0

Sorry about the breadth and complexity of this question. I realise that for most people this question irrelevant as most projects or PhDs are predetermined and that theories and methodologies for each discipline or topic are quite limited. However, for me that is not the case. My research topic is rather broad, and I was keen to explore the topic from different angles before deciding on a specific angle. When the angle was decided, it was not clear which theory to use and whether the project should be one of theory development or theory testing. I have since decided which theory to use but I would like to apply more intellectually rigour to my project.

Anyways, my broad question is directed to people and researchers who have experience or had to decide which theories to use for their project or PhD. What influenced your decision-making and what factors were relevant? I could not find any books or resources that helped me go through my consideration process, so I apologise for this question.

More specifically my topic is broadly about new migrants. I was keen to explore whether migrants who are new to Australia found or felt there are educational differences in the way they are taught in the workplace. Obviously, learning theories are relevant, acculturation theories less so, etc.

Scott Seidman
  • 31,120
  • 4
  • 52
  • 121
Poidah
  • 3,383
  • 12
  • 28
  • 2
    What does your advisor say? – henning Nov 21 '19 at 07:41
  • They are supportive of the theory I have selected. I am still in the process of discussing it. – Poidah Nov 21 '19 at 07:49
  • It is a bit out of their area of expertise and I have tried canvas across disciplines but I am not feeling totally confident and satisfied. – Poidah Nov 21 '19 at 08:06
  • It is difficult to answer your question as this is one of the main difficulties of doing theoretical research. For me, it has always been a mixture of what looks more promising and what looks more interesting/"fun" to work with. – wimi Nov 21 '19 at 08:10
  • Good point @wimi. I come from a science and more positivist training so the "mixture" is less appreciated. Fun and interesting can be important points. Thanks wimi – Poidah Nov 21 '19 at 08:35

1 Answers1

1

Whenever you have a problem you can't solve, solve any part of it, then take another look at the problem and see if the partial answer has helped. (I can't find the exact source of this formula, but I think it's R.A. Heinlein.)

First, what sort of methodology are you considering? Interviews? Surveys? Online or in person? What sort of statistics are you contemplating, if any? How will you assure you have something representative rather than the result of a small number of people with unusual opinions? How will you select your subjects? Is there any method to deal with self-selection? For example, people who are unusually happy or unusually unhappy are possibly more likely to want to talk about their experiences.

You write of differences. How will you determine the situation before they migrated in order to observe the difference? Suppose, for example, that you had a significant number of people with very little work experience outside of Australia, maybe because they migrated in the age range of 10 to 18, and so didn't have jobs previously. How would you discern "coming of age" factors from "changing culture" factors? Possibly you need to do some research on the situation in the places the migrants migrated from. Possibly you need to compare to the perception of experiences of non-migrants in the same age range. Possibly you need to control for financial status, education status, etc.

How will you detect the just so story in your work? If you do detect it, what will you do about it?

There are a bunch more such questions you could ask yourself. I'm a humble safety analyst, so I know very little about sociological studies. I do know something about how to avoid fooling yourself.

Once you have some notion of that, then you can look at theories that you could possibly test using such methods. Then you can refine your methods to target specific hypotheses. Be sure that the research you propose has a reasonable expectation of actually testing your hypotheses. Also, have some thought as to whether there might be alternate hypotheses that have similar explanatory power.

And after you've done all that then stand back and critically examine the whole thing. Maybe there are other methods that would be a similar amount of work (or even less work), but that would test these hypotheses more accurately. Maybe there is already a block of research that tests them. Maybe there is some reason to think some of your hypotheses are wildly wrong before you even start.

puppetsock
  • 1,077
  • 5
  • 7
  • Thanks puppetsock. I appreciate the perspectives and it is helpful to think about those different ways. – Poidah Nov 21 '19 at 21:26