urgent question
Surely two sources are better than one? But that's not my field. How bout sticking the results of one in an appendix?
I agree with Eds in that a second sample set to prove your argument is always better than a single sample. If the time constraints are days rather than weeks for the analysis then could you submit a little late?
The analysis of interview data would take a month at least, and it was intended to reply to and support a secondary research question. That is why I was thinking to drop it. But I don't know if it is possible, and I am hesitant to ask the supervisor because I am worried that he gets the impression that things are not under control and that might weaken my argument in front of him. Some pieces of advice and thoughts plssss
Is it that you don't have the time for the second data set or that you think it's too much effort to add without it improving your thesis. I'm not sure I understand the problem properly, how long until you submit?
Is it that you don't have the time for the second data set or that you think it's too much effort to add without it improving your thesis. I'm not sure I understand the problem properly, how long until you submit?
Presumably there was a rationale for doing both. I would imagine the interviews were there to provide qualitative insight to support or refute the quantitative data.
Whether this is needed, useless, or can be omitted depends very much on your hypothesis and research design. If your survey metric has been validated though other correlations, is well-designed, and has a sample appropriate for your analysis techniques (i.e. is not a fudged ANOVA of 30 Likert responses), then it may stand alone as a contribution worthy of a PhD. Similarly you can build a PhD purely on qualitative work but will have to defend other questions such as how representative the sample is, and how findings might generalise.
Surely if you have the data, a survey would be quick to analyse (though, admittedly, if you have 1,000 hand-filled papers rather than a SurveyMonkey report I'd see the problem). Since you've already undertaken the qualitative work, then if time really is an issue an option might be to pick several interviews as detailed case studies rather than report them all? If transcription is the problem then see if you have budget as a student to pay for this using a service.
Presumably there was a rationale for doing both. I would imagine the interviews were there to provide qualitative insight to support or refute the quantitative data.
Whether this is needed, useless, or can be omitted depends very much on your hypothesis and research design. If your survey metric has been validated though other correlations, is well-designed, and has a sample appropriate for your analysis techniques (i.e. is not a fudged ANOVA of 30 Likert responses), then it may stand alone as a contribution worthy of a PhD. Similarly you can build a PhD purely on qualitative work but will have to defend other questions such as how representative the sample is, and how findings might generalise.
Surely if you have the data, a survey would be quick to analyse (though, admittedly, if you have 1,000 hand-filled papers rather than a SurveyMonkey report I'd see the problem). Since you've already undertaken the qualitative work, then if time really is an issue an option might be to pick several interviews as detailed case studies rather than report them all? If transcription is the problem then see if you have budget as a student to pay for this using a service.
Thanks a lot for this valuable comment. I will think it through.
Yes often functional costs like that can be put down to the department- a bit like your photocopying (probably!).
Do you mean validity in instrument/questionnaire design or qualitative validity? I'm doing quantitative research developing an instrument. im a little worried that you are unsure of measuring validity at this stage of your research. Qualitative/quantitiative validity is very different and greatly depends on what you are doing. I'm developing an instrument and looking at construct validity using confirmatory and exploratory factor analysis (also a stage before this developing items to satisfy content validity) and criterion validity at the very least where you validate your instrument with a valid and reliable measure. Reliability is also important with questionnaire design so looking at internal consistency (cronbach's alpha). Otherwise you would be analysing responses of the questionnaire means, standard deviations etc. Who is providing guidance for you? I need more information to help you
