Posts Tagged: statistics

Research plans and data

Does the research plan ever survive contact with data or actual use? Sometimes seems that it does not. Even plans which seemed a good idea at the time you may find ‘not quite working’ when you actually go to analyse your data. Spent a few hours lately trying to work out the best way to analyse data. What I *thought* I would do simply wasn’t working out. Effective and careful planning helps resolve many of the problems with our research. More time taken in the planning, particularly focusing on how the data will be analysed, and the biases or limitations, will pay dividends later in the research project. Then – always check the assumptions of the statistical tests and make sure you haven’t violated any of the crucial assumptions. It’s alawys fun. Working hard at the start of the project does relieve plenty of pressure near the end of the project. A little time invested at the start, with a little support, makes everything much easier later on and also ensures that you have considered OTHER factors which may still become important, if the research plan changes a little. So – plan carefully, design your research method carefully, and ensure that you understand the assumptions of any tests you will use and why you will use these tests.