 2020). Nineteen lead teachers¹ participated in a video-based intervention comprising a lecture on relational competence, followed by a presentation of the relational competence model (RCM), a tool for analysing teachers’ relational competence (Aspelin and Jönsson 2019). The intervention ended with the participants collaboratively analysing a pedagogical video sequence using the RCM model. The feasibility and effect of the intervention on teachers’ sense of understanding relational competence and the ability to notice and verbalise non-verbal communication were measured pre- and post-test using a self-report online survey. The findings indicated that the intervention was feasible and endorsed its preliminary efficacy for an improved understanding of relational competence and skills to verbalise non-verbal communication. In a subsequent larger controlled study (Ewe & Aspelin, 2022), qualitative changes in teachers’ understanding of relational competence were examined pre and post a similar teacher intervention. However, while the intervention in the pilot focused on students in general, the intervention in the main study specifically focused on teachers understanding of relational competence in relation to students with ADHD. Teachers from two schools participated in the study, with one school establishing the intervention group and the other forming the control. The results indicated that the intervention not only improved teachers’ understanding of how to perceive and respond to students’ non-verbal signals sensitively but also how to oscillate between students’ signals and their own emotions and reactions.

Our previous research outlined above indicates that teachers’ experience improved their understanding of relational competence with regard to ND students; however, following relational competence intervention, whether the effects can be generalised to the classroom situation and whether the perception of an improved teacher-student relationship is shared by ND students remain unknown. Thus, it is unclear whether intervention effects are limited to a theoretical or unidirectional level that does not make a difference in practice or whether they are perceived by ND students as making a relational difference. Hiebert and Stigler (2000) argued that a major obstacle to teachers’ professional development might be teachers’ belief that they are implementing acquired novel knowledge and understanding in practice while in actuality retaining their traditional patterns or not reaching out to students. This result is supported by Soine and Lumpe (2014), who found limited evidence for correlations between teachers’ professional development and impacts on students’ learning outcomes. However, the research results of Antoniou and Kyriakides (2013) demonstrated positive improvements in both teachers’ skills and students’ achievement after interventions related to teachers’ professional development.

Additionally, Holmqvist (2011) states that adding a theoretical framework to professional development interventions increases the possibility of changes in practice, as increased theoretical knowledge entails a change in perspective. Fielding (2001) points out that proxy reports collected from teachers, guardians, and researchers often possess interpretive precedence over students’ primary perspectives, risking misinterpretation of students’ perspectives or disregarding them altogether. Therefore, to investigate whether enhanced theoretical relational understanding also induces changes in practice and student experiences, this study assessed whether professional development of teachers’ relational competence improved not only teachers’ but also students’ perceptions of TSR. For this purpose, we are now following up with teachers and their ND and NT students on experienced TSR after teachers had experienced professional development for relational competence (Ewe & Aspelin, 2022). We hypothesised that professional development for relational competence in teachers would not only improve teachers’ perceptions of relational quality but also report moderate and even significant improvements experienced by students.

Method

Design and context

This study was approved by the Swedish Ethics Review Board (no: 2019-03533). We conducted a controlled ex post facto study to measure teachers’ and students’ perceptions of their TSR on three occasions (pre- and post-test and follow-up assessments) during an ongoing professional development intervention study in the spring of 2020 (February – May). The study is described above, in Appendix A, and elsewhere (Ewe & Aspelin, 2022). All participating teachers in the current study have also taken part in the intervention study. To measure if the intervention led to changes in teachers’ as well as students’ perceptions of their TSR, teachers and students at both schools were asked to rate their perceptions of their TSR thrice during one semester. These ratings were obtained pre, post and follow-up to the teacher intervention study (Table 1) with the pre-test administered one week prior to the intervention study’s pre-test at the intervention school (InS), and the post-test conducted one week following the intervention study’s post-test. Finally, the follow-up test was performed approximately two months past the post-test.

Table 1. The date on which questionnaires were sent to the schools.

School	Pre-test	Post-test	Follow-up-test
InS	February 3, 2020	March 9, 2020	May 4, 2020
CoS	March 2, 2020	March 18, 2020	May 4, 2020

Assessments were conducted using the teacher and student version of the ‘involvement’ section of the Teacher as a Social Context scale (TASC) as the primary outcome (Belmont et al. 1992; Wellborn et al. 1992). The questionnaire was administered as a web survey, wherein the questionnaires were sent to teachers and students via their respective school emails. Because of school organisational issues, assessments at the intervention school and the control school (CoS) were not fully synchronised. The CoS started their pre-test ratings three weeks after the InS. In addition, while there was a five-week period between the pre- and post-test ratings sent out to the InS, the same period comprised two weeks for the CoS. This difference was attributed to the beginning of the Coronavirus disease 2019 (COVID-19) pandemic. As it was uncertain whether the schools would remain open, the decision was made to send the post-test ratings to the CoS two weeks after the pre-test ratings.

Participants

A total of 33 in-service teachers and 232 students aged 10 to 16 years (Table 2) from two elementary schools of two medium-sized urban municipalities in southern Sweden participated in the study.

Table 2. Distribution of students in the different grades.

Students age in year	Grade	InS (total)	InS (ND students)	CoS (total)	CoS (ND students)
10	4	-	-	25	1
11	5	2	2	34	5
12	6	11	2	15	2
13	7	31	6	29	1
14	8	24	3	24	5
15	9	7	4	30	4
	Average grade	7.31	7.29	6.52	6.89
	Median value	7	7	7	7.5

The average class grade of the participating students of the InS was somewhat higher (M = 7.3)² than the average class grade of the CoS students (M = 6.5).³ In the total sample of 232 students, 35 (15%) had parent-reported⁴ diagnoses of ADHD and/or ASD or were showing behaviours indicative of these diagnoses. Participants of the InS comprised 13 teachers and 75 students (40 female and 35 male), of whom 17 (23%) were ND students (8 female and 9 male). Participants in the CoS consisted of 20 teachers and 157 students (81 female and 76 male), of whom 18 (11%) were ND students (7 female and 11 male). The number of students in each class was approximately 20 in the InS compared to 27 in the CoS.

Procedure

Teachers at both schools participated in a professional development intervention study (Ewe & Aspelin, 2022). Please see Appendix A for a description of the intervention performed. These teachers, along with their students, were asked to also participate in the present study. The teachers were asked to express their interest in participating in the study via Google forms. Informed consent was then collected from teachers at both schools. Forms for legal caregivers and related information were sent to students’ homes. Two parent meetings were held at the InS to disseminate oral information and collect consent forms. Shortly before the pre-test assessment, the first author visited both the InS and CoS to help students and teachers understand how to rate the Teacher as a social Context (TASC) scale (Belmont et al. 1992; Wellborn et al. 1992).

Instrument

Adapted subscales of both the teacher and student versions of the TASC scale were used as the primary outcome measure, which captures TSRs in terms of involvement, namely, affection, attunement, dedication of resources, and dependability (Belmont et al. 1992; Wellborn et al. 1992). Earlier use of the instrument (Skinner and Belmont 1993) provided evidence of reciprocal relationships between teacher involvement and student engagement, where teachers’ individual involvement with students had the strongest impact on students’ perception of not only relatedness but also competence and self-determination. Contrarily, a lack of teacher involvement resulted in higher student perceptions of coerciveness and less consistency. Teachers tend to respond with more involvement in relation to students with high behavioural commitment, while the lack or strained levels of student behavioural involvement tends to negatively affect teacher responses. The scale has been successfully adapted and used in Spanish and Indonesian school contexts (Iglesias-García et al. 2019; Maulana et al. 2016). Ahn et al. (2019) highlights the wide use of both verbatim and adapted versions of the short form of the TASC manual (TASC) (Belmont et al. 1992; Wellborn et al. 1992) among researchers.

The original involvement subscales in both teacher and student versions include 14 items operationalised on a 4-point Likert scale. Some adaptions in the instrument were made because of the focus of the study. First, the Likert scale in both the teacher and student questionnaires was extended to contain a 6-point Likert scale (with 1 being disagree and 6 being strongly agree) instead of the original 4-point scale. This procedure was performed to increase the sensitivity of the instrument. Further, the items in the manuals were translated from English to Swedish. Some changes were made to the student version to gauge if a teacher intervention focusing teachers’ relational competence in relation to ND students could affect students’ perceptions of being understood and listened to by their teacher. This was executed based on the results in (Ewe, 2019), which showed that students with ADHD perceived higher levels of teacher rejection compared to their NT peers. The following changes were made: One item was added to the affection part in the student version [‘I have a good relationship with my teacher.’] since this was one of the prominent aspects of the study focus. Further, the dedication of resources section was renamed in the student version to ‘Communication’ and the original items [‘My teacher spends time with me’; ‘My teacher talks with me’] were remoulded to correspond with the study focus [‘I understand what my teacher says when he/she talks to me.’; ‘My teacher understands me when I talk to him/her.’]. One more item was added to this part [‘My teacher listens to me.’]. Further, one question was added to the dependability part [‘I can talk to my teacher when I am not feeling well.’] to capture if students felt secure and trusted their teachers enough to talk to them about sensitive issues.

The changes resulted in a total of three added items, which resulted in a total of 17 items in the student version. After the adaptions made, Cronbach’s alpha was r_α = .79 in the affection subscale, r_α = .75 in the attunement subscale, r_α = .86 in the communication subscale, and r_α = .82 in the dependability subscale which indicates high internal consistency. No changes were made to the teacher version, which consisted of 14 items, as the original version was considered to measure an eventual change in teachers’ perceptions of their TSR. The changes in the student version resulted in a maximum total score of 102 for students compared to 84 for teachers. At pre-test, Cronbach’s alpha was r_α = .91 in the student sample and r_α =.79 in the teacher sample.

The conversion of the items in the student version is not expected to affect the results, as the study intends to identify changes in total scores on the different measurement occasions for students and teachers rather than analyse the comparison between teachers and students. Thus, higher ratings indicated a more positive perception of TSR in this study. However, as the TASC manual includes both positive and negative items, all responses of negative items were recoded to positive scores (i.e. for the statement ‘My teacher just doesn’t understand me, ’ the value 1 was reversed to 6; 2 was reversed to 5 and so on).

The teacher and student versions of the TASC were collected and administered via the survey tool, Sunet Survey. All the students who participated in the study rated their relationships with one teacher. Simultaneously, each teacher rated their relationship with up to five students.

Data analysis

Primary outcome data (survey responses of the TASC) were collected online via the Sunet Survey before being manually aggregated and compiled in Excel and further analysed using IBM SPSS Statistics (version 27). Descriptive TASC data of student and teacher ratings of relational quality across student types and genders were computed for pre- and post-test and follow-up assessments of the InS and CoS. Potential professional development intervention main and interaction effects of the type of school (InS vs. CoS) were investigated by student group (ND vs. NT) and student gender. Student and teacher TASC data were analysed using a two three-way analysis of variance (ANOVA) for repeated measures (MANOVA), applying an alpha level of 5% and reporting effect sizes (eta²). Due to bias risk due to unevenly distributed missing data over assessment points between the InS and CoS, we applied intention-to-treat analyses using the last observation carried forward imputation. Post-hoc power analysis in G-Power (3.1.9.7) indicated the adequacy of 1-beta to detect differences in a general linear model given the sample size and alpha was .70 for a small effect (f = .20) and > .99 for both medium (f = .50) and large (f = .80) effects for students’ TASC ratings and .10, .44, and .90 for small, medium, and large effects for teachers’ TASC ratings.

Results

Participating students’ response rate for the TASC (Appendix B) was generally high in both schools in the pre-test (InS = 89.3%; CoS = 81.5%), although student participation decreased from pre- to post-test and follow-up assessment (InS = 81.3% [post], 77.3% [follow-up]; CoS = 62.4%, [post], 57.3% [follow-up]).

Teacher ratings (Appendix C) were similarly high in both schools in the pre-test (InS = 87.1%; CoS = 84.4%) and dropped from pre- to post-test and follow-up assessment (InS = 71.0% [post], 72.6% [follow-up]; CoS = 77.5% [post], 46.1% [follow-up]). Descriptive TASC results for teachers and students from pre- to post-test and follow-up assessment are presented in Tables 3 and 4.

Table 3. Total TASC scores of students of the intervention and Control schools at pre- and post-test and follow-up assessments.

Students (max 102)	Pre	N	Post	N	Follow-up	N
InS^6	Mean (%)		Mean (%)		Mean (%)
ND^7 (all)	77.6 (76)	17	79.1 (78)	15	79.1 (78)	16
ND male	81.6 (80)	9	83.1 (81)	7	81.4 (80)	8
ND female	73.3 (72)	8	74.6 (73)	8	76.4 (75)	8
NT^8 (all)	77.6 (76)	50	78.1 (77)	46	75.2 (74)	42
NT male	77.2 (76)	20	75.4 (74)	20	72.0 (71)	13
NT female	77.9 (76)	30	79.9 (78)	26	77.5 (76)	29
CoS^9	Mean (%)		Mean (%)		Mean (%)
ND (all)	82.0 (80)	13	82.2 (81)	6	82.9 (81)	6
ND male	8.3 (79)	7	81.6 (80)	3	81.1 (80)	4
ND female	83.5 (82)	6	83.0 (81)	3	85.2 (83)	2
NT (all)	8.1 (79)	115	78.5 (77)	92	79.0 (77)	84
NT male	82.3 (81)	59	81.8 (80)	42	82.0 (80)	40
NT female	78.0 (76)	56	75.1 (74)	50	76.1 (75)	45

Table 4. Total TASC scores of teachers of the intervention and Control schools at pre- and post-test and follow-up assessments.

Teachers (max 84)	Pre	N	Post	N	Follow-up	N
InS^10	Mean (%)		Mean (%)		Mean (%)
ND^11 (all)	64.6 (77)	17	64.8 (75)	16	65.8 (78)	15
ND male	65.2 (78)	9	63.7 (67)	9	67.8 (81)	8
ND female	63.9 (76)	8	61.8 (74)	7	63.6 (76)	7
NT^12 (all)	62.8 (75)	37	65.1 (78)	28	64.6 (77)	30
NT male	60.1 (71)	14	61.9 (74)	10	62.3 (74)	10
NT female	64.1 (76)	23	66.7 (79)	18	65.8 (78)	20
CoS^13	Mean (%)		Mean (%)		Mean (%)
ND (all)	60.1 (72)	16	58.9 (70)	14	58.8 (70)	6
ND boys	60.6 (72)	10	58.7 (70)	9	58.5 (70)	5
ND girls	59.3 (71)	6	59.3 (71)	5	59.3 (71)	1
NT (all)	63.4 (75)	59	62.4 (74)	54	61.8 (74)	33
NT boys	61.6 (73)	26	61.6 (73)	24	60.9 (72)	16
NT girls	64.9 (77)	33	63.1 (75)	30	62.5 (74)	17

There was a significant interaction effect of the type of school (InS) × student group × gender on the total TASC scores of students (F_1/203 = 3.2, p = .03) with a small effect (eta² = .02). The effect was driven by both female ND students reporting better TSR (TASC = 73.2 [pre], 74.6 [post], 76.4 [follow-up]), and NT boys reporting poorer relational quality with teachers (TASC = 77.2 [pre], 75.4 [post], 71.9 [follow-up]). Thus, significant positive effects were only found among ND female students. Nonetheless, when comparing InS student ratings of their TSR between the pre-, post-, and follow-up assessments, a general improvement was observed for ND students, while the trend for NT students was the opposite (see Figure 1).

Figure 1. Mean of the total TASC scores for students of the InS from pre-test to follow-up assessments.

No significant differences were observed between ND male students and NT female students at InS. However, a descriptive analysis revealed that male ND students comprised the student group with the highest TSR ratings at the InS on all measurement occasions. No main or interaction effect were significant for the CoS student ratings.

Among teachers, there was a significant main effect of type of school (InS) (F_1/32 = 3.2, p = .03), with a small effect size (eta² = .02) (Figure 2).

Figure 2. Mean of the total TASC scores in teacher ratings Comparing InS and CoS from pre-test to post-test and follow-up assessments.

Teachers in the InS reported enhanced relations with their ND and NT students (TASC 63.2 [pre], 64.4 [post], 64.9 [follow-up]), while there was no improvement in the CoS (TASC = 62.7 [pre], 61.6 [post), 61.1 [follow-up]). No main or interaction effects were significant for the teacher ratings at CoS (Figure 3).

Figure 3. Mean of the TASC scores Comparing teacher ratings of ND and NT students from pre-test to post-test and follow-up assessments.

Discussion

This study investigated whether professional development of relational competence in teachers would not only improve teachers’ perceptions of their TSRs but also generate changes in students’ perceptions of their relationship with their teacher. We expected an overall significant improvement in InS teachers’ perceptions of relational quality, along with a moderate but significant improvement experienced by InS students. The hypothesis regarding InS teachers’ ratings was fulfilled (p = .03) across the student groups and across genders while significant changes in student ratings (p = .03) were only driven by ND female students and NT male students where the former reported improved TSR, while the latter reported reduced TSR.

The results regarding CoS teachers’ student ratings were, unlike those of InS teachers, consistent with that reported in earlier research (Blacher et al. 2014; Zendarski et al. 2020), with NT students displaying the highest TSR ratings. Moreover, the lowest initial CoS teacher ratings were, similarly to the teacher ratings at InS, observed in ND female students, matching the research results of Lawrence et al. (2017) which implies that teachers are generally more accepting of ND boys’ behaviour, while ND girls’ behaviour tends to be problematised to a greater degree.

As expressed above, there was a significant improvement regarding InS teachers’ ratings (p = .03) between pre-test and follow-up assessments, which was witnessed across student types and genders. The results indicated an overall change in teachers’ perceptions of their TSR. No similar improvement was observed in the CoS. Rather, a decrease in teacher ratings of the CoS among the three measurement occasions was observed (Figure 2). Unfortunately, the data collected did not provide an explanation for this decrease. A possible explanation might be the beginning of the COVID-19 pandemic, which coincided with the study period during the Spring of 2020. It might have increased workload and stress among teachers, which in turn might have had a negative impact on teachers’ perceptions of their TSR.

It can be assumed that the teachers at both schools were aware of the aim of the study, and by that extension, demonstrated an initial increased interest in their relations with ND students. Unfortunately, there was a greater loss of participants from the CoS, especially in the follow-up assessment, which made it difficult to draw any relevant conclusions of the follow-up assessments.

Students of the CoS initially rated their TSR higher than students of the InS. However, ND and NT students at both schools rated their TSR fairly high during the pre-test ratings, which indicates that students at both schools initially felt that they had a good relationship with their teachers. Therefore, the study results may have been different if the initial student ratings indicated strained levels of TSR. In addition, the pre-test ratings, in contrast to previous research (Eisenhower, Blacher, and Bush 2015; Ewe, 2019), revealed that ND students perceived their TSR as high or even higher than their NT peers. This was true for all student groups in both schools, except for those of ND female students in InS picturing the lowest TSR rating of all student groups. The result is congruent with previous research demonstrating that ND female students tend to experience the highest levels of teacher rejection, while ND male students seem to perceive their TSR similarly to NT male students in general (Rogers et al. 2015).

Student ratings were essential in this study, as the students were not part of any intervention which meant that the Hawthorne effect (Cohen, Manion, and Morrison 2011) could be ruled out. Consequently, their ratings would reflect if the teacher intervention (Ewe & Aspelin, 2022) did affect teachers’ relational competence while interacting with ND students in practice.

A comprehensive comparison between ND and NT students at the InS indicated an overall positive development in ND students’ perceptions of their TSR between the pre-test and follow-up assessments, whereas the development in NT students’ perceptions moved in the opposite direction. However, significant effects (p = .03) were solely driven by ND female students and NT male students. While the results for ND girls revealed a positive trend, the results for NT boys were the opposite. Thus, an argument can be made that the intervention led to a shift in focus among teachers towards the targeted student group at the expense of another student group.

On one hand, the result can be considered a positive finding, aligning with those of Antoniou and Kyriakides (2013), and indicating correlations between teachers’ professional development and notable improvements for both teachers and the targeted student group in practice. Also, the positive finding among ND female students can be considered an important result, especially given earlier research that showcases feelings of teacher rejection (Ewe, 2019) as common among this student group. One tentative explanation regarding the positive outcome might be teachers’ heightened focus on their relations towards this student group, potentially leading to increased teacher-student interaction and thus aiding teachers in detecting the needs that students attempt to mask (Fugate and Gentry 2016; Moyse and Porter 2015; Solomon et al. 2012).

On the other hand, the result indicates a struggle for teachers to be relational sufficient for all students since the result expose that directing teachers’ focus towards one student group (ND students) risk straining their relations with other student groups (NT students). This was especially evident regarding NT male students declining results between pre-test, and follow-up assessment. While the data in this study does not provide an explicit explanation for the negative trend, a potential reason could be inferred from previous research indicating that boys generally receive most of the teacher’s attention in the classroom (Eliasson, Sørensen, and Karlsson 2016). Thus, when teachers’ attention is directed at one student group, it can be assumed to inevitably affects other student groups. This result holds importance as it highlights the challenges teachers face in meeting and satisfying all students while having limited time to educate large classes. Furthermore, this challenge is compounded by the increasing intention to include all students in the same classroom, which places a heavy workload on teachers while they strive to be relationally sufficient and meet the needs of every student.

The result of this study can therefore be considered important while organising schools and resources to offer inclusive and accessible education for all students, with or without neurodevelopmental symptoms. This is not least as mainstream classrooms following an inclusive education approach can sometimes be a challenge for teachers (Florian 2019; Perrin, Jury, and Desombre 2021). Variation in students’ needs might result in a dilemma for teachers, and as there is no ‘general need’ in a varied classroom, the challenge to adapt and respond to all students’ needs becomes more difficult as the variation of students’ relational needs widens (Haug 2017). The results of this study may provide a tentative explanation of the increased challenges faced by teachers of being sufficient and satisfy all their students.

Using a CoS was beneficial as it increases the reliability of the study. The decision to use intention-to-treat and last observation carried forward analysis decreased the risk of bias due to the large dropout rate at CoS. However, as only two ND female students and four ND male students rated their TSR in the follow-up test, these results were not discussed further despite intention-to-treat analyses. The same applies for CoS teachers’ estimations of ND girls as only one teacher completed the follow-up test regarding this student group.

Limitations

This study possesses several limitations that warrant addressing in future research. Firstly, the absence of a randomised sample suggests the potential for bias within the sample. Consequently, we have taken into account the conformity of the two schools when discussing the results. Secondly, the substantial dropout rate observed in the CoS during follow-up tests has rendered it impossible to compare outcomes between the two schools concerning teachers’ evaluations of their ND students and vice versa. This situation could potentially have a significant impact on the study’s validity.

Furthermore, it is important to acknowledge the presence of the Hawthorne effect (Cohen, Manion, and Morrison 2011), as the participating teachers were conscious of their involvement in a professional development intervention study. Additionally, the relatively brief time span between the pre-test and follow-up assessment represents a limitation that necessitates consideration when discussing the potential influence of participants’ attitudes, relationships, and responses.

Finally, the study detected noteworthy changes in perceptions of TSR among teachers, ND female students, and NT male students. However, the study is unable to delve into the underlying reasons for these changes due to the absence of qualitative data. As a result, further investigations are imperative to provide a deeper understanding of these shifts.

Conclusion

The results can be perceived as important contributions to existing research, as they indicate that relationships between teachers and ND students can be improved by using video-based teacher intervention as a tool for teachers’ professional development. This result is especially important, as earlier research tended to focus on the state of the TSR rather than on how to develop the relationships (Eisenhower, Blacher, and Bush 2015; Ewe, 2019). The results suggest that teachers changed not only their perspectives of how they perceived their relationships with their students but also how they changed their interactions with their students in practice in a way that affected students’ perception of their TSR. Further, the results are especially important because they shed light on students’ perspectives. Students themselves were able to make their voices heard in areas that ultimately affect them, without anyone else being given interpretive priority, which seems common in studies concerning ND students (Fielding 2001).

As the research is iterative, the attempt to answer some questions leads to new questions. Therefore, this study needs to be followed by further research examining the qualitative and elusive aspects of the relationship between teachers and ND students to create a complete understanding of teachers’ relational competence. One way of doing this might be through micro-observations of teachers’ interactions with ND students to visualise promoting patterns in teachers’ relational competence in the pedagogical practice.

Ethical approval

This study was approved by the Swedish Ethics Review Board (no: 2019–03533).

Informed consent

Informed consent was collected from teachers as well as from guardians and students from the age of 15 of both schools.

Author contributions

The first author, Linda Plantin Ewe, composed the study conception, design of the study and material preparation, data collection, initial manual data analysis, and writing of the first draft. Mona Holmqvist designed the study and contributed to material preparation. Sven Bölte, conducted the statistical analysis using SPSS. All authors commented and contributed on previous versions of the manuscript and read and approved the final manuscript.

Acknowledgments

We would like to thank Editage (www.editage.com) for English language editing and journal submission support.

Disclosure statement

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

Data availability statement

The datasets generated and/or analysed during the study are available from the corresponding author on reasonable request.

Additional information

Funding

This study is part of the Swedish National Research School “Special Education for Teacher Educators” (SET) program, which was graciously funded by the Swedish Research Council (grant no. 2017-060639).

Notes on contributors

Linda Plantin Ewe

Linda Plantin Ewe holds a Ph.D. in Special Education and works as a researcher and special needs teacher educator at Kristianstad University. Her research revolves around students with neurodevelopmental disorders, with a specific emphasis on investigating the relationships between teachers and students to enhance both social and academic accessibility for all students. Drawing upon her background as an elementary school teacher in math and science, as well as her role as a SENCO, Linda contributes valuable insights to her research. Her expertise encompasses several years of experience working in special schools that cater to students with neurodevelopmental disorders.

Mona Holmqvist

Mona Holmqvist, Mona Holmqvist, Ph.D., is a professor of Educational Sciences at Lund University, Sweden. She holds a Master of Arts in Primary and Secondary Education (Biology, Chemistry, Psychology and Pedagogy). Since 1996, she has worked as a teacher educator and researcher at the university level. Holmqvist has been the principal investigator of several external funded research projects, including PI of a national graduate school for teacher educators in special education (SET), funded by The Swedish Research Council 2018-2022. She has also been a council member of the World Association of Lesson Studies (WALS), a global association of research on professional development for teachers 2007-2022. Furthermore, she is appointed by the Swedish Government as a member of the National Agency for Special Needs Education and Schools (SPSM) Authorities Board, member of the editorial board of International Journal for Lesson and Learning Studies, editorial consultant for Australasian Journal of Special and Inclusive education, and member of the Editorial Committee of Research on Teaching and Learning (ForskUL). Among commissions of trust she has hold appointment at the Swedish Autism and Asperger Society. She has published several books and scientific articles in the field of education; which about 40 are listed on Web of Science.

Sven Bölte

Sven Bölte, Ph.D., is a professor of Child and Adolescent Psychiatric Science at the Department of Women’s and Children’s Health at Karolinska Institutet (KI) and a senior clinical psychologist at the Division of Child and Adolescent Psychiatry at the Center for Psychiatry Research in Stockholm, Sweden. He is also the director of the KI Center of Neurodevelopmental Disorders (“KIND”), researcher at the Autism Research Group at Curtin University, Western Australia editor in chief of the Scandinavian Journal of Child and Adolescent Psychology and Psychiatry, and associate editor of Child and Adolescent Psychiatry and Mental Health. He is the founder of the Scientific Society Autism Spectrum and an international ADOS and ADI-R trainer. His work has been recognised by “Life Watch Nordiska Priset,” “Årets Ljus” (Society Attention), “Psynk award GNET” (Sweden’s Municipalities & Regions), Autism CRC (Australia), Achievement in Autism Research, and Fellow of the International Society for Autism Research (INSAR). Among his commissions of trust are appointments at the Swedish Research Council, the European Network of Hyperkinetic Disorders, the National Society Attention scientific board, the Swedish Autism and Asperger Society, the Swedish Psychiatry Foundation, Swedish National Board of Institutional Care, and the National Agency for Special Needs Education and Schools. Professor Bölte has published more than 500 original articles, reviews, book chapters, and assessment and intervention tools in the fields of autism spectrum, ADHD, and other neurodevelopmental conditions, and has been cited more than 30,100 times (H-index 85).

Notes

1. A ‘Lead Teacher’ is a Swedish career position for teachers who are exceptionally talented in their work.

2. The 7th grade consists of students aged 13‒14 in Sweden.

3. The 6th grade consists of students aged 12‒13 in Sweden.

4. Swedish schools are not allowed to register students’ diagnoses, and diagnostic information must be collected via their legal guardians.

5. Aspelin and Jönsson (2019). ‘Relational competence in teacher education. Concept analysis and report from a pilot study’. Teacher Development, 23(2): 264–283. https://doi.org/10.1080/13664530.2019.1570323

6. Intervention school.

7. Neurodivergent students (i.e. students with ASD and/or ADHD).

8. Neurotypical students (i.e. students without ASD and/or ADHD.

9. Control school.

10. Intervention school.

11. Neurodivergent students (i.e. students with ASD and/or ADHD.

12. Neurotypical students (i.e. students without ASD and/or ADHD.

13. Control school.

Appendix A

Procedure and design of the teacher professional development intervention

The foundation for the present study is established through a teacher intervention conducted by Ewe and Aspelin (2022). A total of 53 teachers participated, with 33 from the intervention school (InS) and 20 from the control school (CoS). The intervention aimed to enhance teachers’ comprehension of relational competence. Every teacher from CoS who participated in the intervention study also took part in the present study. However, only 13 out of the 33 teachers from InS who were involved in the intervention study also participated in the current study. An apparent drop-out effect might potentially be attributed to the coincidental outbreak of COVID-19 in the spring of 2020.

The intervention’s evaluation was carried out through pre-, and post-tests administered as web questionnaires with open-ended responses, spaced three weeks apart. Both the InS and the CoS participants completed the tests individually, using their work computers and headphones. However, only teachers from InS completed the intervention, while participants from CoS exclusively undertook the pre- and post-tests. This methodology was employed to effectively assess the intervention’s impact on InS participants.

Initially, at both schools separately, one of the researchers delivered a concise introduction, presenting prior research on the teacher-student relationship. Subsequently, the participants took the pre-test, which incorporated an embedded video sequence portraying a teacher’s interaction with students. The specially produced video sequence was to be watched by the participants, who then answered related questions. The pre-test aimed at measuring the baseline understanding of relational competence among the participants.

Following the pre-test, an intervention was conducted at InS. The intervention was centred around video reflection and encompassed three components: i) a lecture on relational competence, ii) a presentation of the relational competence model⁵ (RCM), and iii) a collaborative session. Initially, one of the researchers delivered a lecture outlining the concept of relational competence, supplemented by a presentation of existing research in the field. This was succeeded by an introduction to the RCM, a model designed to analyse relational competence, along with an explanation of how the model could be employed to assess teacher-student interactions. The final component involved a collaborative session, where participating teachers collectively analysed two video sequences depicting educational scenarios, utilising the RCM model.

Finally, participants from both schools completed the post-test, which featured the same video sequence presented in the pre-test. However, the InS participants were now allowed to use the RCM model while performing the test. The post-test aimed to visually identify any potential heightened comprehension of relational competence among InS participants. The inclusion of CoS was to ensure that any observed developments did not occur naturally, independent of the intervention’s influence.

Appendix B

Table B1. Response rate of student ratings.

Students		No. of estimations	Response rate pre-test	Response rate post-test	Response rate follow-up
InS	Total	75	67 (89.3%)	61 (81.3%)	58 (77.3%)
	ND	17	17 (100%)	15 (88.2%)	16 (94.1%)
	− girls	8	8 (100%)	8 (100%)	8 (100%)
	− boys	9	9 (100%)	7 (77.8%)	8 (88.9%)
	NT	58	50 (86.2%)	46 (79.3%)	42 (72.4%)
	− girls	32	30 (93.8%)	26 (81.3%)	29 (90.6%)
	− boys	26	20 (76.9%)	20 (76.9%)	13 (50%)
CoS	Total	157	128 (81.5%)	98 (62.4%)	90 (57.3%)
	ND	18	13 (72.2%)	6 (33.3%)	6 (33.3%)
	− girls	7	6 (85.7%)	3 (42.9%)	2 (28.6%)
	− boys	11	7 (63.6%)	3 (27.3%)	4 (36.4%)
	NT	139	115 (82.7%)	92 (66.2%)	84 (60.4%)
	− girls	74	56 (75.7%)	50 (67.6%)	45 (60.8%)
	− boys	65	59 (90.8%)	42 (64.6%)	40 (61.5%)
Total		232	195 (84.1%)	159 (68.5%)	148 (63.8%)

Appendix C

Table C1. Response rate of teacher ratings.

Teacher estimations		No. of possible estimations	Response rate pre-test	Response rate post-test	Response rate follow-up
InS	Total	62	54 (87.1%)	44 (71.0%)	45 (72.6%)
	ND	17	17 (100%)	16 (94.1%)	15 (88.2%)
	− girls	8	8 (100%)	7 (87.5%)	7 (87.5%)
	−boys	9	9 (100%)	9 (100%)	8 (88.9%)
	NT	45	37 (82.2%)	28 (62.2%)	30 (66.7%)
	− girls	29	23 (79.3%)	18 (62.1%)	20 (69.0%)
	− boys	17	14 (82.4%)	10 (58.8%)	10 (58.8%)
CoS	Total	86	76 (84.4%)	69 (77.5%)	41 (46.1%)
	ND	19	16 (84.2%)	14 (73.7%)	6 (31.6%)
	− girls	7	6 (85.7%)	5 (71.4%)	1 (14.3%)
	− boys	12	10 (83.3%)	9 (75%)	5 (41.7%)
	NT	67	59 (88.1%)	54 (8.6%)	33 (49.3%)
	− girls	37	33 (89.2%)	30 (81.1%)	17 (45.9%)
	− boys	30	26 (86.7%)	24 (8.0%)	16 (53.3%)
Total		151	130 (86.1%)	113 (74.8%)	86 (57.0%)