prev next front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |21 |22 |23 |24 |25 |26 |27 |28 |review
When all the data is clean, we can perform the Statistical Analysis as in the original protocol.
Always start with a description of the population, with histograms or crosstabulatin of the descriptive variables.
Descriptive variables are not tested.
Then compute the Prevalences, OddsRatios, etc which are the objective of the Survey. Rememeber to only test the significance of the Objectives and not of the Descriptives of the population studied. After the main analysis is finished, you can , and must, explore the data for facts that are curious, interesting or usefull. These are findings, they can not be tested in the clasical way, although the significance can be computed. All these findings fall into the category of New hypothesis generated by the study