Ethnography:

I want to debunk a common preconception that ethnography is a qualitative method. Not true! Ethnography is a multi-methods approach. While it’s true that the primary method in ethnography is participant-observation, which is largely qualitative, I think the best uses of ethnography are where quant. and qual. are used collaboratively. Surveys questions should be informed by qualitative observations and interviews. Surveys can also be used to ‘check’ theories as they emerge through qualitative analysis.

Statistics:

I have begun to think that statistics are more mis-used than properly used. I don’t just mean election polls. A discussion today of the meaning of reliability and validity reminded me of an important point: findings can be both reliable and valid, but meangingless. Social scientists sometimes have a habit of hiding behind the technical jargon of statistics: sample sizes and methods, error, confidence intervals, etc. But when you break down their arguments, they too often draw causal relationships using data which are meaningless in the context of the study. Be on the lookout!