I'm doing a small-scale research project into perceptions residents of an area have about their surrounding natural environment.
When we designed the questionnaire we set up a grid on Google Forms asking people to rank particular aspects according to their importance.
However, upon going into the field for our first pilot run (we printed the questions onto paper forms and approached residents in the street) we realised that a). some people didn't consider certain criteria we had selected as important, so couldn't really rank it relative to other criteria they did consider important, and b). as everyone was in a rush respondents tended to just treat it as though it were a Likert Scale anyway because the format was confusing.
My question is, how should we manage the data? Is it valid to keep the other responses and drop responses from the two questions that were resulting in confusion, or should all the data be scrapped? And, would it be valid on the responses we can tell were obviously entered as if on a Likert Scale, or would that be too presumptuous and hence invalidate the results?
The only reason I even ask is that we are quite pressed for time. We didn't get a huge amount of data but it was a few hours work and it would be a shame to start from scratch and lose all of the other good responses we got.
Does anyone have any thoughts on the matter?
When we designed the questionnaire we set up a grid on Google Forms asking people to rank particular aspects according to their importance.
However, upon going into the field for our first pilot run (we printed the questions onto paper forms and approached residents in the street) we realised that a). some people didn't consider certain criteria we had selected as important, so couldn't really rank it relative to other criteria they did consider important, and b). as everyone was in a rush respondents tended to just treat it as though it were a Likert Scale anyway because the format was confusing.
My question is, how should we manage the data? Is it valid to keep the other responses and drop responses from the two questions that were resulting in confusion, or should all the data be scrapped? And, would it be valid on the responses we can tell were obviously entered as if on a Likert Scale, or would that be too presumptuous and hence invalidate the results?
The only reason I even ask is that we are quite pressed for time. We didn't get a huge amount of data but it was a few hours work and it would be a shame to start from scratch and lose all of the other good responses we got.
Does anyone have any thoughts on the matter?
via International Skeptics Forum http://ift.tt/1bhqTc6
Aucun commentaire:
Enregistrer un commentaire