Chi Square Tests


  • The Chi-Square Distribution
  • Chi-square goodness of fit tests $拟合优度检验$
  • A chi-square test for independence $独立性检验$

Chi-Square Tests

  • Test of Multinomial Experiment
    • General Test
    • Test for homogeneity
    • Test for normal distribution
  • Test for independence

The Chi-Square Distribution

The chi-square $\chi^2$ distribution depends on the number of degrees of freedom.

  • There is a family of chi-square distributions.
  • It is skewed to the right.
  • It is non-negative.
\[\chi^2 = \sum_{i=1}^{k} \frac{(f_i - E_i)^2}{E_i}\]

Test of Multinomial Experiment

  1. H0: multinomial probabilities are p1, p2, … , pk Ha: at least one of the probabilities differs from p1, p2,…, pk
  2. Choose $\alpha$
  3. Multinomial Experiment Assumption: The excepted cell frequencies are >5
  4. Critical value is $\chi^2_\alpha$, with k-1 degrees of freedom
  5. Test statistic:
\[\chi^2 = \sum_{i=1}^{k} \frac{(f_i - E_i)^2}{E_i}\]
  1. Reject H0 if $\chi^2$ > $\chi^2_\alpha$

General Test Example

Marital status

Test for homogeneity Example

Support calls

Chi-Square Test for Independence

Chi-square test of independence is used to determine if there is a significant relationship between two qualitative (categorical) variables.

A contingency table is used to investigate whether two traits or characteristics are related.

  1. H0: X and Y are independent Ha: X and Y are dependent

  2. Assumption: The excepted cell frequencies are > 5
  3. Set the level of significance
\[\hat{E}_{ij} = \frac{r_i c_j}{n}\]
  1. Critical value is $\chi^2_\alpha$, with degrees of freedom (r-1)(c-1)
  2. Test statistic:
\[\chi^2 = \sum_{i=1}^{r} \sum_{j=1}^{c} \frac{(f_{ij} - E_{ij})^2}{E_{ij}}\]

where: $f_{ij}$ = observed frequency in cell $(i, j)$ $E_{ij}$ = expected frequency in cell $(i, j)$ $r$ = number of rows $c$ = number of columns

  1. Reject H0 if $\chi^2 > \chi^2_\alpha$

Chi-Square Test for Independence Example


Chi-squared test of goodness of fit:

\[\chi^2 = \sum_{i=1}^{k} \frac{(f_i - E_i)^2}{E_i}\]

d.f. = k - 1

Chi-squared test of independence in a contingency table:

\[\chi^2 = \sum_{i=1}^{r} \sum_{j=1}^{c} \frac{(f_{ij} - E_{ij})^2}{E_{ij}}\]

d.f. = (r - 1)(c - 1)

These formulas and degrees of freedom calculations are fundamental to performing chi-squared tests in statistics.