Module - Chi Square
On examining the 2008 internal party election results in the great state of ABC, a Chi Square test for independence was conducted.
The two variables of choice were:
1. Candidate A (male) and Candidate B (male) , and
2. Voters' Gender: Females (n=2.6 million) / Males (n=2.5 million)
The Chi Square was very high and statistically significant, which means that for some reason there was a strong gender preference for one of the candidates.
This is the important part:
3. If you were to be a hopeful 2012 candidate for your party, what could you LEARN and SAY about the 2008 voting results in State ABC?
What could you not learn from it?
Learning Outcomes
Upon successful completion of this module, the student will be able to satisfy the following outcomes:
- Case
o Discuss non-parametric distribution and tests, and distinguish between observed and expected datum.
o Choose a Chi Square test when appropriate per the study hypotheses and variables' level of measurement.
o Review, conduct, and interpret statistical Chi Square tests.
- SLP
o Implement within a "real-world" project: The definition of research questions and their derivative hypotheses; Examine data; perform appropriate statistical test: t test, ANOVA/ANCOVA, Chi Square; interpret and report results.
- Discussion
o Discuss what a Chi Square test for independence can explain.
Module Overview
1. In the previous modules, we looked at Inferential statistical procedures which utilize research data sampled from a study population to make inferences about that particular study population.
2. The basic assumption underlying these statistical methods is that the data conforms to a normal distribution.
3. Whenever, based upon our examination of the descriptives of our data (as covered in module 1), we have non-parametric variables, we will need to utilize non-parametric statistical tests.
4. In this module we will focus on: The Chi Square test.
5. The Chi Square looks at two nominal variables with their respective categories (sub-groups), and identifies the differences between the observed datum and an expected datum. For example: The academic success level of students (Pass/Fail) and their Gender (Females/Males).
6. In a "perfect world" we would expect that the proportion of success (or failure) would be similar between the males and females.
7. However, when data was collected, we found out that the observed data is different than our expected ones.
8. The Chi Square statistic and its level of significance (p value) allow us to decide whether the two variables (success and gender) are independent of each other, and consequently decide on our null hypothesis.
9. It is recommended that you FIRST view the modular Case and SLP assignments, and see what is required of you to do. With this in mind, review the Background page for reading and support materials.
Attachment:- Assignment Background.rar