Chi- square

Chi- square

1. Assists in whether or not a systematic association exsists between two variables

2. Expected cell frequencies and actual observed frequencies that are found in the cross tab can then be used to calculate chi -square

3. The greater the discrepancy between the expected and actual, the LARGER the value of the statistic

4. To determine whether a systematic association is present is calculated from the cross-tabulation is estimated

5. Degrees of freedom = number of observations less the number of constraints

6. Chi- square distribution is often skewed as the shape depends on the number of degrees of freedom (DF)

7. DF increases , distribution becomes more symmetrical

8. Can be used in Goodness-of-fit tests to determine whether a model fits with observed data

9. Only be estimated on the counts of data

10. Used when there are TWO nominal data variables

11. Can show whether an observed relations between two variables arose as a result of chance

12. Can only be used on actual numbers not; percentages, means, proportions etc.

13. There are several types of Chi-square that can vary in use depending on the hypothesis

14. DF = (number of columns - 1) X (Number of rows - 1)

15. Working out chi square =