Close Home Forum Sign up / Log in

Data collection - Interviews

E

Just wanted to know how many interviews you're thinking of doing in your PhD if this is your adopted form of data collection. I am intending to do between 20-30 by 2hrs. Does that sound robust? I know it depends on a variety of other factors like research issue, point of saturation, etc. However surely there must be a suggested range somewhere. For example how many cases is required for a Phd? Many would say 'how long is a piece of string', however Eisenhardt suggests 4-10 is reasonable regardless.

T

I have been going by saturation level. Done about 15 interviews, have about 2/3 more to go. I think 20- 30 is a lot at 2 hours each, that aside the transribing will be a killer unless if you have someone else to do it for you. having said, you are better off transcribing your own data as it helps you to get more familiar with it. I however get the impression that its more to do with the quality of data than the number of interviews.

P

I think I will have about 50 at between 1 to 2 hours each. Plus onbservations over 8 month period.
It does depend on your topic. Some people in my department have done 70 interviews, other have done 15. It depends what you do with them after.
I think your supervisor is the best person to guide you on this.

S

i'm planning for 10x3 interviews plus 10 others. that's three consecutive interviews with 10 different people, estimated at 1h each. i have other materials too.
but if i find "saturation" earlier, i'll not do the remaining planned interviews - that is, i'll fight for not having to. just because that's what i initially proposed doesn't mean i have to stick to it if it turns out it doesn't make any sense! (thinks me)

E

Thankyou all for the replies. Judging from what you three have said, i reckon 30 should be enough for one to find some form of saturation .

While transcribing is indeed pain-staking, doing it yourself really does get you closer to the data. Infact i find transcribing and making notes and coding should preferably be happening all at once

D

I am just finishing phase one of a longitudinal study (three interviews with each respondent over a 6 month period). Using existing similar studies I had calculated I needed 66 respondents at phase one in order to stave off high attrition rates and complete the study with 24.
In reality after 10 weeks I have only managed to get 30 respondents and now have to pray like hell they all remain to the end of the 6 months....otherwise I could have a very small sample!
I found I reached saturation on some areas of the interview quickly but the odd inquiry has continued to throw up new observations right to the end!!

C

some people like to put a number on it but i think it really depends on your own project. I'm personally aiming for about 25-30 interviews of about an hour each, but I don't think that I have to reach this target, or that if I do that it then shows I have succeeded in my data collection. There is a lot to be considered when deciding what/how/when/where is adequate.

9634