Download An Introduction to Causal Analysis in Sociology by Ian Birnbaum PDF

By Ian Birnbaum

Show description

Read or Download An Introduction to Causal Analysis in Sociology PDF

Best introduction books

An Introduction to LTE (2nd Edition)

Following on from the profitable first variation (March 2012), this e-book offers a transparent clarification of what LTE does and the way it really works. The content material is expressed at a structures point, providing readers the chance to understand the most important components that make LTE the recent subject among proprietors and operators around the globe.

Getting Started in Online Investing

This authoritative consultant to making an investment on-line, a part of a brand new growth of the relied on CliffsNotes model, can help new traders be successful.

Additional info for An Introduction to Causal Analysis in Sociology

Example text

When we consider causal analysis, however, we will find that this view is not applicable; and it is to causal analysis that we now turn. 30 I An Introduction to Causal Analysis in Sociology Causal Analysis Earlier, we saw how to best predict X4 from X 1 , X 2 and X 3. In our analysis we did not need to make any causal assumptions about these variables -our aim was simply that of prediction. Of course, our choice of X 1 , X 2 and X 3 was guided both by convenience and by our theoretical preconceptions: convenience, because we generally try to choose independent variables which are relatively easy to measure; theoretical preconceptions, because we try to choose variables which we feel have a strong association with the variable we wish to predict.

Hypothesis testing will be considered for causal analysis in Section 0). The computer, of course, makes all of this easy, but also extends the dangers of 'data-dredging'. Such procedures are best seen as exploratory only, results arising being regarded themselves as hypotheses which should be tested on other data (this is the method of replication). More will be said on all this in the sections dealing with causal analysis, and we leave until then a detailed consideration of assessing the separate influences of variables.

We must conclude, then, that standardisation has its uses, and hence that path coefficients interpreted as path regression coefficients have their uses as causal parameters. The variance interpretation of path coefficients, however, is not to be recommended. M Using Discrete Variables in Causal Analysis If an exogenous variable is discrete cardinal then the discussion is just as in Section G and no problem arises. Again, if an exogenous variable is noncardinal the discussion in Section G applies, with the conclusion that we should generally make an analysis in terms of sub-populations for each category of the non-cardinal variable.

Download PDF sample

Rated 4.37 of 5 – based on 28 votes