Keywords

1 Introduction

It has been hypothesized that special need students’ understanding, engagement, behaviour and concentration are among the main problems in teaching those with special educational needs. Understanding is especially difficult when learning material with a high cognitive load. Hence, an inappropriate instructional design can impose a high extraneous cognitive load that interferes with the students’ level of understanding any learning content. Cognitive Load Theory (CLT) focuses on presenting educational instruction to decrease extraneous cognitive load by considering split attention effect, redundancy effect and modality effect [43]. Errey et al. [20] stated that high extraneous load occurs when the learner tries to extract information from multiple sources and subsequently integrate it. More importantly, complicated or irrelevant information should be reduced when designing multimedia messages for special needs children, even more than for typically developing learners [28].

Building on Mayer s’ theory [32], people can learn more deeply from words and pictures than from words alone. Hence, we can reduce extraneous process by presenting text with near corresponding pictures [32].

The most popular methods to teach special needs in UK are symbol systems (Blissymbolics, Picture Communication Symbols (PECS), Widgit, Signalong and Makaton) as well as images and Diagrams. Building on our previous work [14, 15], we refined the previous platform from a pilot study by employing cognitive load theory principles. In this study, we employ split attention effect, redundancy effect and modality effect on educational content. This assumption challenged our research; hence, we suggested re-designing educational instructions in a specific way, which could reduce extraneous cognitive load, then improve the learning outcomes. We achieve our goals by using semantic annotation techniques by allowing the staff to select the required metadata or add the metadata of their choice. The motivations for promoting semantic annotation tools and cognitive load theory (CLT) motivated the design of a new platform which could support students with a variety of learning needs.

Therefore, we designed the SEN Teaching Platform (SENTP) ontology model using protégé 5, select semantic annotation tool (Amaya) coordinated with a web application.

We contributed with a detailed practical evaluation at seven schools caring for special needs in the UK. So during the study of real analysis, we applied new SENTP framework based on cognitive load theory to increase the understanding ability of concept affecting at the better able to learn, engage and concentrate.

2 Research Design and Platform Process

The study follows a design research approach that starts with learning about the problem space leading through to design artefact evolvement and evaluation. Hevner [24] described the process as an effective solution to a problem. The effectiveness of the solution must be provable through an iterative evaluation of the design artefact(s). The artefact resulting from the Design Science Research (DSR) in this work was to induce the characterisations of the SENTP model. This study aims to build and refine a number of micro-designs (content, annotation and process). Importantly, core theories of learning and memory systems, including those related to cognitive load, direct the design of the SENTP [31, 34]. Artefacts (such as web content and the SENTP architecture) are refined to minimise the cognitive load and enable efficient use of working memory in order to improve communication, aid understanding, and reduce the effort and time needed for resource preparation. Typically optimal performance can be achieved by offering presentation strategies that reduce cognitive load [29]. Consequently, the annotation techniques used with the Amaya [2] tool offer a number of types of annotation for field testing such as images, information, symbol systems, pictures, information, and audio. The final refined framework can be summarized in four main steps, as illustrated in Table 1. The table presents the Iteration Steps, Method and Input-Output Model.

Table 1. Iteration steps, method and input–output model
Full size table

3 Design and Build SENTP Framework

This section describes the design of a SENTP framework and subsequent development of content. The design based on our previous studies [14, 15]. In response to the users’ requirements from previous studies and based on further research in literature, we listed a set of objectives, which is described in Table 2.

Table 2. SENTP framework with CLT principles
Full size table

Cognitive load (CL) refers to the amount of cognitive demand imposed by a particular task on a person, which is related to the limited capacity of working memory [11, 12] The rationale behind CLT is that the quality of instructional design is enhanced if attention is paid to the limitations of working memory. The objective of using ideas surrounding CL in the SENTP is to examine a key question: “To what extent can semantic annotation techniques reduce the burden of CL for SEN learners?”

There are two aspects to CL, explained by Sweller [39]:

  • Reducing intrinsic load: The design of the SENTP should consider the ability of semantic annotation to lower the cognitive load by reducing task complexity, as explained by Ayres [4]. This will be done by adding different forms of annotations, real images, and improving the presentation layout of the User Interface (UI) by using different colours and fonts relevant to the needs of the SEN user.

  • Reducing any extraneous Cognitive Load (CL) imposed by the instructional design itself through the integration of the annotations.

Figure 1 illustrates the proposed SENTP framework. Amaya is certainly a good starting point for creating educational content. In this design, it is an annotation tool that supports the teaching and learning of SEN students. However, it needs to be modified to meet the SENTP requirements listed in Sect. 3. The features that need to be modified are as follows:

Fig. 1.
figure 1

SENTP Framework

Full size image
  1. 1.

    The SENTP should have an option to display visuals (images, symbol systems such as Makaton, PECS, and Widgit) while verbally demonstrating the platform or the provision of audio annotation using headphones with visuals to reduce the contiguity effect [40].

  2. 2.

    A combination of text and visuals, such as images or symbol systems or text and sound, can reduce the split-attention effect. Cognitive capacity in working memory is limited so that if a learning task requires too much capacity, learning will be hampered. The recommended solution is to design instructional systems that optimize the use of working memory capacity and avoid cognitive overload. These results in reducing the time required to keep information active in working memory, without the need to integrate information resources mentally.

  3. 3.

    The learning content should include short text to re reduce the intrinsic load.

  4. 4.

    Visuals should include enough information to reduce the redundancy effect. Importantly, educational content enables exploration of the influences of semantic annotation on SEN teaching and learning, including motivation, understanding, communication and satisfaction. Field testing was used to examine the effectiveness of the SENTP.

Figure 2 presenting a SENTP diagram of employing CLT principles channel, diagrams and spoken text that rely on both auditory and visual modalities can be used.

Fig. 2.
figure 2

SENTP with cognitive load theory (CLT) framework

Full size image

4 Building SENTP Ontology

The architectural approach of the semantic web is relatively straightforward, creating a layer on the existing web that enables advanced automatic processing of the web content so that data can be shared and processed by both humans and computers [27]. Ontologies from part of the semantic web and provide a mean for defining a domain-specific language for education- Special Educational Needs (SEN) in particular.

The basic components of the semantic web are; metadata, semantic web languages, ontologies, semantic mark-up of pages and services [18]. They can be summarised as follows:

4.1 Metadata

Metadata is data about data which means the data that describes another piece of data as shown in Fig. 3. Some developers consider the Metadata as the heart of e-learning [38].

Fig. 3.
figure 3

Representation of metadata

Full size image

4.2 Semantic Web Languages

The basic ontology language and simple models used for combining data and representing information on the web. They have typically used the Resource Description Framework (RDF), which could be represented as a labelled graph [18, 8, 22, and 1] and are based on XML (then called RDF/XML). Many of the languages based on XML. Resources are described using RDF statements, which are represented as subject, predicate and object (see Fig. 4). SENTP ontology is designed and implemented for a wider design using desktop system (Protégé 5). Protégé has become widely used software for building and maintaining ontologies [36].

Fig. 4.
figure 4

Representation of RDF statement

Full size image

5 SENTP Implementation

We designed educational ontology and website using HTML that integrates with the selected annotation tool ‘Amaya’. The selection was based on the cost and the school convenience. First, the class teacher selects a poem, style, and type of annotation required for the class demonstration. The platform is prepared beforehand with the kind of annotation required (e.g. images or Makaton symbols). The annotation options are wide-ranging, depending on SEN age and needs. Figure 5 presents the poetry webpage with different annotation options.

Fig. 5.
figure 5

SENTP Graphical User Interface (GUI)

Full size image

Figure 6 presents the ‘Bedtime’ poem with the selected words for annotation and an annotation symbol ‘I’ with Makaton.

Fig. 6.
figure 6

Presents the ‘Bedtime’ poem annotations

Full size image

Figure 7 presents another annotation with text and image.

Fig. 7.
figure 7

Illustrate the integration of the picture with the text

Full size image

6 Data Collection

Data were collected whilst field testing the designed artefacts. The aim was to gather data in order to assess and further develop the SENTP framework and educational content. Data collected from February 2013 to October 2013 and follow qualitative approach. Semi-structured interviews were used, in addition to field notes and researcher/staff observation. The interviews provided an opportunity to explore personal experiences that may otherwise have been hard to observe [31, 41]. The data collection activities are described below.

6.1 Participant Recruitment

Field-testing annotation interviews were carried out at seven schools in the UK. Table 3 provides an overall description of the participants.

Table 3. The overall description of the participants
Full size table

6.2 Research Instruments

The main research tools were the interview questions framework, and the website supported by Amaya software. The questions were direct and open-ended to allow participants to be more engaged and detail their experiences. An example learning website was designed using HTML, supported by Amaya, containing poetry of different styles as the sample of teaching materials. The NVivo11 [37] software package was employed to carry out thorough and reliable qualitative data analysis. It is a very reliable management tool that can aid in analysing the data [44].

A prototype was presented in schools on a laptop and a projector in a classroom. A digital voice recorder, “Olympus VN-8600PC” was used along with a small notebook and a pen for extra notes.

All the interviews were managed by the researcher using the interview sheet. All recordings were transferred onto a personal laptop and two USB drives secured with a password known only to the researcher.

7 Data Sources

Following the ethical approval given by Brunel University, sixty-one schools caring for SEN students were approached via email, telephone and the postal service. Each school was sent a covering letter along with an information sheet. Seven schools agreed to participate in the research. Of those seven, six completed the interview procedure, and one subsequently withdrew. The data was collected over a six-month period due in part to the scheduling and timing pressures within a typical school. In total, twenty-three interviews were conducted. Table 4 Group demographics are included in Table 5.

Table 4. Group demographic data
Full size table
Table 5. ‎ Open-coding concepts categorization
Full size table

8 Thematic Analysis

Thematic analysis is used as part of the wider design process to elicit future requirements and more importantly determine artefact effectiveness. Consequently, the SENTP design with the Amaya annotation is assessed during interviews, using CLT principles to examine reductions in cognitive load. All interview data were analysed using thematic analysis [7, 9]. After gaining familiarity with the data, the transcription analysis process involved listening to the interviews, reading through the data and uncovering possible themes [7]. First codes were generated from the transcript information. The qualitative data analysis software NVivo11 was used to facilitate the thematic analysis, and transcript data was exported to NVivo11 that then coded the features from the entire dataset. Themes were identified and reviewed. Each theme captured something important about the data in relation to the requirements and problems being addressed. All data relevant to each theme was extracted and matched with specific codes. A model of themes is presented to show the connections and relationships between the themes and the subthemes [6, 7]. Below, nine codes are describing the themes and sub-themes. Table 4 outlines the details of the themes and sub-themes.

9 SENTP Evaluation

The evaluation shows that SENTP within the SEN domain makes significant contributions towards SEN teaching and learning. One of the most important visits was to the special needs secondary school where a demonstration was conducted by T7-M-SMO. This was a challenging classroom, with different levels of severity of special needs. The teacher presented the class with four poems, asking students if they wanted more content after each poem, instead of teaching just one, as agreed before the lesson. The immediate feedback from the students and the teacher reaction demonstrated that the entire class was engaged and motivated during the demonstration as noted by T2-M-P: ‘they were well involved, they can take part with their actions, with their hands, fingers, very engaged and looking at the computer screen and watching all the images, very involved’ and part of the email sent by T7-M-SMO ‘. In consultation with the class staff, it was felt that the session was very successful. This was made clear by the high level of pupil engagement during the lesson ‘… the design of the prototype shows promise’. Also, T8 confirmed that some children they may lose their attention because of their physical disability and the traditional way they are using to teach them ‘TA8-M-SMO: ‘It does make it easier because our children will not have the ability and understanding of turning pages and going there because they lose their attention, their attention is only a couple of seconds’. This shows that SENTP can keep the concentration and engagement of SEN student with physical disability last longer. This can propose a better understanding of the learning materials.

Using the SENTP encouraged group work method as well. The class teacher, T7-M-SMO reported that he will suggest group work method for his class in addition to the current approach of independent learning as the headteacher attended and observed the demonstration and establish that the demonstration was successful: ‘I am going to have a word with him and see what he thought because our English and Maths is usually done at workstations and I think there is space for group work as well; they work very well’. He confirmed that in his email when he said ‘The prototype could be used for target groups during teaching and would be a valuable resource when finalised. This shows that SENTP can support group work to reduce the one-to-one staff demand and to overcome the difficulty in learning with others in small or large group settings.

Furthermore, the tool was shown to be useful for class management, as indicated by T1-M-SMA, a teacher in the special secondary school. Her class includes children with a mix of severe issues, and she has good background experience: ‘we have children with severe learning difficulties, including children with Down’s syndrome. I have experience teaching autistic children; we also have children with genetic disorders and severely challenging behaviour; we have a huge range of children’. T1-M-SMA commented on classroom management: ‘all the students were quiet and listened when the lesson started’. This shows that SENTP support teachers with managing the special need class that they could easily distracted with behaviour problems.

All the participants agreed that visuals are important to SEN. The manager at the pre-school indicated that the image annotations within the SENTP could improve engagement and attention ‘It’s visual, isn’t it? It keeps their attention’. The teaching assistant at the pre-school indicated that having different options for annotation types can offer various types of teaching methods: ‘They give a broad range of ideas and thinking, and we can use different ways to teach children’. This point was furthered in an email sent by the teacher from a special secondary school: ‘The ability to have instant access to images etc. and not have to rely on on-the-spot searching would contribute to the pace of the lessons and thereby minimise anxious behaviour and increase understanding’. Hence, SENTP could reduce the student cognitive load because it increases attention.

Most of the participants pointed out that the annotation included within the SENTP is fun and interesting, as indicated by the teaching assistant from the special secondary school: ‘I think it is quite engaging; children enjoy looking at the images in the classes’. The teaching assistant from the special secondary school agreed: ‘Actually, the student sitting next to me actually participated because he was signing what he saw, what you said […] he seemed to be enjoying it, so yes’. The teacher from the pre-school said: ‘The session went quite well. It was very easy-going, the children really enjoyed it, and I think they benefited from it’. SENTP can reduce a behaviour problem which is one of the main concerns in teaching special needs. SENTP could reduce student frustration and improve their mood as noted by many participants, and the teaching assistant from the special secondary school confirmed that ‘It takes away quite a lot of the frustration of not understanding. It takes out the boredom of not understanding until all of them have understood’. These benefits of the SENTP are of particular importance in reducing behaviour problems of special need students as they tend to have low tolerance levels and high frustration levels.

The participants agreed that the SENTP could be adapted to subjects other than poetry such as Religious Studies, Science, History and Math, as indicated by the teacher from the special secondary school: ‘I think it is really good; that is what I am left with today. That the concept of a click in the text and it pops up with a photo is very good, something we could use for poems, for all kind of things, anything that has text’. This demonstrates that SENTP can be utilised in teaching different learning content to support SEN students such as the one who struggles with their poor handwriting skills and the difficulty in following complicated directions or remembering directions for extended periods of time.

The teachers consider the SENTP as an ‘easy to use’ tool, which the teaching assistant touched on: ‘Yes, I would like to use it because it is the simpler way to teach and grab children’s attention, and [it works] on different levels for different children’.

Many participants believed that the SENTP can save preparation time; the teaching assistant from the special secondary school said: ‘it’s good for teaching; we can concentrate on assessing more students because we have to assess them on regular bases’. This shows that SENTP can save the staff time and effort to free them for other significant work.

Other points related to the ability of the SENTP to support the teaching staff were noted by some of the interviewees who said: ‘I think the speech therapist would be very interested to see this software’ […] ‘it’s interesting’.

In summary, since student’s capacity to learn a concept is directly associated to how much cognitive load is used to comprehend the material and their working memory is fixed size. The significant problem is when learning is impaired when processing requirements exceeds the capacity of working memory. We designed the educational content instruction with semantic annotation to minimise cognitive load. To measure cognitive load we used. Not many of the teaching staff aware of the cognitive load theory and its effect on learning, which makes measuring not precise. Hence, ten (50% of the participants) believed that SENTP could reduce spilt attention effect, and then enhance learning (see Fig. 8). Only five (25% of the participants) believed that SENTP could reduce redundancy effect (see Fig. 9).

Fig. 8.
figure 8

Split attention effect

Full size image
Fig. 9.
figure 9

Redundancy effect

Full size image

10 Conclusion and Future Works

A considerable gap exists between Semantic Web utilisation in the field of mainstream education when compared to special educational needs education. The teaching methods available in a special needs school are typically based on time-consuming, manual methods. SEN can affect a child’s ability to learn, their behaviour and ability to socialise. Reading, writing, understanding, concentration and physical abilities are also more limited (Chen, 2011). This paper presents a novel approach to special needs teaching and learning and finds that Semantic Web annotation techniques can reduce the SEN cognitive load within the classroom. Consequently, the designs and resulting system (developed using Amaya) and the usage methodology enhances the learning process of SEN through the use of a range of annotation types.

Design practice and contributions underpinned all of this research. Design Science Research methods directed the constructs, models, methods and instantiations employed. The artefacts include both larger frameworks (e.g. SENTP) and smaller media content. Furthermore, SENTP ontology is designed and implemented for a wider design using Protégé 5.

Participant requirements defined the application of CLT principles within a number of technological artefacts. The platform was extended by following a set of methodological guidelines to reduce the SEN cognitive load, reducing the split-attention and redundancy effects. Interviews were conducted to identify the impacts of semantic annotation techniques when teaching poetry for students with a wide range of SENs and with different levels of understanding. Interview analysis supported a combination of text with images, sound, or symbols in order to reduce the SEN cognitive load. Consequently, the classroom benefitted from reductions in behavioural problems and increasing SEN understanding. Poetry teaching material was used that supported CLT, increasing SEN engagement and motivation. The platform can also support teaching staff with class management techniques, including resource preparation. Schools use different types of sign and symbol systems, many of which are integrated into the platform. Children with additional languages are also possible end-users of the proposed approach.