Designing and planning research

A common, and costly, mistake is to leap in and begin working without designing what it is you will do and planning how to carry it out. Spend time on working this out and discussing it with your supervisor and any available experts.

After this, you will most likely need to submit for approval a brief proposal providing details of your intended studies. You should provide information about: topic; aims of the research, or the problem to be investigated; relevance to your discipline; and the central concepts of the study.

Identifying the central concepts of your study often means considering possible attitudes, experiences, behaviours, social processes, reactions - whatever is interesting and possible to examine. Take into account such factors as your own personality, age, gender, interests and work load; the participants, their available roles, the activities that will take place, the times at which things happen and the places in which they happen; the accessibility and availability of material, of members of a group or an organisation, of individuals to be interviewed or documentary sources.

Methods to be used
Many of these remarks apply across the board; and not to just one method. Indeed, it is difficult, and usually fruitless, to compartmentalise research methods. As with most aspects of PhD work, you arrive at the best method to use by asking questions.

Experimental and quasi-experimental studies.

What design is most appropriate?
What factors are to be controlled by you in the experimental design?
What factors are to be knowingly ignored, or dealt with by randomisation?
How is random allocation of treatments to experimental units to be achieved?
Where are the experimental units coming from?

Survey research

What form of survey is most appropriate for the task at hand: mail-out, face-to-face interview, telephone?
Who are the intended respondents?
How will the respondents be chosen-by a randomisation process, by a quota, by purposive selection, from administrative records, from an organisational hierarchy?
If randomisation has a role, how will it be introduced - by your, by a private survey organisation, by a governmental survey organisation?

Interview based studies

How many members of a group would you attempt to interview? How would you approach them initially?
What form will the interviews take - semi-structured, unstructured?
What would be the approximate duration of the interviews? Would you use a single interview per person or a series of interviews?
What topics would you attempt to cover in the interviews?

Participant observation

What role or roles would you adopt in the setting? What level of participation would you employ - complete, participant as observer, observer as participant?
What would be the best activities in the setting in which to participate? What would be the best physical locations from which to observe?
What use, if any, should be made of informants? How would you go about selecting them and establishing a relationship with them?

Textual analysis

What sort of textual material will you use - printed, graphical, audio or visual?
How will you obtain the texts to be analysed?
What other aspects of the text will you examine, eg, origins of the material?
How will you select a sample of texts to be analysed from the range of available material?

Conversation analysis

Will you record conversations for the analysis or use conversations recorded for other purposes?
If you record conversations for the analysis, how will you manage your role as observer? How will you obtain existing recordings? How will you gain the consent of the participants?
If you use conversations recorded for other purposes, how will you obtain existing recordings?

For each methodology, ask yourself how should the data be

recorded?
analysed?

These may not be simple decisions to make. But, you need to clarify the steps involved in each part of the study, and establish a timetable for each part.

Of course, it is also important to recognise that you may well bring personal biases to the study. Examine how these might affect the data collection process and your interpretation of the results of your research. There are invariable ethical implications to proposed research programs.

If possible, seek professional advice before setting about data collection, especially on statistical matters that you may be unfamiliar with. Most universities employ professional statisticians within their mathematics, economics, agricultural, biological and social science departments. These people have to be experienced in the real-life issues of data collection and data analysis relevant to their discipline. Use them.

	Looking at the fit between your aims and your research design.
	Dealing with data.
	Now I see how I should have done it all along. Is it too late to change?
	I am three quarters of the way through but it isn't as fruitful as I thought it would be. How can I rescue my thesis?
	Finding, formulating and exploring your topic.
	Deciding on your structure.
	Seeing a plot emerge.
	Developing a picture of the thesis as a whole.
	Preparing an outline for three years?
	Is it really necessary to start with a really detailed outline of your thesis?
	Seeking, receiving and handling feedback.
	Strategies for getting the best feedback possible.
	Overcoming reluctance to seek feedback.