Potion wrote:
bascially; what is wrong with your method?
is it the sample? the survey? the interviewer or setting for which the survey was done? would another method be better for what youre trying to research?
few examples:
- related to the sample: the sample is too small, respondents refuse to answer (either by getting tired or provoked and therefore do not finish the survey), the respondents feel theyre not qualitfied to answer, the respondents have little to no interest in whats being surveyed
- regarding the questions: the questions were ambigous, leading or too closed; the choices listed dont match the question, the questions arent mutually exlusive (ex. 15-20, 20-25)
- problems with validity: does the data actually answer the hypothesis, have you measured what you want to measure with the right instrument
- problems with reliablilty: can you trust the data? can someone welse use your method and research the same phenomenon and their dataset and yours will "conform"