i think the most important thing is to break the stereotype,You shouldn't look at what you're studying from your point of view or from the point of view of modern mainstream values.for example,when Guangzhi Zhang studying the inheritance law of Shang Dynasty,he pointed out that in the Shang Dynasty, "son" did not necessarily refer to the biological son, but could refer to all the male descendants of his next generation, including nephews.if you can't pay enough attention to this problem,this can certainly be a hurdle.

The over-reliance on interview forms of data collection are notoriously unreliable. We all know that people may say what they think the researchers wants to hear, what makes the informant look good, or just provide an answer to avoid having to admit not knowing the answer or to provide certainty when ambiguity might be a more realistic assessment. Qualitative data often suffer from these methodological problems. Ethnologists recognized these issues when it was the only method commonly employed technique, assuming that "cultural norms" are the reasons people do things. Fortunately, many behaviors can be observed and quantified. Observation can count the number seconds each component activity is performed and the number of bouts of those activities. This can be done ethnographically even before the investigator develops competence in the language of the population one is trying to learn from for a particular research question. The hard part is that this can require long fieldwork to observe different individuals, women and men, people of different ages, and to understand variation in seasonally or situationally appropriate behavioral responses. Behavior observation methods have been developed in animal behavior studies, and were somewhat commonly employed in anthropology in the 1980s-90s. There is a current backsliding to interview methods again, because it is quick. Many professors want students to finish their degrees in short periods of time, and it was grad students who usually had the time to spend 1 year or more in the field before taking up some academic or other professional employment that curtailed such long fieldwork. Additionally, evolutionary psychology has, perhaps too enthusiastically, popularized their view that interview data "reveals" evolutionarily-derived modern behaviors and garners headlines by asserting that all correlations are equivalent to identifying what activities have been acted on by natural selection and what the alleged deep-time outcomes are. Questionnaires will garner answers, but are they either accurate or relevant? I do believe the evolutionary psychology popularity has unfortunately influenced over-confidence in the utility of interview data and made such fast data collection attractive, to the detriment of opportunities to encourage researchers to collect more fine-scaled observations of what people do that do provide reliable, replicable, and exciting quantified views of what people actually do, often at odds with what they say they do.