I Could Like Maybe
Write very brief sketch of plans for 3 methodological approaches to your topic.
My Research Problem/Question |
For each method, how would it work, what would it take, plan of action; how would the analysis look? what sort of time and resources would it take to make it happen? Pros and cons
Method #1: | Method #2: | Method #3: | |
How? | How would it work? | ||
Raw Data | What would raw data look like? | ||
Cases | What are your cases? | ||
Variables | What are your variables? |
My Method of Choice |
Create and Analyze a Fake Dataset
This exercise requires a little bit of free imagination. It's late September and the leaves are turning. Senior year is off to a great start. And your data has started pouring in. It's magically cleaning itself, coding itself, and arranging itself in your computer. There it is, over there. Can you see it? It looks like you have about 30 cases. Looking across the data table it appears that you have collected data on 8 variables, 3 of which look like INDEPENDENT variables and 5 DEPENDENT. They have all been neatly coded either with numeric or mnemonic codes.
Independent Variables | Dependent Variables | |||||||
Case | Var1 | Var2 | Var3 | Var4 | Var5 | Var6 | Var7 | Var8 |
1 | ||||||||
2 | ||||||||
3 | ||||||||
4 | ||||||||
5 | ||||||||
6 | ||||||||
7 | ||||||||
8 | ||||||||
9 | ||||||||
10 |
You realize that the forces of magic have done their part and now it's up to you to pull an informational rabbit out of this data hat.
Univariate
For each of the INDEPENDENT VARIABLES, suggest how they might be distributed if your study is to really investigate their role in connection with your DEPENDENT VARIABLES.
For each of the DEPENDENT VARIABLES, suggest the range of variation you'd expect to find among the randomly sampled subjects/informants/respondents in your study.
variable 1 Univariate frequency table or histogram |
variable 2 Univariate frequency table or histogram |
variable 3 Univariate frequency table or histogram |
variable 4 Univariate frequency table or histogram |
variable 5 Univariate frequency table or histogram |
variable 6 Univariate frequency table or histogram |
variable 7 Univariate frequency table or histogram |
variable 8 Univariate frequency table or histogram |
BIVARIATE
Start with the most general question you have relating one of your DVs and IVs. Think about what you hypothesize is the relationship. Now sketch what the bivariate distribution would look like if your hypothesis was strongly supported by the data. Then what it would look like if its opposite were strongly confirmed by the data. Then what it would look like if the data were completely inconclusive as far as this hypothesis goes. Sketch both a scatterplot cloud and a 2x2 table (or m x n if 2x2 is not appropriate).
Next, imagine that the first bivariate analysis suggested no detectable relationship but you suspect that if you partial out or elaborate the table or control for a third variable (another one of your DVs). Sketch the elaborated 2x2 table that shows something different going on for different values of this third variable.
Presentation
Consider building a slide show such as this (template available here):