1. Researching and Writing Dissertations Roy Horn Researching and Writing Dissertations

2. Roy Horn Researching and Writing Dissertations Data Analysis WEEK 8

3. Researching and Writing Dissertations Roy Horn Reading Recommended text: Researching and Writing Dissertations Horn, R CIPD: London 2 nd Edition, 2012 ISBN: 978 184398 302 6 Chapter 8: Data, and How to Analyse It, page 146. Chapter 9: Data analysis and Representation, page 169.

4. Researching and Writing Dissertations Roy Horn Lecture outline l How to prepare and validate data. l How to put data into an electronic form suitable for analysis.

5. Researching and Writing Dissertations Roy Horn Learning objectives l Understand how to validate data before entry. l Understand the basic descriptive measures used with quantitative data. l Be able to represent data in appropriate ways.

6. Researching and Writing Dissertations Roy HornActivity Experiential Activity Consider the type of data that your research will generate. How are you intending to analyse this data? Work alone 10 minutes prep 2 minutes feedback Feedback to the group – ‘my research will generate…and I will analyse it...’

7. Researching and Writing Dissertations Roy Horn Coding closed questions You will need to consider ways to code your data for entry into a computer system. Each response to a question needs an individual code that allows for data analysis. For example, a common question is ‘What gender are you?’ This could be coded as Male or Female in a spreadsheet. However, subsequent analysis is easier and more accurate if a code is used – say: 1 Male. 2 Female.

8. Researching and Writing Dissertations Roy Horn Coding closed questions There must also be a code for a non-response or a missing response, and with gender it is sometimes useful to know the reason for a non-response. So the full coding of the question would be: 1 Male. 2 Female. 3 Would prefer not to respond. 4 Unwilling to be characterised in this manner. 5 This information is confidential. 9 Missing. Note that 9, 99, 999 or 9999 is ordinarily used as the code for missing data.

9. Researching and Writing Dissertations Roy Horn Coding closed questions If you are intending to use a secondary data source and carry out comparative analysis with that source, then coding your own research using the same coding system will make the comparison considerably easier. If you are collecting data on the socio-economic class of your respondents, you could use the standard national statistics classification, then any subsequent analysis with other data sets will be easier.

10. Researching and Writing Dissertations Roy Horn Coding closed questions Analytic socio-economic classes 1 Higher managerial and professional occupations. 1.1 Large employers and higher managerial occupations. 1.2 Higher professional occupations. 2 Lower managerial and professional occupations. 3 Intermediate occupations. 4 Small employers and own-account workers. 5 Lower supervisory and technical occupations. 6 Semi-routine occupations. 7 Routine occupations. 8 Never worked, and long-term unemployed.

11. Researching and Writing Dissertations Roy Horn Coding closed questions For a scaled question such as ‘How enjoyable did you find your university studies? Answer on the scale 1–6’, answer codes might be: Very enjoyable 1 Quite enjoyable 2 Mostly enjoyable 3 Not very enjoyable 4 Not enjoyable 5 Dreadful 6

12. Researching and Writing Dissertations Roy Horn Coding closed questions Likert scales are used to test the agreement or disagreement of a respondent to a proposition. A Likert scale question is coded in the same way as other scaled questions. Agree 1 Tend to agree 2 Tend to disagree 3 Disagree 4

13. Researching and Writing Dissertations Roy Horn Coding closed questions If you are using survey or questionnaire software, such as Snap, the data coding is done automatically. When the data is entered into a spreadsheet the cells will contain only the code number. A narrative reminder of the coding of the question is often present in the column heading. When using SPSS you set up the variables and their characteristics as you enter the data.

14. Researching and Writing Dissertations Roy Horn Coding open questions Even in a questionnaire-based study it is common to find open questions – those that do not have predetermined answers. The most commonly-occurring open question is represented by the word ‘Other’, which allows for a response that has not been predetermined.

15. Researching and Writing Dissertations Roy Horn Coding open questions For example: Q What motivated you to join your current employer? Mark the most important motivator! (one answer only) Salary 1 More responsibility 2 Employer’s reputation 3 Desire for a change 4 Improved fringe benefits 5 More interesting job role 6 More flexible working arrangements 7 Other – please state 8.....................

16. Researching and Writing Dissertations Roy Horn Coding open questions Note that the ‘Other’ response is still coded 8 in this example, but will also have a text entry column in the data table for the other motivation(s). What sort of things might be entered in Other? Such things as: l My partner moved jobs to a different part of the country. l The transport links to my old job were just too poor.

17. Researching and Writing Dissertations Roy Horn Coding open questions Continued… l I wanted a shorter journey to work. l Two of my friends worked at the same company. l My mum is the managing director. l I worked here before and I really liked it. l This job is completely different from my last job. I was a teacher and now I am an administrator.

18. Researching and Writing Dissertations Roy Horn Coding open questions The most common way to deal with open data is to create code categories for the responses, so that the data above is entered into electronic form as: Geographical change of area 11 Transport problems 12 Shorter journeys to work 13 Social connection to existing workers 14 Kinship relationship to existing worker 15 Previously employed at the organisation and enjoyed the experience 16 Change of career 17

19. Researching and Writing Dissertations Roy Horn Coding open questions One main question often asked of supervisors is ‘Do I set up a category for every response?’ This is an important question and requires a measure of balance in answering. If you simply code every response with a new category, then you will have done little more than list all the responses, which was the original unanalysed position.

20. Researching and Writing Dissertations Roy Horn Coding open questions However, if you encounter a similar response more than once – let’s say: l The buses are terrible to my current place of work. l The traffic to that part of town is unimaginably bad. l The train line to work is dangerous and I don’t feel safe. l The bus stop is about a mile away from work, and I am fed up with getting drenched walking the last mile.

21. Researching and Writing Dissertations Roy Horn Coding open questions These could all be coded as ‘Transport problems’ and the number of responses recorded. The data in a spreadsheet can be ‘back’ coded after analysis. So the column after an ‘Other’ response contains the narrative detail and then the next column the analysed code, eg 11 Geographical change of area.

22. Researching and Writing Dissertations Roy Horn Validating data There are two stages to this process. ‘In the field’ is a research term used to describe the actual collection of data. In dissertation research you are very likely to be the person collecting the data. Field editing, or checking the validity of the data at the point of entry, is vital. The key issues in field editing are: l To get accurate answers. l Record an answer to every question. l Ensure that any writing is legible.

23. Researching and Writing Dissertations Roy Horn Validating data If you are asking the respondent the questions, plan for about two to three minutes between the respondents so that you can check and correct the entries on the completed form. You will be surprised how often after only a few seconds you cannot quite read or remember the answer just given. When the respondent is self-completing the questionnaire, you must check the form before the person walks away.

24. Researching and Writing Dissertations Roy Horn Validating data It should only take a minute to look over the form but if anything is not clear or is missing, you will have the chance to ask if your respondent missed the question or chose not to respond. You may also need to ask the question: ‘What does that say?’ Check through this list after every respondent’s questionnaire completion: l All questions have been answered or a non- response given. l Answers are in the correct format.

25. Researching and Writing Dissertations Roy Horn Validating data Continued … l Inappropriate answers have been identified and corrected. l Answers in the wrong place on the form have been replaced correctly. l Open questions’ answers are legible. l Biographical data is present and ‘visually’ looks correct. l Any note the interviewer has made is legible.

26. Researching and Writing Dissertations Roy Horn Validating data Office editing involves a more thorough check of the questionnaire for the same items in the checklist above. This editing process can make small corrections to ‘obvious’ errors such as marking the answers in the wrong format, or putting the answer in the wrong place, or using the wrong data format.

27. Researching and Writing Dissertations Roy Horn Validating data But this editing process should never be adding missing answers that you think the respondent would have made – in academic dissertations this would be regarded as a form of plagiarism.

28. Researching and Writing Dissertations Roy Horn Validating data Errors can occur when data is entered and once entered are quite difficult to find and remove. Standard transposition errors can occur frequently and create havoc in your data. A transposition error occurs where the entry of two digits is reversed.

29. Researching and Writing Dissertations Roy Horn Validating data Because questionnaires use codes for data, there are plenty of opportunities to create havoc. The code 12 is certainly a possible code for a question with a large number of possible answers. The code 21 is not likely to represent useful data. When entering data on a numeric keypad it is very easy to enter 21 for the code 12.

30. Researching and Writing Dissertations Roy Horn Validating data Most of the data analysis software can set up ‘input masks’ that significantly reduce the possibility of entering data wrongly. An input mask restricts the data that can be entered to a predetermined range. So if a question has 14 answers and uses 15 for missing data, the input mask can restrict the entry of data in that cell to the range 1–15.

31. Researching and Writing Dissertations Roy Horn Data analysis Questionnaire data analysis is normally undertaken using one of the following software products: l E xcel spreadsheet. l Access database. l Survey software, such as Snap. l SPSS software.

32. Researching and Writing Dissertations Roy Horn

33. Researching and Writing Dissertations Roy Horn

34. Researching and Writing Dissertations Roy Horn

35. Researching and Writing Dissertations Roy HornActivity Case Study Carry out one of the case studies. ‘The Great Book-Buying Survey’ or ‘Why I Buy Books’. Work in groups of three 180 minutes 20 minutes feedback Feedback your reflections using four PowerPoint slides

36. Researching and Writing Dissertations Roy Horn Preparation activity for next week Writing my dissertation Prepare two slides of PowerPoint presentation setting out: 1. My current style of writing looks like this... 2. Set out the time scale for your writing. Work alone Prepare for next week’s workshop

37. Researching and Writing Dissertations Roy Horn Conclusion l Consider how you will validate and enter data from your dissertation. l Spend some time considering what will be the best software for your dissertation data and its analysis.

38. Researching and Writing Dissertations Roy Horn Next Week l Developing good writing habits. l The macro/micro approach to writing. l How to use creative thinking. l How to write effective paragraphs. l Linking and signposting. l How to structure and format your dissertation. l How to avoid errors in your argument. l Referencing and bibliographies. l Effective introductions and conclusions.

39. Researching and Writing Dissertations Roy Horn Researching and Writing Dissertations The End