Can the Engagement Model act as a replacement for the P- scale assessment system? Issues in monitoring the progress of students with autism and severe learning difficulties TEREZA READ

The Engagement Model was launched in January 2020, endeavouring to address the weaknesses of the P- scales assessment for students not yet involved in a subject- specific curriculum. This paper will discuss how and if the tensions between previously adopted assessment systems as discussed in teacher interviews can be reconciled through the Engagement Model in relation to students with autism and severe learning difficulties. The interview findings suggested that some of the problems with assessment, when applied in this context, are related to consistency and transferability, lack of formal recognition of non- academic progress, familiarity with the students, observation skills and training, workload and time, and subjectivity of judgement amongst professionals. When compared with the aims of the Engagement Model, the findings of the research suggest that even though it addresses some of the issues raised, it cannot act as a substitute to the P- scale system as it serves a different purpose.

The Engagement Model is based on Carpenter et al.'s (2016) project which focused on the engagement of students with autism in formal educational contexts. The professionals' response to Carpenter's model suggested that 95% of the participants would use the material in 'some way ' (p. 16), while 26% would fully adopt it. The percentages indicated that the material could be of value; however, adopting the system exclusively was received with hesitation, a testament to the fact that its scope could have been too ambitious.
In addition to the controversial results of Carpenter's initial study, various contradictive statements within the statutory guidance itself may reveal that the purpose of the Engagement Model had not been clearly defined by its originators. This paper will aim to examine if and how the engagement model can address issues around the pre-existing P-scale assessment system and the tensions between concurrently employed summative and formative assessment systems, aiming to monitor the progress of students with autism and severe learning difficulties (SLD).
To discuss the possible uses and limitations of the Engagement Model, this paper will be utilising current literature around assessment for students with autism and SLD, semi-structured teacher interview findings around the topic and a focused discussion of the Engagement model's statutory guidance. Following this discussion, an alternative assessment model, demonstrating how the Engagement Model can be utilised effectively within a wider assessment system will be proposed based on relevant previous research (Aidonopoulou-Read, 2019).
The statutory guidance (STA, 2020) endorses freedom for schools to use a bespoke curriculum, also common practice with P-scales (Ndaji and Tymms, 2009). This type of freedom, however, can be problematic for mainstream settings, which may not have the expertise to devise such bespoke systems. Therefore, this level of flexibility can create problems with inconsistency due to the lack of guidance.
It is questionable whether the Engagement Model addresses the issue of what assessment outcomes should be valued. Since it acknowledges (STA, 2020) that engagement is a procedural matter not related to outcomes like the P-scales are, replacing the one with the other may not be appropriate. This is not necessarily to question the value of engaging students in education; however, the replacement of a standardised tool such as the P-scales with a non-standardised tool such as the Engagement Model is inappropriate.

The purposes of assessment
The distinction between the purposes of assessment is essential to ascertain the reasons for adopting assessment systems. This, however, can be impractical as, to be holistic, an assessment system may require the simultaneous employment of various monitoring systems. Important to consider when choosing what assessment system to employ based on the target a practitioner aims to address is the differentiation between formative, summative and ipsative assessment and the role of feedback within those. The implications of autism and SLD and the obstacles those may pose in relation to assessment, also form part of the theoretical discussion that follows. Pellegrino et al. (2016) remarked that formative assessment is on-going and summative assessment is periodic and gives the teacher information about the grade-related progress of their students. Taras (2005) argues that 'The process of assessment leads to summative assessment, that is, a judgement which encapsulates all the evidence up to a given point' (p. 468). Black (2003), moreover, explains that even though summative and formative assessment are different, it is unrealistic to expect teachers to keep them separate, while Wilam (2000) highlights that formative and summative assessment should coexist, 'no matter what the tensions between the two might be' (p. 16).

Formative, summative and ipsative assessment
Even though, for practical purposes, formative and summative assessment systems ought to coexist, differences between ipsative (against one's own self) (Hughes, 2014), summative and formative assessment targets can lead to the simultaneous employment of various assessment systems. This can lead to time-consuming, inefficient schemes that may not work concurrently with each school's adopted curriculum. In the case of children with autism and SLD the application of generic assessment systems would be inappropriate as they form a different profile from each other (Jordan, 2001)

The role of feedback in assessment
Integral to any effective assessment system is the presence of feedback. For feedback to be effective, it needs to be regular, at the right for the student level, constructive, detailed and timely and it needs to be a dialogue between the teacher and the student (Nicol, 2010). Weurlander et al. (2012) comment that feedback is the 'key component ' (p. 748) to formative assessment and students in the position of receiving feedback need to be able to understand and act on it to make progress.
'Continuous assessment' and systematic observation and recording need to be clarified and understood by teachers' (McNicholas, 2000, p. 153). The lack of guidance and training as to how this can be achieved makes such practices less likely to form a natural part of the lesson (Shute and Kim, 2014). As Taras (2013) points out giving feedback that is '… dialogic, negotiated and understood by all' (p. 34) can be a challenge.
Reciprocal feedback, an important part of formative assessment, can foster student progress. Informal formative assessment (Ruiz-Primo, 2011) encourages reciprocity of feedback through the close observation of body language, crucial in the case of students with autism and SLD due to their communication challenges. The Engagement Model, adopts the observation of body language (STA, 2020) as a positive form of feedback, potentially establishing it as a formative assessment tool.

Challenges in assessment for students with autism and SLD
Reciprocal feedback, part of dialogic formative assessment, can be challenging for these students because of their difficulties related to expressing emotions and using facial expressions to communicate (Kroncke et al., 2016) related to autism and cognitive difficulties related to SLD. Interpreting student feedback as part of the Engagement Model, therefore, can be a difficulty, especially for inexperienced practitioners. Students with autism often display repetitive behaviours that would be classed as disengaged, however, they are idiosyncratic (Simmons and Watson, 2014): these can carry meaning related to the content of the lesson, but they often are unrelated, which underlines the necessity of training for accurate observation. A lot of those behaviours can be misinterpreted even by experienced practitioners (Aidonopoulou-Read, 2019); therefore, a parallel assessment and reporting system may be needed to reflect on student performance and the meaning of such behaviours. The need for training, also present with the P-scales, cannot be overcome through its replacement with the Engagement Model.
Children with autism and SLD have a deviant and delayed learning pattern (deviant is attributed to autism and delayed to SLD) but neither of those can be ignored (Jordan, 2001). The P-scales provide these multiple, smaller steps of learning (Imray and Hinchcliffe, 2012) which address the issue of delayed but not deviant learning, since they still respond to a linear developing pattern. Students with autism and SLD also appear to belong in an 'in-between group' that has specific strengths and difficulties (Jordan, 2001). Because of this unique profile, they may be able to achieve higher P level skills, however, these would not form part of their assessment and learning, while The Engagement Model is used for their assessment. This may disadvantage and limit these students, who, once engaged through captivating day-to-day activities, can achieve higher academic skills (Aidonopoulou-Read, 2019), making the model less appropriate as a standalone assessment tool for these students.

Rationale of the research
The main premise of the Engagement model is that it aims to address issues with the P-scales, which were deemed too complicated to address through the modification of the existing P-scales system (STA, 2020). Nevertheless, the specific issues it aimed to address were not identified or summarised within the statutory guidance framework. Therefore, of particular relevance is Phase 1 of prior research on assessment for students with autism and SLD which took place in 2009 that aimed to identify the problems with the P-levels and day-to-day assessment systems used in what, for anonymity purposes, has been named Highland School, a school for students with communication and interaction difficulties.

Methodology
This qualitative research aimed to establish the teachers' challenges when it came to applying assessment with their students with autism and SLD. The interview questions focused on formative assessment and how it is applied in conjunction with the P-Scales, which was used as a summative assessment tool in Highland School at the time. Details on what the teachers' understanding of the different types of assessment was (formative, summative and ipsative) and the way they used available tools to complete those assessments needed to be established before identifying the potential difficulties linked with those. For that reason, 264 Support for Learning • Volume 36 • Number 2 • 2021 © 2021 NASEN semi-structured interviews were considered to be an appropriate tool for gathering data through the use of context-specific questions, while also allowing for teachers to express their views based on their positionality.

Developing the interview schedule
Questions were formed based on literature and personal professional experience. Conversations with colleagues also informed the choice of questions, as those suggested several intricacies in relation to assessing non-verbal students with autism and severe learning difficulties. The school's most recent (at the time) Ofsted inspection in 2009 had also identified gaps in assessment in relation to those students.
The question types employed, as categorised by Spradley (1979) and Patton (2002), included descriptive questions (for instance question 1 asked: Could you give me a general description of the students in your class?), experience questions (e.g. question 3: What type of information does assessment give practitioners in the field?) and knowledge questions (e.g. question 4: What would you define as formative assessment?). The interview schedule was piloted with three former teacher colleagues, who had no connections with Highland school and their feedback was considered when revising the interview schedule, which led to some minor modifications in the wording of the questions to aid clarity.

The research context
Highland school is a special needs school with 300 students on roll. It is a wellresourced school with three separate parts (1. Primary and 2. Secondary for students with autism and other communication difficulties and 3. Primary for students with profound and multiple learning difficulties). The research took place in the Primary part of the school for students with communication and interaction difficulties as the research focus was on students with autism and SLD who were enrolled in this part of the school, while all teachers agreed to participate. The majority of the students performed between levels P1 and P8.
At the time, the school was following a modified version of the P scales in which levels had been broken down into discrete steps for summative assessment purposes. Formative assessment was mainly taking place as 'feedback' to the students through commenting on their performance in relation to their targets at the beginning and end of the lesson, while any other methods used were experimental and heavily depended on teacher creativity. Verbal feedback, it was felt, was not adequate for students with autism and SLD due to their communication challenges. Therefore, attempts to apply formative assessment were largely experimental and not always aligned with summative assessment. This made the gap between the two types of assessment difficult to bridge and resulted in a high level of paperwork and uncertainty amongst the practitioners. This uncertainty highlighted the necessity for the present research and provided a firm rationale for this approach.

Participants
Teachers in the case study school were chosen for their expertise, as this was crucial for the purposes of the research, which required a detailed understanding of the assessment processes in a special needs setting. The interviewees were given randomly chosen, background-unrelated pseudonyms to avoid revealing their identity. All teachers held Qualified Teacher Status and had experience in both mainstream and special needs settings. Details on their background and expertise have been included to contextualise this study: Laura: trained in England, initially worked in a different special needs school. Worked in the specific school for five years prior to the interviews. She had been working as a KS2 teacher both with P level and National Curriculum students.
Joanna: trained abroad, she worked with some mainstream school pupils sporadically in her country of origin. That was her first full-time job and her first job as a special needs school teacher in early years.

Sylvia:
A mainstream-trained teacher she moved to this school to work as an early years teacher. She had been in the school for four years prior to the interview.
Darren: Mainstream school background. He had been in the school for three years prior to the interviews. He had been working with KS2 students and it was his first special needs school post.
Elizabeth: A young, mainstream-trained teacher. This was the beginning of her second year teaching children with special needs.
Gwen: Gwen completed her training in mainstream schools and when she got hired in the special needs school she was in her NQT (newly qualified teacher) year. This was her second year in the specific setting and she had been primarily working with non-verbal students.

Sophie:
A mainstream teacher for about a year before she moved to the current setting. She had been working with early years and KS1 students. That was the beginning of her second year in school.

Helen:
Helen was an experienced mainstream teacher before she moved to this school. She had been working in the current school for over five years. Helen was a curriculum coordinator and she had been working as a middle manager for a number of years prior to the interview.

Ethical issues
The BERA Code of Ethics (2011) was employed. Whereas '… the social scientist realizes is that while the outside … does not know the meanings of the patterns, the insider is so immersed that he may be oblivious to the fact that patterns exist …' (Wax, 1986, p. 3); the researcher's dual role served as a contextual familiarity lens though which the unfamiliar (assessment) was questioned via the interviews.
Being an insider researcher can impose certain difficulties with objectivity in research. As the 15 th teacher in the school, remaining objective and not adopting the practitioner role while being a researcher was a challenge. Nevertheless, a topic like assessment did require a certain level of expertise and understanding, which was the insider's advantage. Contextual understanding and informal knowledge of the difficulties with assessment in the setting also informed the choice of pertinent questions.
Another challenge was managing colleagues' expectations as to what this research could achieve at the time. Some interviewees indicated that they were anticipating the research would provide a 'miracle solution' related to applying formative assessment with students with autism and SLD. Knowledge and expertise were also assumed as part of the researcher's role, on a topic that was under investigation and knowledge certainly was not established prior to the interviews. Therefore, through the clarification of the fact that from a researcher's point of view there were more questions to be asked than answers to be given, it was attempted to adjust the interviewer's and interviewees' position to that of equal professionals aiming to find answers in practice-based questions. Brown and Mclntyre's (1993) qualitative analysis method was used to inform the thematic analysis of the interviews. Based on this method, a random sample of three transcriptions was read and analysed manually, identifying initial codes. Following the second suggested stage of analysis, points of similarity and difference amongst the transcripts and in relation to the research questions were identified, using NVivo coding and establishing similarities and differences between manual and NVivo coding to aid a more nuanced analysis of the data. During the final stage of the analysis, emerging themes were identified as suggested through Brown and Mclntyre's method. Once the list of themes was finalised, all transcripts were coded using NVivo generated codes, followed by further manual coding to encourage more context-specific analysis and interpretation of the data.

© 2021 NASEN
Within the following data analysis section, quotes have been selected based on their relevance with the dominant themes and their ability to either summarise popular highlighted issues amongst teachers or to underline a unique and insightful opinion linked with the research themes.
Documentary analysis has been used in this paper as a tool to analyse the Engagement Model document in order to evaluate its potential uses alongside the clarity and usability of the document in education settings. Documentary analysis being used in conjunction with interview research is particularly relevant on this occasion as '… where the policy being researched is contemporary or recent, the policy-makers and implementers, plus of course those affected by the policy, may all also be the subject of research, typically using interviews … alongside documentary analysis' (p. 5).
Through the concurrent interview and documentary analysis, a number of themes related to the difficulties teachers identified and the Engagement Model attempted to solve have been identified and are discussed in the following sections.

The Engagement Model and subject-specific learning
Replacing the P-scales, a formal monitoring scheme, with the Engagement model, a formative assessment tool, can be considered unfounded. Even though the Engagement Model guidance (STA, 2020) suggests that students are not engaged in a subject-specific curriculum until they reach level P5, this is not entirely accurate. Reading, for example, is practised long before the recognition of words, something recognised through the P-scales which identify pre-reading skills such as 'Pupil recognises adult visually', which is one of the P3(i) level criteria (EQUALS www.equals.co.uk).
Identified through teacher interviews was the problem of clarity around P-scales targets, specifically related to foundation subjects. It was indicated that using the P-scales for core subjects was easier: 'I'd probably say the main subjects (are easier to assess) because the objectives are clearer. More specific' (Teacher early years).
Foundation subjects are topic based, however, core subjects are skills based: a generic system like the Engagement model is more appropriate for foundation subjects, for which the P-scales proved problematic: 'Knowledge and understanding of the world generally, history, geography, religion nearly impossible because that's obviously tied up with a lot of things' (Teacher and middle manager KS2).
Appropriate would also be the extension of the Early Years curriculum and assessment beyond the Early Years: 'Which is why I actually think foundation stage should be moved up to at least seven which I think is the general consensus in the early years world' (Middle manager KS1). Using existing, established systematic tools as opposed to replacing them with more generic ones, especially for core subjects, would be more fit for purpose.

Non-linear progress and generalisability of skills
The Engagement Model claims to monitor both lateral and linear progress (STA, 2020). Students with autism make non-linear progress (Jordan, 2001), which needs to be addressed by any assessment system used with these students: 'They've got big gaps in something else … and they can do something else that's higher. So, with our particular kids and maybe with special needs, it doesn't really show a true reflection of their ability' (Teacher KS1). Even though the five areas of engagement are 'not hierarchical' and 'there is no expectation that pupils need to demonstrate progress in all 5 areas' (STA, 2020, p. 10), no reference is made as to how non-linear (or, indeed, linear) progress will be monitored. Generic guidance around progress within the model can make it highly challenging for teachers to decide when and how to transition a child from the Engagement Model to a subject-specific curriculum. As students with autism have difficulties with the generalisation of skills (Jordan, 2001), monitoring lateral progress is imperative and appears to be addressed through the Engagement Model.

Consistency and transferability
Teacher interviews established transferability of information from one teacher to the next as well as one setting to the next as a major weakness of the system: 'If you're not there or when they move up to the next class, I don't think all of that information goes up. Of course, there's the end of year reports. There's a lot of information that's stored in your own head that doesn't get passed on' (Teacher early years). Naturally, and in the absence of a tightly framed monitoring system, some teachers will keep more detailed notes than others, providing an inconsistent monitoring system in which information passed on is not of the same level of accuracy: 'It depends on the staff a lot … Some … are better paperwork keepers than others. A lot of teachers do keep an awful lot in their heads. Don't … write a lot down' (Teacher and middle manager, KS2). The fact that observational data within the Engagement Model are presented in narrative form (STA, 2020) can cause issues of interpretation from one teacher to the next and reading narratives can be time-consuming. The way the Engagement model is employed can also be a matter of personal preference, creating inconsistencies in reporting across the board: 'I used to do my own thing for formative assessment and I think if every teacher is doing their own thing … and then you're all doing the same thing for summative, I don't see how that would marry up so much' (Middle manager, KS1). Based on the statutory recommendations, the Engagement Model would serve both as a formative and summative assessment tool and as its use depends on each practitioner's preferences, inconsistencies could occur in both summative and formative assessment.

Monitoring non-academic progress
The Engagement Model encourages monitoring progress related to the areas identified in the SEND Code of Practice (2015) and Education and Health Care (EHC) Plans (Children and Families Act, 2014). Teacher interviews highlighted that within the P-scale system non-academic progress is overlooked: it remains unclear, however, how this type of progress can be monitored or embedded within the model. As engagement is a process (Carpenter et al., 2016) and not an outcome, it would be difficult for the individual teacher to judge how well a student performs in relation to their individual targets or how success in all areas of learning can be noted. When asked how they monitor any progress unrelated to the lesson objectives one of the teachers responded: 'It probably wouldn't (be noted) … It should, but probably wouldn't be. It's not that I wouldn't choose to, it's just that I haven't got a system which would, is flexible enough necessarily to allow that' (Teacher KS2), which is an issue also not addressed through the Engagement Model. The question of where to note progress is also central, as inconsistencies in this respect will present issues of transferability since teachers will not know where to look for the information: 'So going to the toilet or something like that, it's not going to be on my weekly plan. It was … an issue where we'd record these things, an issue I don't think we'd ever resolve' (Teacher KS1).

Observation skills and training
The subtlety of the communication signals given by students with autism and SLD can be easily missed by the untrained practitioner. One of the main strengths of the Engagement Model is that it brings attention to those as they would otherwise be missed due to the lack of familiarity with the context and training (Nind and Strnadova, 2020): 'If the person doesn't know, can't really pick up on the small steps, the small things that a child does, that is an achievement. If you miss those small little things. Then what's the point' (Teacher early years). However, even within that framework, subjectivity can create inconsistencies in reporting progress as what one practitioner would class as communication and engagement, another would not: 'It is problematic because our guy may make eye contact for a minute and then he goes out next year and they want eye contact for five minutes …' (Teacher early years).
Further to this, consistency in assessment processes requires regular training. Since results in relation to the Engagement Model will not be formally reported (STA, 2020) there is less of an incentive to share good practice between schools through training. This can lead to the same issues raised through the teacher interviews: 'No-one has said to me 'oh, this is how you assess these children' I feel like what I'm doing is experimental' (Teacher KS2). 'I'm sure others would like some more training or some more ideas on how to do that' (Teacher early years).

Flexibility and workload
The Engagement Model offers flexibility (STA, 2020), a requirement for students with autism and SLD, whose needs and strengths can be different (Jordan, 2001). A flexible assessment system can encourage individualisation, however, without discrete and clear steps of progress and guidance, this level of flexibility can increase workload and create irregularities across the system. With reference to how individual progress is reported, teachers responded that, 'you'd have a huge book for each child … using in an informal way. I then feed into my planning which would be probably a good thing to do, but just limited time' (Teacher early years). This type of monitoring system also accentuates issues of transferability discussed earlier. As individualisation depends on the monitoring systems devised by the class teacher they may not be meaningful to others and the information may be lost when a child transfers from one group to the next, creating a new cycle of time-consuming processes for the new teacher.
Individualising assessment can increase teacher workload: 'It's difficult because each child is so different. You have to cater your assessment and everything you do to each specific child' (Teacher KS2). Simultaneous employment of various assessment systems can also pose an issue with monitoring and workload as 272 Support for Learning • Volume 36 • Number 2 • 2021 © 2021 NASEN highlighted by Highland school's specialist music teacher: 'I think it's difficult when you're working with seven children in a whole kind of variety of behaviours that might be going on at the summative assessment time'. Therefore, the suggestion of a 'flexible' assessment system can be less inviting when one considers the implications. Detailed advice within the guidance document on how to successfully utilise several assessment systems would be beneficial.
There are issues of efficiency when multiple assessment systems are employed: 'You kind of make the assessment that the school demands of you or the government demands of you. Whether it is done in a formal way, writing it down on a daily thing, I think it's unrealistic. Writing on a piece of paper which I'm never going to look at again, no-one else is going to look at again, why am I doing it?' (Teacher KS2). Multiplication of information that is not monitored into a 'common language' and a tool that everyone uses consistently results in time-wasting and inefficiency which results in failure of any adopted assessment system.

Accountability
The Engagement Model dictates that formal reporting of the results to the DfE will not be required, however, for the purposes of Ofsted the expectation will be to 'see evidence of the pupil's attainment, a focus on the outcomes and rigorous approach to the monitoring and evaluation of any SEND support provided' (STA, 2020, p. 28). This is problematic on two fronts: first, when results do not need to be published the system may be considered flawed or of less value than that of the National Curriculum. Second, reporting to Ofsted, especially under several undefined and far from concrete criteria, was identified as challenging by teachers as often they felt they had to 'perform' in order to satisfy the expectations of those observing them: 'Most people do find that some things that we have to do because of curriculum, because of inspections' (Middle manager KS1).
The framework, which the Engagement Model refers to is generic and does not address any of the concerns around the availability of rigorous reporting systems that enable teachers to communicate with others about the progress of students with additional needs, effectively.
Having a common language and available published data which enable one to compare how well students in one school perform in relation to students in another school is not only a matter of accountability, but also a matter of equity. Having special educational needs is a major contributor to socioeconomic inequality (Holt et al., 2019), which the lack of access to standardised data on progress can promote further. To make an informed choice, a parent would ideally have access to performance data in relation to how well the school addresses the needs of children also when their performance is lower than the P4 level, something that the statutory guidance linked with the Engagement Model does not promote.

A more efficient assessment model
Even though the Engagement model cannot be a replacement for the P-levels as it is a formative tool, it can play multiple roles within a school's assessment system and in relation to students with autism and SLD. Some suggestions around how this can be achieved, based on assessment principles adopted in prior research on monitoring progress for students with autism and SLD are discussed in this section.
A crucial first step towards learning in relation to students with autism is the establishment of engagement (Aidonopoulou-Read, 2019), while it has also been recognised that engaged students, inherently interested in learning are more likely to achieve their academic targets and reciprocate feedback (Aidonopoulou-Read, 2019). The engagement model was constructed to shift students with autism towards this ready state for learning (Carpenter et al., 2016) which can, in turn, develop into an inherent interest in lessons (Aidonopoulou-Read, 2019). For this reason, and for the case of students with autism and SLD, positioning the Engagement Model as a preparatory stage for formal schooling with the aim to progress to formal curriculum learning would be of higher relevance and value based on its design and focus. Its simultaneous use with relevant interventions such as Attention Autism (Watson et al., 2017) can be deemed appropriate as the intervention focuses on increasing attention and engagement and it is not subject-based.
The Engagement Model, however, can continue to serve a formative purpose as students become further acquainted with the formal curriculum. Engagement when it comes to students with autism and SLD can be determined through their body language and a combination of idiosyncratic behaviours, which the students © 2021 NASEN can eventually establish carry meaning through their interactions with others. The Engagement model would, in this case, benefit from being systematised and presented in a more explicit manner, similar to the way the behaviour checklist was utilised to evaluate the students' level of engagement in previous research (Aidonopoulou-Read, 2019). Within the behaviour checklist, a table including typical behaviours of children with autism and SLD as witnessed in Highland school was devised, which was then modified and individualised. This type of breakdown can create discrete steps, highlighting how each of the engagement criteria can be achieved and establishing the engagement model as a tool with measurable outcomes, which are transferable and comparable. Clarity of expectations can establish the tool as being more objective and it can address the problem exposed through the interviews around consistency and transferability.
It would be advisable that a systematic summative assessment tool including targets relevant to the students' level and needs is employed in parallel to the Engagement Model, giving valuable information on progress (academic and ipsative). This can ensure that subjective observations and interpretations of student behaviour remain accurate, especially since idiosyncratic behaviours in children with autism do not always carry meaning related to environmental stimuli, but may be related to anxiety or sensory differences (Simmons and Watson, 2014).