Data Analysis

I finished my final interview today and closed my survey. I am really pleased with all the information I have gathered, but with that now comes the responsibility of doing something with it. The problem I am having is that social science is messy. Nothing is actually 'answered' because you are dealing with so many interesting, varied and equally valid opinions. How then do approach bringing everything to some kind of cohesive conclusion?

At the final campus session :( last Thursday (I will genuinely be sad not to see some of you guys again) this is exactly what we talked about so I thought I'd mention some of the things that were said in the hope that it helps you and, by reiterating it, helps me too.

When analysing interviews, which constitute the main part of my inquiry, Paula talked about two things. Firstly, it is important to mention common responses or information that typifies what has been said. If lots of people have mentioned it and agreed on it then it must be important. Secondly, it is important to mention things that are significant. It may only have been said by one person but if you think it is crucial information then include it. I think probably a word of warning here would be to make sure you are not including something that goes against the general trend just because it proves your point. Don't cherry pick your information.

I think it would also be interesting to compare and contrast your findings, not only with each other but also with your literature. Does what you have found out support what you've read or contradict it. For me personally my gathered data does a bit of both but then that is also important to highlight.

I feel one of the most important things that Paula said is that in your analysis make sure you are giving an interpretation of the data not your opinion. Obviously you will have thoughts about what you have researched but support these thoughts with quotes and statistics that give it weight and context.

Another important thing is that we have been collecting evidence around certain ideas, not proof. Particularly when talking about information gathered in surveys it is important to say that data is indicative rather than proof of certain things. For a survey to be statistically accurate you are supposed to have at least a thousand respondents. I don't know if anyone has managed that but I personally only managed to get ninety-one back. I am very pleased with this number and I think it is enough people to give an indication of trends and ideas but not enough to prove anything for certain.

Finally, I would say that everyone should look to their journals for things they have learnt and things that have changed. Journal entries do count as evidence of changed behaviour in yourself and others.

I hope some of this has helped. It strangely enough has clarified things for me. Keep up the good work everyone. It's not long now.