CHAPTER 4: RESEARCH METHODOLOGY
4.3 RESEARCH METHODS
4.3.1 Quantitative Data: Self-Completion Survey
By distributing a self-completion scoping survey prior to conducting in-depth interviews, I sought to collect a large amount of quantitative data on basic aspects of what is currently happening regarding HRE provision in English primary schools. The data was useful for three particular reasons. Firstly, from a research process point of view, it provided me with my sample for the follow-up interviews.10
Secondly, the collection of a large quantity of relevant data provided me with useful background information on the nature and extent of teaching in this area, thus enabling me to make tentative generalisations about HRE provision in England.11
And finally, I wanted to avoid having to ask my interviewees – whose available time was inevitably limited – simple, fact-finding questions. Having basic information about their teaching practice to hand thus allowed me to probe more deeply into the reasons for such practice and into their thoughts and opinions on HRE.
A copy of the survey is attached in Appendix 4, but in summary the questions served as indicators for establishing the following: (i) whether values are being taught in primary classrooms and what particular values are focused upon; (ii) whether education about, through and for human rights is currently being incorporated into classroom teaching; (iii) if so, how this is being done, and if not, why this is not being done; and (iv) whether teachers are referring to the international HRE framework and to human rights terminology. To increase the likelihood of a greater response
8 Verma, G. & K. Mallick, Researching Education: Perspectives and Techniques (The Falmer Press, London
1999), 122.
9 Bryman, A., Social Research Methods (4th edn, Oxford University Press, USA 2012), 645-646. 10 Ibid, 644.
rate, I kept the survey relatively short and used closed questions. Respondents were then provided with the opportunity to elaborate on some of their closed answers by way of a textual response box. Four of the survey questions required respondents to assess their teaching practice by determining either the accuracy of statements or the frequency with which specified topics are addressed in their classrooms using rating scales.
The survey was hosted by the website SurveyMonkey. I considered an online survey to be appropriate for the target population of primary teachers in England as all teachers are given a school email address. Sampling issues arising from discrepancies in internet access and availability did not therefore arise. In order to obtain the greatest number of responses and to build as representative a sample as possible, I engaged in extensive promotion and distribution of the survey link. I commenced this process by emailing every local authority in England requesting that they forward the survey link onto all maintained primary schools within their control.12 Only a
small number of local authorities responded saying that they were willing or able to do so.13 A greater number advised that they were unable to contact the schools
directly with the link but that I could contact the schools myself, and most provided me with a list of the schools under their control.14 For these local authorities, and
also for those that did not respond to my original email request, I then emailed individually the head teacher of each school within their control requesting that the survey link be distributed to all teaching staff members.
The survey was thus distributed to a reasonable section of the sampling frame15 – H.
Russell Bernard defines this as ‘a list of units of analysis from which you take a sample and to which you generalize’16 – which in my case was all maintained primary schools
in England. Though some head teachers may have acted as gatekeepers and neglected to forward the survey link on to their staff, the scope and volume of
12 I emailed only a selection of London Borough Councils. A maintained school is one that is funded
by the local education authority.
13 These were: Bath and North East Somerset, Buckinghamshire, Bromley, Cambridgeshire, East
Riding of Yorkshire, Peterborough, Portsmouth, Royal Borough of Kensington and Chelsea, Sheffield and Southend.
14 These were: Brent, Derbyshire, Devon, Essex, Gloucestershire, Harrow, Herefordshire,
Hertfordshire, Lancashire, Leeds, Somerset, Waltham Forest, West Sussex and York. This information is also publicly available through local authority websites.
15 With the exception of primary schools in some London boroughs.
16 Bernard, H.R., Social Research Methods: Qualitative and Quantitative Approaches (Sage Publications,
responses suggests that a large and diverse sample of teachers completed the survey. Figure 4.1 shows a relatively even split of respondents across the primary year groups17 – head teachers were also able to complete the survey and a small number
did so18 – and a table showing the counties represented by the respondents is
included in Appendix 5. Whilst certain local authorities clearly promoted the survey to a greater extent than others, responses were received from teachers in 28 of the 48 ceremonial counties of England.19 Those counties with the highest response rates
represented an appreciable geographic split: West Sussex [12%] in the South East; Gloucestershire [9%] in the West Midlands; Devon [8%] on the South West coast; and Lancashire [7%] in the North of England.
Whilst I did not actively distribute the survey link to primary schools not under local authority control, one teacher from an independent school did leave contact details for a follow-up interview, indicating that the survey had reached some such schools. I decided not to exclude this teacher from the follow-up interviews, particularly as her school followed the National Curriculum. However, as the survey did not query specifically whether respondents taught at local authority maintained schools, the
17 Regarding the data on school year taught, where teachers indicated that they taught more than one
year, they were included within the total for each year group specified in their response. Figure 4.1 thus represents percentages of a total of 515 year group entries and not percentages of the 374 individual responses to the survey question.
18 The nominal category of ‘Other’ [representing 0.4% of respondents] included a Head of Reading
and Planning, Preparation and Assessment and a specialist teacher in a Speech and Language Base.
19 Those counties not represented are: Bedfordshire, Berkshire, Cheshire, Cornwall, Cumbria, East
Sussex, Isle of Wight, Kent, Leicestershire, Lincolnshire, Merseyside, Norfolk, Northumberland, Nottinghamshire, Rutland, Shropshire, Suffolk, Surrey, Wiltshire and Worcestershire.
10% 13% 16% 13% 13% 13% 18% 4% 0% Figure 4.1 Early Years Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Head teachers
differences in approach between state-funded and private schools is not identified as a research variable within this study. Similarly, whilst the number of teachers from schools with a religious character represented in the qualitative interviews (19, or 43% of the sample) indicates that a significant proportion of survey responses were received from teachers in denominational schools, the survey did not specifically query this. Again, therefore, the differences in approach between denominational and non-denominational schools does not constitute a research variable in this study.20
Whilst my method of survey distribution did limit the scope for sample bias – emailing every school within a county if the local authority was unwilling to contact them directly – I acknowledge that the survey sample cannot be said to be representative due to sampling limitations and high non-response rates.21 For
example, some of the local authorities or head teachers who said that they would distribute the survey may have neglected to do so and, of course, a number of those to whom the survey was distributed will have chosen not to respond.22 As Alan
Bryman highlights, ‘[t]he significance of a response rate is that, unless it can be proven that those who do not participate do not differ from those who do, there is likely to be the risk of bias’.23 It is arguable, for example, that teachers with strong
opinions either in favour of or against HRE would be more likely to respond to my survey than those with less polarised views. The likelihood of teachers with moderate views responding was increased, however, through local authorities and school management directly requesting that teachers spare a small amount of their time to complete the survey.24
20 I am hoping to carry out more detailed research into the differences in HRE provision between
denominational and non-denominational schools in my future academic career.
21 For further information on sampling and sample bias, see Oppenheim, A.N., Questionnaire Design,
Interviewing and Attitude Measurement (2nd edn, Pinter Publishers, London 1992), 42-44; Bryman (n 9) at
199-206; & Corbetta, P., Social Research: Theory, Methods and Techniques (SAGE Publications, London 2003), 224-225.
22 For detailed discussion of non-response error, see Corbetta, Social Research (n 21) at 224-225. 23 Bryman, Social Research Methods (n 9) at 235.
24 I was copied in to some such emails and also advised by a number of head teachers that they would