Choosing appropriate research methodologies and methods

  Choosing appropriate research methodologies and methods

Introduction
It is vital you pick approach research methodologies and methods for your thesis - your research afterall is what your whole dissertation will rest on.

 Choosing qualitative or quantitative research methodologies
Your research will dictate the kinds of research methodologies you use to underpin your work and methods you use in order to collect data. If you wish to collect quantitative data you are probably measuring variables and verifying existing theories or hypotheses or questioning them. Data is often used to generate new hypotheses based on the results of data collected about different variables. One’s colleagues are often much happier about the ability to verify quantitative data as many people feel safe only with numbers and statistics.
However, often collections of statistics and number crunching are not the answer to understanding meanings, beliefs and experience, which are better understood through qualitative data. And quantitative data, it must be remembered, are also collected in accordance with certain research vehicles and underlying research questions. Even the production of numbers is guided by the kinds of questions asked of the subjects, so is essentially subjective, although it appears less so than qualitative research data.
Qualitative research
This is carried out when we wish to understand meanings, look at, describe and understand experience, ideas, beliefs and values, intangibles such as these. Example: an area of study that would benefit from qualitative research would be that of students’ learning styles and approaches to study, which are described and understood subjectively by students.


Using quantitative and qualitative research methods together
This is a common approach and helps you to 'triangulate' ie to back up one set of findings from one method of data collection underpinned by one methodology, with another very different method underpinned by another methodology - for example, you might give out a questionnaire (normally quantitative) to gather statistical data about responses, and then back this up and research in more depth by interviewing (normally qualitative) selected members of your questionnaire sample.


Research methods in brief:
Look at the very brief outlines of different methods below.Consider which you intend using and whether you could also find it more useful to combine the quantitative with the qualitative. You will be familiar with many of these methods from your work and from MA, MSc or BA study already.
  • Qualitative research methods include
  • Quantitative research methods
Qualitative research methods
  • Interviews
Interviews enable face to face discussion with human subjects. If you are going to use interviews you will have to decide whether you will take notes (distracting), tape the interview (accurate but time consuming) rely on your memory (foolish) or write in their answers (can lead to closed questioning for time’s sake). If you decide to interview you will need to draw up an interview schedule of questions which can be either closed or open questions, or a mixture of these. Closed questions tend to be used for asking for and receiving answers about fixed facts such as name, numbers, and so on. They do not require speculation and they tend to produce short answers. With closed questions you could even give your interviewees a small selection of possible answers from which to choose. If you do this you will be able to manage the data and quantify the responses quite easily. The Household Survey and Census ask closed questions, and often market researchers who stop you in the street do too. You might ask them to indicate how true for them a certain statement was felt to be, and this too can provide both a closed response, and one which can be quantified (30% of those asked said they never ate rice, while 45% said they did so regularly at least once a week... and so on).
The problem with closed questions is that they limit the response the interviewee can give and do not enable them to think deeply or test their real feelings or values.
If you ask open questions such as ‘what do you think about the increase in traffic?’ you could elicit an almost endless number of responses. This would give you a very good idea of the variety of ideas and feelings people have, it would enable them to think and talk for longer and so show their feelings and views more fully. But it is very difficult to quantify these results. You will find that you will need to read all the comments through and to categorise them after you have received them, or merely report them in their diversity and make general statements, or pick out particular comments if they seem to fit your purpose. If you decide to use interviews:
  • Identify your sample.
  • Draw up a set of questions that seem appropriate to what you need to find out.
  • Do start with some basic closed questions (name etc.).
  • Don't ask leading questions.
  • Try them out with a colleague .
  • Pilot them, then refine the questions so that they are genuinely engaged with your research object.
  • Contact your interviewees and ask permission, explain the interview and its use.
  • Carry out interviews and keep notes/tape.
  • Transcribe.
  • Thematically analyse results and relate these findings to others from your other research methods.

Quantitative research methods:
  • Questionnaires
Questionnaires often seem a logical and easy option as a way of collecting information from people. They are actually rather difficult to design and because of the frequency of their use in all contexts in the modern world, the response rate is nearly always going to be a problem (low) unless you have ways of making people complete them and hand them in on the spot (and this of course limits your sample, how long the questionnaire can be and the kinds of questions asked). As with interviews, you can decide to use closed or open questions, and can also offer respondents multiple choice questions from which to choose the statement which most nearly describes their response to a statement or item. Their layout is an art form in itself because in poorly laid out questionnaires respondents tend, for example, to repeat their ticking of boxes in the same pattern. If given a choice of response on a scale 1-5, they will usually opt for the middle point, and often tend to miss out subsections to questions. You need to take expert advice in setting up a questionnaire, ensure that all the information about the respondents which you need is included and filled in, and ensure that you actually get them returned. Expecting people to pay to return postal questionnaires is sheer folly, and drawing up a really lengthy questionnaire will also inhibit response rates. You will need to ensure that questions are clear, and that you have reliable ways of collecting and managing the data. Setting up a questionnaire that can be read by an optical mark reader is an excellent idea if you wish to collect large numbers of responses and analyse them statistically rather than reading each questionnaire and entering data manually.
You would find it useful to consult the range of full and excellent research books available. These will deal in much greater depth with the reasons for, processes of holding, and processes of analysing data from the variety of research methods available to you.

 Developing and using a questionnaire - some tips
  • Identify your research questions
  • Identify your sample
  • Draw up a list of appropriate questions and try them out with a colleague
  • Pilot them
  • Ensure questions are well laid out and it is clear how to 'score them' (tick, circle, delete)
  • Ensure questions are not leading and confusing
  • Code up the questionnaire so you can analyse it afterwards
  • Gain permission to use questionnaires from your sample
  • Ensure they put their names or numbers on so you can identify them but keep real names confidential
  • Hand them out/post them with reply paid envelopes
  • Ensure you collect in as many as possible
  • Follow up if you get a small return
  • Analyse statistically if possible and / or thematically

0 comments:

Post a Comment