CHI SQUARED

• WHAT IS CHI-SQUARED TESTING FOR?

  differences from expectations
• TYPE OF DATA FOR CHI-SQUARED?

  2 or more categories
• NULL HYPOTHESIS?

  there is no difference between observed and expected values for all the categories
• DEGREES OF FREEDOM?

  number of categories - 1
• CONCLUSION FOR CHI-SQUARED IF TEST STATISTIC SMALLER THAN CRITICAL VALUE?

  1. t < CV
  2. Accept null hypothesis
  3. No significant difference between observed and expected
  4. With 5% chance of error
• CONCLUSION FOR CHI-SQUARED IF TEST STATISTIC GREATER THAN CRITICAL VALUE?

  1. t > CV
  2. Reject null hypothesis
  3. Significant difference between observed and expected
  4. With 5% chance of error