Difficulty in integrating and processing environmental and sensory information has been observed in various neurodevelopmental disorders, especially attention-deficit hyperactivity disorder (ADHD) (Ghanizadeh, 2011; Zimmer et al., 2012). The American Psychiatric Association (APA, 2013) considers ADHD the most prevalent psychiatric disorder among the young child population. In a communal classroom of ± 30 children, two to three are diagnosed with ADHD (Barkley, 2018; Micoulaud-Franchi et al., 2016), increasing the likelihood for them also to experience sensory integration or processing difficulties (Ghanizadeh, 2011). Table 1 summarises the responses and characteristics the child is confronted with to a different degree, intensity and nature.

Professionals can benefit from using psychometric-founded instruments to interpret environmental and sensory stimuli to adapt to communal spaces, for example, the Classroom Climate Scale (López et al., 2018), Classroom Sensory Environment Assessment (Miller-Kuhaneck & Kellehers, 2018) and the Sensory Gating Deficit and Distractibility Questionnaire for adults with ADHD (Micoulaud-Franchi et al., 2016). However, these mentioned instruments focus predominantly on a specific aspect of sensory integration (e.g. environmental design) or an age group (adults), which does not cater for the needs of a South African early childhood teacher.

Delaying interventions will increase the prospects of children with ADHD being unsuccessful in their school trajectory (Huerta, 2017). Therefore, looking to early childhood teachers as ‘path changers’ is not unfounded. Introducing young children to quality learning environments is not a luxury but rather a necessity. Learning spaces ought to be perceived as emotionally safe, socially smart, environmentally friendly and cognitively supportive (Jensen, 2003). Ergonomics for children integrates a wide swath of disciplines (e.g. psychology, rehabilitation, education, architecture, law) to ensure developmentally appropriate practices (Lueder & Rice, 2007). More specifically, sensory ergonomics is considered a trusted strategy to cater for children’s special educational and developmental needs by constructing conducive classroom conditions (Lombard, 2015; eds. Lueder & Rice, 2007). Studying children’s ability to self-regulate physical, emotional and cognitive responses (sensation) within the learning environment (ergonomics) sheds insight into whether their sensory nervous system is promoting or hindering their functioning (Brown, 2002; Lombard, 2015).

Early childhood education in South Africa, also known as Foundation Phase, is children’s first compulsory entry to schooling. Six-year-old children enter the schooling system as preschoolers in an informal classroom before transitioning to a more formal school setting in subsequent grades. Although informal (Grade R) schooling and formal (Grades 1–3) schooling are clustered as early childhood education, the curriculum design, pedagogical approach, teaching and learning support material, and classroom setup for preschoolers and schoolers differ significantly (Van Heerden & Du Preez, 2021). Herewith some challenges teachers in South African schools come to face:

Firstly, it is not uncommon for developing countries to have mainstream and multi-aged classrooms in the early years that are significantly larger (n = ≥ 45) than international classrooms (n = ± 24) (Howie et al., 2017), suggesting that larger classes increase the likelihood of hosting more children with ADHD (Perold, Louw, & Kleynhans, 2010). Secondly, South African teachers are often situated in disempowering learning environments in terms of physical size, socio-economic status, high child-to-teacher ratios, limited access to multidisciplinary teams for guidance and a lack of developmentally appropriate resources and equipment (Balfour, Mitchell, & Moletsane, 2008). Thirdly, Foundation Phase teachers may not be well informed about neurological disorders (e.g. ADHD, sensory processing disorder [SPD] and associated subtypes) and interdisciplinary interventions (e.g. sensory ergonomics) to adapt to learning environments (Brown, 2002; Perold et al., 2010). Fourthy, optimising an environment for rich sensory stimuli and SPD is the specialisation field of registered occupational therapists (Ayres, 1979; Lombard, 2015), leaving South African teachers to decipher child–environment synergy for themselves. Fifthly, too few training opportunities for continuous professional teacher development (CPTD) is offered to broaden their knowledge and skills on special educational needs (De Clercq & Phiri, 2013). Lastly, teacher-friendly psycho-educational resources to assist teachers in creating conducive learning environments are rarely freely available.

Considering all mentioned dichotomies, a prospect is offered to recouple cross-disciplinary knowledge systems to promote conducive classroom conditions for children with ADHD who could also experience sensory integration/processing challenges. The nexus this article would like to present is a well-designed psycho-educational assessment instrument for Grade 1–3 teachers when creating inclusive, conducive and sensory ergonomic learning spaces for children diagnosed with ADHD. The researchers designed an instrument that assesses environmental conditions which teachers could use to enhance the classroom climate.

This study’s main objectives and processes are as follows:

Evidence for instrument validity commences with an adept conceptualisation phase that requires identifying a suitable conceptual or theoretical framework, relevant extensive scholarly literature and a panel of experts to scrutinise the construct(s) and provide feedback on the content validity (Du Preez & De Klerk, 2019; Michell, 1997; Zhou, 2019). Existing scholarly theories were utilised to generate items for the SCTQ to measure sensory ergonomics as a latent construct. The SCTQ is based on meaning-making frameworks about sensory integration/processing, sensory ergonomics and the triad characteristics of ADHD. Modulating sensory input is imperative for everyday functioning as it influences productivity, focus, attention, communication and interaction (Alnajjar et al., 2015; APA, 2013; Brown, 2002; Lombard, 2015), hitherto a challenge for children with ADHD.

Factual statements (items) were derived from literature and categorised into subscales to represent the dimensions of the latent construct. The aim was to create concise and unambiguous items that would measure one central idea and remain consistent with the purpose of the measurement (Murphy & Davidshofer, 2005; eds. Schweizer & DiStefano, 2016). The final version of the SCTQ is summarised in Table 2, listing the construct, then the question number and the item statement followed by Likert scale options. The questionnaire is self-administered and can be used by early childhood teachers themselves or informed observers.

As presented in Table 2, the three underlying latent constructs, or subscales, represent the conceptual framework. These latent constructs offer item statements that suggest practical ways to adapt, change and manipulate the existing communal learning space. Sensory integration/processing interventions assist children with ADHD in modifying and regulating their own social and academic behaviour, which they find challenging, as alluded to in the triad of characteristics. The instrument is a multidimensional questionnaire that assesses aspects of overarching sensory ergonomics in the early childhood classroom. The SCTQ’s constructs are given as follows:

Please note that the olfactory system (smell and taste) is not included for manipulation, as Lombard (2015) strongly advises to keep it neutral or natural.

A purposive and convenience sampling technique was used to recruit the panel of experts, fieldworkers and teacher participants.

Firstly, researchers consulted with the panel of experts to appraise the content validity of the SCTQ and its generated statements (items). The panel of experts was multidisciplinary, including two early childhood specialists, one registered occupational therapist, two registered educational psychologists and one research psychologist. Secondly, a protocol document was compiled and the fieldworkers were trained on handling topics such as ethical procedure to obtain informed consent, ensuring safe and anonymous participation, how to self-administer the SCTQ, authoring a qualitative interview report and capturing the raw data in Microsoft Word and Excel. The fieldworkers were qualified early childhood teachers enrolled for their postgraduate studies in Learning Support. They were authorised by the Department of Basic Education to visit any primary school in South Africa and issue a copy of the registered ethics certificate granted by the University of Pretoria’s Ethics Committee of the Faculty of Education. Thirdly, the participants, who self-administered the SCTQ, were qualified and appointed in-service teachers within the early childhood sector either in Grade 1, 2 or 3. The participants were mainly female, as most South African teachers in Foundation Phase are women (Petersen, 2014; Sak, 2015). The teacher participants’ biographical profiles are tabulated in Table 3.

As depicted in Table 3, both rounds of piloting presented a teacher sample profile of having less experience in teaching in the early years; schools with more nongovernment income and; teachers perceiving their ADHD knowledge as adequate.

Ethical approval was obtained from the University of Pretoria, Faculty of Education, Ethics Committee (reference number: UP09/04/01).

The combination of Rasch statistics, qualitative interviews of the fieldworkers and the review of the items by panel experts led to the revision of the instrument (Cavanagh & Romanoski, 2006; eds. Cavanagh & Waugh, 2011). The first version of the SCTQ had 55 items, and the model fit was significantly different from the data (p < 0.05). Four items had significant fit residuals, indicating potential problems. Thirty-one out of the original 55 items displayed disordered thresholds and categories, with less than 10% of participants endorsing the option. The five-point Likert scale categories were combined into three categories, which resulted in a better fit and no disordered categories. Differential item functioning was detected for five items compared to all the demographic characteristics. The refining of the items resolved the DIF. Principal component analysis in RUMM 2030 was used to investigate unidimensionality, and the three factors were confirmed to be independent. The cross-disciplinary panel of experts revised the first version of the SCTQ so that items could be combined or deleted. The revised instrument has 24 items and an improved overall RMT fit (χ² = 132.744, p = 0.015).

The next round of data collection for the SCTQ revealed that each construct fit the Rasch model, had acceptable reliability and was unidimensional (Table 4).

Four items were identified that had problematic categories. Reliability indices were above 0.700 and deemed acceptable (Linacre, 2021). Differential item functioning was absent for grade, school type, teacher knowledge of ADHD and teaching experience. The second version of the SCTQ showed evidence of internally reliable and valid inferences to assess classroom sensory ergonomics in the classroom. The individual item fit statistics are shown in Table 5 with their standard error (SE), fit residual values, degrees of freedom (df), chi-square and probability values.

As seen in Table 5, none of the items had fit residuals above or below 2.5, and none of the items significantly misfit the Rasch model.

The correlations among the latent traits are shown in Table 6. As derived from the Rasch logit conversion, the constructs had moderate to strong relationships (0.378–0.628). The logit scales ranged from −3 to +5, and attention (co)regulation was the most difficult to implement in the classroom (M = 0.361, SE = 1.426).

An examination of the items and latent trait structures was conducted with a Bayesian CFA. The posterior predictive p is close to 0.50 (p = 0.38). The 95% highest posterior density interval did not contain a 0. All items had convergence statistics lower than 1.1. The posterior predictive p showed a good fit, but the model could be improved. One item did not significantly load onto its factor, question 6 (‘labelling of resources and designated areas’). The convergence statistics were acceptable for the specified structure.