Page 3 of4 Chi-Square Test Chi-Square tests (Chi rhymes with guy) are designed to compare observed frequencies within groups to their expected frequencies. The formula is: Which means that you take the observed frequency, subtract from it the expected frequency and square the result and divided by the expected frequency. You then sum all ofthe observed frequencies in the study. The bigger the difference between expected and observed, the bigger the chi-square, the more likely it is that the observed frequency did not come from the population on which the null hypothesis is based. The simplest Chi-Square is deciding whether or not a coin is "fair", which means that it comes up heads and tails an equal amount. To test this, you record the number ofheads and tails in 200 toss.es. The expected value is that there would be 100 heads and 100 tails (the null hypothesis). We did a Chi-Square test in class and found that this frequency table departs from expectation at the .05 level. HEADS
This is the end of the preview. Sign up
access the rest of the document.