3. Pilot Work
3.1.4. Approaching research with an SEND sample
research, decisions had to be made about how the children with SEND would be identified and what measures to use in order for the items to be accessible.
3.1.4.1. Measuring SEND. SEND has been measured in research in a variety of ways. When recruiting an SEND sample, the way in which the diagnosis is
collected/reported/measured can be carried out in different ways. Studies that have utilised a qualitative approach (i.e. interviews) often ask for volunteers to come forward to take part. Glazzard (2010) conducted interviews with students aged 14-15 years to explore the impact of dyslexia on their self-esteem. Glazzard (2010) approached two schools and nine
students volunteered to be interviewed. There are obvious limitations to this approach, with the primary limitation being sample bias as only certain children who would be able to talk about their issues would come forward.
Another way in which researchers have accessed children with SEND is by
reviewing files to find students that have the appropriate diagnosis (Baumeister, Storch & Geffken, 2008; Ingesson, 2007). Ingesson (2007) also carried out interviews with 75 individuals aged between 14-25 years to explore their experiences of growing up with dyslexia. Ingesson (2007) recruited a sample of individuals with dyslexia by reviewing assessment records at a dyslexia clinic at which the individuals were diagnosed and then contacted them to take part in the research, but this is very ethically sensitive. Baumeister et al. (2008) also used existing records to find their sample of 77 learning difficulty diagnosed individuals aged 5-18 years. Baumeister et al. (2008) used the University of Florida Division of Child and Adolescent Psychiatry to search for individuals with a learning disability, along with accompanying measures of behaviour, depression, and anxiety. While this method of recruiting a diagnosed sample would be the most accurate, it would also be time consuming to review many files to find the few participants that meet inclusion criteria. Additionally, for Ingesson (2007), who recruited their sample for additional interviews, many individuals may not have liked being contacted; indeed, a quarter of the original sample did not want to participate.
Similarly, another way to recruit an SEND sample is to gain access to a sample of SEND individuals who are already taking part in longitudinal research, as this group would be readily accessible. Kaukiainen et al. (2002) used this approach and recruited a group of 141 children (both with and without SEND) who were midway through another research project on learning disabilities (LD). They were able to use the scores from tests
administered by the original team to identify which students have LD. Students were categorised as having LD if they fell in the bottom 20% of the scores for the assessments. This method would be beneficial in terms of time, as it would cut out a lot of time spent on identifying the SEND children; however the sample may have been influenced by the previous research they had taken part in and demand characteristics could have been increased.
Finally, there are self-report methods for identifying a SEND sample. These are usually done in an open-ended way, asking children to write if they have a diagnosis. Sentenac et al. (2011) investigated bullying in children with a disability or chronic illness (D/CI) in a sample of 12,048 students aged between 11-15 years. They identified the individuals with D/CI in the sample with the inclusion of several questions adapted from previous large-scale surveys. They asked ‘Do you have a long-term illness, disability, or medical condition (like diabetes, arthritis, allergy, or cerebral palsy) that has been diagnosed by a doctor?’ and if so, did the students think it restricted their attendance or participation at school. The children were then categorised into three groups: non D/CI (82% of sample), D/CI without restriction (13%), and D/CI with restriction (6%). Restriction, in this study, referred to whether the children experienced restriction in attendance or participation in school. Rose, Espelage, Aragon, and Elliot (2011) used a similar methodology to identify students with a disability. Students were asked whether they had a disability, if they knew what it was and if they could describe it, and if they attended special classes. Eighteen percent of the sample recorded themselves as having a disability. This is the method that was used for this pilot work due to the data protection act limiting access to the children’s diagnosis records in schools and the age of the children (11+) means they will be able to comment on their own SEND diagnosis.
3.1.4.2. Measuring bullying in an SEND sample. Very little research has
several researchers have worked around the challenges of working with children with SEND is to use qualitative approaches, e.g. interviews, in order to ensure that the children fully understand what they are being asked. Additionally, more innovative methodologies have been utilised for research with children with severe SEND, such as ASD, through the use of play and other activities with which children with ASD could engage.
However, for purely quantitative research, very little adjustments are made to take SEND into account. Many studies use standardised questionnaires, such as the My Life in School Checklist or the Olweus Bully/Victim Questionnaire; however, these scales may not be aimed at a reading/writing age appropriate for children with SEND, especially those with below average skills. Several researchers have recruited parents/teachers to complete surveys about the children’s experiences of bullying and/or victimisation. For example, Van Roekel, Scholte and Didden (2010) investigated bullying in adolescents with ASD and asked both the children and their teachers to complete a questionnaire. They found that teachers reported higher levels of bullying than the children, which could be due to a number of potential reasons: social desirability in the children, lack of understanding the questions, children not understanding they are being bullied, or the teachers reporting events that the students do not consider to be bullying, i.e. “banter” or teasing (van Roekel, Scholte & Didden, 2010). However, this paper only investigated children with ASD, a disability which commonly has an accompanying symptom of being unable to understand social situations, and so the difference between teacher and student reports may be due to this specific disability. Estell et al. (2009) also recruited both students and teachers to take part in research into bullying and victimisation, and the research took place in a
mainstream school and included children without disabilities and those with a variety of disabilities (learning disabilities, mild mental retardation, emotional and behavioural disorders, or general special needs). Estell et al. (2009) used a peer-nomination method and also asked teachers to complete a similar scale in which they rated each child on social skills competency. They found that teachers and students nominations were similar, with both groups reporting that children with disabilities were more likely to be bullies, however only teachers reported these children as also being victims (Estell et al., 2009).
When researchers want to rely entirely on student reports, there are a limited number of materials designed especially for an SEND sample. However, Fink, Deighton, Humphrey and Wolpert (2014) recently developed the Bullying Behaviour and Experience Scale specifically to measure both physical and verbal bullying and peer victimisation experiences in children with SEND. Fink et al. (2014) developed the scale with children with SEND and also carried out a readability analysis to ensure that the sentences were of an appropriate length, without too many words, and the individual words were not too long. The readability of this scale was approximately 8 years of age, which indicates that older children with weak reading ages would be able to access and understand this scale. As this scale has been specifically designed for use on students with SEND, it is the most appropriate to use for the current research to provide reliable and valid data.