Statistics 3 Contingency Tables AS AQA
- Created by: Brandon
- Created on: 29-05-13 12:53
View mindmap
- Contingency Tables
- A Contingency Table shows the frequencies of two or more variables simultaneously.
- X2 Distribution
- X2 = SIGMA (O-E)2 */* E
- Compare with critical value. Degrees' of Freedom
- An (m x n) Rows-1 x Column-1 i.e. Rows=2 Column is 3, 2-1x3-1 = 2
- At 5% X2 Crit Value = 5.991
- An (m x n) Rows-1 x Column-1 i.e. Rows=2 Column is 3, 2-1x3-1 = 2
- Compare with critical value. Degrees' of Freedom
- Only work if Os are frequencies + Es are reasonably large i.e. over 5
- X2 = SIGMA (O-E)2 */* E
- X2 Distribution
- O - Original Value
- E - Expected Value
- Calculated after the O.
- E = Row Total x Column total */* Grand Total
- Calculated after the O.
- Ho - Variable is Independent
- H1 - Variable not Independent
- Reject Ho if X2 larger than Critical Value
- H1 - Variable not Independent
- Small Expected Values
- Can lead to a bad Critical Value so combining classes is needed to increase the value of E
- E should be Large
- Classes need to be combined so data still remains a contingency table
- In order to interpret the conclusions, classes with small E's should be combined with similar classes.
- E should be Large
- Can lead to a bad Critical Value so combining classes is needed to increase the value of E
- Yate's Correlation
- 2X2 tables
- So find out X2 but like this -->
- Sigma (O-E/-0.5)2 / E
- I.E. when O+E=x when x is 4 then 4-0.5 is 3.5
- 2X2 tables
- ONE TAILED
- A Contingency Table shows the frequencies of two or more variables simultaneously.
Comments
No comments have yet been made