I am very disturbed after receiving from a the first final draft submited to my supervisors. Among the points highlighted the following worries me:

8. You state on p.23 that you are going toemploy a range of statistical techniques. In actual fact, I only noticed 2instances of cross tabulation. The rest is pure descriptive stats, and I have aproblem here- percentages and frequencies are somewhat thin for a doctorate. Iam convinced that the thesis can be strengthened a great deal by properstatistical analysis- pls consult the statistician.
9.Also, using only a questionnaire at doctoral level is thin. I think it isimperative that you run at least two focus groups, in which you can discuss anddebate the results of the questionnaire. This will strengthen the thesis agreat deal..




ANY ADVICE IS WELCOME

mamara

Hmm these sound like issues that should have been raised with you before you got to the write up stage. Did they not mention them before now?

Thank for your response. It was never brought up. After all these critic he ends up with this note:

"I am impressed with the work which you have done- it is a good piece of work, and will certainly make a contribution.If you attend to 1-10 above, it will be a very good piece of work".

How do I get "focus groups" involved at this stage?

regards

This sounds like quite a bit of extra work needed for point 9, have you spoken to supervisors about how best this can be acheived given the possible time restrictions? Re-running stats is a pain and I have had to do this to repsresent further significances ie p=0.000 explained as p<0.001 or p<0.0001, so this is annoying but not major. Point 9 seems to be running extra work over and above your original protocol which they must have approved. I would suggest having a discussion with them asap to agree the best way forward. Otherwise do take the well deserved praise that it is good research with a contribution, as that is half the battle won.

I agree that it should have been brought up sooner but I also agree that purely descriptive stats are insufficient (in the uni I worked at they are considered inadequate for an undergrad dissertation). Increasingly at PG level multiple methods or multiple studies (eg multiple groups, multiple locations or multiple experiments) are being required

Hi Mamara- I can imagine you must be stressed out after having this feedback. To start adding focus groups now seems to be quite a major thing to do. How long do you have left to submit? I would wait and see what your main supervisor suggests before getting too wound up. Bear in mind there is a difference between what would be nice to include in your PhD and what is absolutely necessary to pass. If I hadn't put my foot down several times with my supervisor I would testing participants well into my fourth year. Finally we have separated what I/she would like to do in the grand scheme of things, and what I actually need to do to pass my PhD. However if your other supervisor also recommends doing this then you might not have much choice- it's better to take time now and make changes then getting to the viva and having to resubmit (although this does seem like the kind of thing they should have mentioned ages ago, like about 2 years ago!). I can't really comment on your stats as I know nothing about your project. Certainly in my subject I have to use quite complex stats and would never be able to just present descriptives, but it could be different for different people/topics etc. Really I think you need to look at the aims of the study and see whether your analysis addresses them. It might be worth seeing a statistician to explore your options...I hope you get it sorted out. I think the main thing is find out what they think you need to do to pass, not what they would like you to do just for the hell of it! Best wishes, KB

Mamara,

Sorry to hear you are having this dropped on you when you felt you were so far into the process. As others have said it sounds like something that should have been raised by supervisors much earlier.

You ask what you should do regarding the two issues - firstly the implied weakness in your stats; it may be that depending upon what you are looking at that your existing data can be analysed differently. Take up the suggestion of explaining what you have and are trying to do with a statistician. Some additional number crunching may well provide greater robustness to your results.

The other point regarding needing more data or data from a different angle (focus groups) will surely depend upon what your null hypothesis is, how large your sample and how generalisable your data is. You must sit down with your supervisory team and have a serious talk. It may be worth mentioning the concerns to the statistician who may be able to make comments about the robustness of any answers you are getting from the existing data.

I think you should take comfort from the fact that they are suggesting you do more statistical analysis not that you collect more statistics. What stats package are you using? Do you know an expert who can help you to produce the additional stats? I don't think there is anythign wrong in getting additional help with the stats. It sounds like the focus group is a good idea - can you set something up fairly quickly?

I would definitely go back to your supervisor and ask for advice on how to take this forward. It sounds like it is a slight set back (and obviously not something nice at this stage but much better to find out now than have it thrown at you at your viva)

Hi Mamara, I agree with the others. It's a bit late in the day for a fundamental shift in your methodological design. Collecting questionnaire-based data is fine for PhD research. It's a very common tool for data collection. What was your research very, very broadly about (so as to maintain anonymity)? Is it a quantitative study? If so (in my view), it wouldn't be a very good idea to start collecting qualitative data with focus groups. Or did you questionnaire collect both quantitative and qualitative data? It would definitely be worth seeking the advice of a statistician to see if there is any further inferential statistics that you can carry out. But, because it's so late in the day for your research and the design has already been implemented in a particular way, there might not be much that can be done in the way of 'powerful' stats.
Try not to worry about things until you have discussed the matter with the statistician and your supervisor further.

Hi all,
while awaiting feedback from the core supervisor, I asked my the co-supervisor the following:
who should make up the focus group?
and the aim of the focus group?

He said the focus group should be made of relevant role players-probably people in the original sample.

The question I was shy to ask him is: How do I integrate the findings from the focus group into the results of the question? What do I do if the findings differ.


Thanks in advance


Regards
Mamara