Assignment 4: Performing quantitative variationist analysis due 31 March at start of class
This assignment is about performing quantitative variationist analysis.
In other words, this is where you get ‘down and dirty’ with sociolinguistic data.
The variable: The one you used for assignment 3. The data: The LING495 mini-corpus. This assignment is worth up to 35% of your final grade for the course; it takes the format of a research report/paper, using empirical data and methods.
1. Circumscribe and code your data following the protocol you devised for assignment 3, working in any revisions Caroline and I suggested.
2. Check your coding for errors and for methodological consistency.
3. Perform distributional analyses of your results.
I. What is the overall distribution of forms?
II. What are the distributions according to each social predictor?
III. What are the distributions according to each linguistic predictor? As part of this exercise, be sure to cross-tabulate all the predictor groups.
Cross-tabs are an exercise in being aware of your data and spotting interactions. If two groups are interacting, you need to either remove one (if they are not orthogonal) or collapse the two to account for the interaction. Either way, you need to motivate your decision(s).
What you hand in: A report of your findings that minimally includes (see handout):
• an introduction outlining your variable and your research question(s)
• a methodology section (based on assignment 3)
• your results (the overall distribution and the factor-by-factor analyses**); include short discussions of your results
• a brief conclusion summarizing your findings; include a discussion of how your results fit in with previous research, both in general and with respect to the variable you have studied e.g. If you found that men use a certain variant more than women, what does this suggest? e.g. If you found that a particular form is not very frequent, what does this suggest?
* You will learn how to do distributional analyses (including cross-tabs) in class. ** Include: all predictor groups; the total N in each predictor group and in the analysis; the Ns for all factors.