ATHK1001 Lecture Notes - Lecture 15: Chi-Squared Distribution, Ultimatum Game, Null Hypothesis
Document Summary
Analysis of categorical data: not all interesting data is qualitative. Lots of data is quantitative: e. g. sales figures, how many people intend to vote labour or coalition at the next election, how many people reject unfair offers in the ultimatum game. Qualitative data has categorical variable(s: often summary data is frequencies, a general approach to tests. In many statistical tests we are concerned about deviation: e. g. deviation from the mean, deviations from expected frequencies. Model: is what we would expect the data to look like, usually when the null hypothesis is true. If data deviates by lot from the model, then reject the model (i. e. reject the null hypothesis): chi square statistic. If the sum is big enough, the chi square tells us there"s statistical. Sum up the differences between observed frequency and expected significance. frequency (which depends on the hypothesis) for each category, each difference being squared and divided by the expected frequency.