In my study, I have listed several strategies taken in the area and finally ask them which are taken by a particular respondent.
I have asked them whether he/she has taken a particular strategy.
Responses were : Yes or No
Poverty was measured by using a Simple poverty scorecard (http://www.simplepovertyscorecard.com/BGD_2010_ENG.pdf).
there is not enough information about what you want to do and how you intend to do it. You also need to define terms., Best, D. Booth
David, is, of course exactly correct.
Most methodologists find that students think far too much and too early about statistical methods. If you spend some time precisely describing the adaptation strategies you intend, how they are to be measured, how poverty is to be measured, and something about the subjects you plan to use, we would have a chance at helping. I am confident that given that information, you will get what you need here.
Mr. John Morris and David Eugene Booth ,
I have asked them whether he/she has taken a particular strategy.
Responses were: Yes or No
Poverty likelihood was measured by using a Simple poverty scorecard
(http://www.simplepovertyscorecard.com/BGD_2010_ENG.pdf).
Sorry, this question was lost for a bit.
Do you have interest in each strategy, or in the total number of strategies that have been undertaken? In the later case you can sum the strategies giving a total number, then, assuming assumptions are met, correlate that sum with poverty. Alternatively, if you want to maintain the individuality of strategies, you could relate each of the dichotomous strategy variables with poverty using either a PPM correlation, or equivalently, a t-test. In this case, you would be well advised to consider one of several methods to maintain Type I error (e.g., Bonferonni, Holm, FDR, etc.). If you would like to consider the predictability of poverty from a model including all strategies, then you could do that with regression.
My understanding was that the criterion is a relatively continuous poverty metric, and that the predictors were each dichotomous in respect to each strategy -- undertaken, or not. If the criterion is dichotomous, then, as S. Béatrice Marianne Ewalds-Kvist suggest, LR (or discriminant analysis) would be plausible, but if not, and the intent is to predict a continuous criterion, with the information I have, I stick with my previous answer.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression