Going into this project, I knew how hard finding the data was going to be. I knew that the exact statistics I wanted would never be available to me, especially because the countries I chose (Ghana, Cote d'Ivoire, Niger, and Nigeria) aren't exactly known for their consistency in keeping up with data collection. So I accepted the fact that my data would come from 2011-2012 if it even exists. I managed to find data for my IV and DV, and that was all good for my simple linear regression. But then having to find data for my control variables, that has been a nightmare. What I've been doing is looking at the research available and forming my control variables off of that. Shouldn't it be the other way around? Honestly I feel like I am going in the complete wrong direction with my project. The biggest hurdle I am facing right now is finding any data about domestic violence in these areas. From what I have read, domestic violence is considered normal in the communities and is underreported. I could only find such data for Ghana, and even then it seems really low.
So here are my two questions, what kind of data do I truest and what kind of data would count as not accurate because of underreporting? Also, does anyone have any ideas about measuring power dynamics within a household that isn't domestic violence experienced by women?
So here are my two questions, what kind of data do I truest and what kind of data would count as not accurate because of underreporting? Also, does anyone have any ideas about measuring power dynamics within a household that isn't domestic violence experienced by women?
As far as what kind of data would count as accurate or not accurate because of underreporting, I think that you could possibly use existing data as long as you state that it is very likely that this is an underreported phenomenon and this may also skew some of your data, though that would probably not work based on what you said. As far as measuring power dynamics within a household that isn't domestic violence (and I'm not sure if this exists for your countries, but it's possible, just ideas) maybe try looking at child marriage rates (if a lot of young girls are married off, it's probably a huge power difference in the household), employment rates for women who are married or in general versus men (it may not correlate to power dynamics within a household necessarily though), or possibly look at gender relations in the country in general (education data, gender-based violence, discourse analysis/content analysis, employment data, etc.). It might be good to look at a combination of the domestic violence data that's available with another data set like one of the ones I suggested as opposed to just one thing, especially since there may be reporting errors.
ReplyDelete