I've seen a couple now so I know some tips to give you.
Think of it like a scientific method - describe each step you take and the decisions you have made about how you will collect your data.
Details are important:
- how much data? Link to an evaluation of how reliable the data sample will be (typical of the whole data pool e.g. if you collect three Beyonce songs and she has written over a hundred, how likely is it that those three are typical and therefore that the data is reliable? If you picked a different sample and did the same quantifications, would the results be likely to be roughly the same?).
- how will you cut any samples down and will this aid reliability (e.g. by not using the first couple of minutes of recording while the participants forget they are being recorded and thus lessen the effect of the observer's paradox)? Or comparability (if you are comparing something like political speeches and are quantifying the frequency of techniques, it might help to select extracts that are comparable in length to aid comparability)?
- What are the comparablity factors if you are comparing and how far can you isolate the variable that you are interested in - what have you done to maximise your comparabililty (e.g. starting each group off with the same introductory script)?
- What are the ethicality issues and how have you dealt with them e.g. what did you make sure of informing participants of? How have you anonymised data where necessary?
- What should you end up with? Make it clear what you are collecting, preferably in a table with useful codes to refer to each piece of data.
Good luck and don't forget to email me as per the lesson instructions a couple of posts further down.
No comments:
Post a Comment