Cross-Cultural Research Methods

Cross-Cultural Research Methods

Carol R. Ember

Language: English

Pages: 236

ISBN: 0759112002

Format: PDF / Kindle (mobi) / ePub

Without ethnography, cross-cultural comparison would not be possible. But without cross-cultural comparison, we would know nothing of what may be universal or variable across human cultures, or why variation exists. Cross-Cultural Research Methods is an introductory teaching tool that shows students and potential researchers how to describe, compare, and analyze patterns that occur in different cultures, that is, how to form and test anthropological, sociological, psychological, medical, or political hypotheses about cultural variation.

Lifeworlds: Essays in Existential Anthropology

The Art of Not Being Governed: An Anarchist History of Upland Southeast Asia (Yale Agrarian Studies)

Beyond Aesthetics: Art and the Technologies of Enchantment

Near a Thousand Tables: A History of Food













with more or less an equal number of negative and positive statements were considered ambivalent; those with mostly positive statements were considered relatively low on men’s fear of sex with women; and those with only positive statements were considered as lacking men’s fear of sex with women. While the variable as operationally defined does not capture everything in a culture’s beliefs about heterosexuality (in all their uniqueness), it does capture some distinguishable similarities and

convey difference and ordinal scales only convey order. Means and median scores are best reserved for interval and ratio scales. If you wish to summarize nominal scores or grouped ordinal scores for a set of cases, it is best to use frequencies or percentage summaries. A modal score can also be given for the scale score with the highest frequency. One other common kind of descriptive statistic is a measure of variability. This kind of measure helps us understand how much the scores are spread

statistical significance of a result. (Consult a statistics book for the formula for computing chi-square and a table in which to look up the p-values for different values of chi-square. You will also need to know the degrees of freedom in a table. A two by two table has one degree of freedom because, given the row and column totals, once you know the number in one cell, you can calculate all the other cell values.) It is important to compute the test of significance first because the strength of

relationship were zero. In our example, the r could be positive or negative, but we would look for the onetailed significance because the direction of the relationship was predicted. The p-value for our hypothetical data set is <0.0005, one tail. This means that the likelihood of there being no negative linear relationship is less than five times in ten thousand. What if the relationship is not linear? Figure 8.5 shows an example of a nonlinear relationship. If we didn’t plot our data but just

asked for a Pearson’s r, we would have gotten an r of 0.00 because the line of best fit in figure 8.5 is flat. (A flat line means that the best predictor of Y for each value of X is the mean of Y.) If the variation on X doesn’t help us predict the variation on Y any better than using the mean of Y, there appears to be no relationship between X and Y. But concluding that there is no relationship is obviously incorrect. There appears to be a strong relationship, but the nature of the relationship

Download sample


About admin