################################################################# ### Testing hypotheses about one proportion using prop.test() ### ################################################################# prop.test(10, 100, p = 0.5, correct = FALSE) ################################################################## ### Testing hypotheses about two proportions using prop.test() ### ################################################################## # Read data table into R pill.study <- matrix(c(23, 34, 35, 132), ncol = 2, byrow = TRUE) colnames(pill.study) <- c("Blood Clot", "No Clot") rownames(pill.study) <- c("Pill", "No pill") # Show data table pill.study # Test whether probabilities are equal for the two groups prop.test(pill.study, correct = FALSE) ################################################################### ### Testing hypotheses about two proportions using chisq.test() ### ################################################################### # Test whether probabilities are equal for the two groups chisq.test(pill.study, correct = FALSE) # Expected values chisq.test(pill.study, correct = FALSE)$expected ################################################################### ### Testing hypotheses in contingency tables using chisq.test() ### ################################################################### # Read data table into R poll <-matrix(c(79, 91, 93, 84, 66, 60, 37, 43, 47), ncol = 3, byrow = TRUE) colnames(poll) <- c("4 weeks", "2 weeks", "1 week") rownames(poll) <- c("Cand1", "Cand2", "Undecided") # Show data table poll # Show column percentages prop.table(poll, 2) # Plot probabilities barplot(t(prop.table(poll, 2)), beside = TRUE, col = 2:4, las = 1, ylim = c(0, 0.5), ylab = "Percent", xlab = "Candidate", main = "Distribution of votes") legend(legend = colnames(poll), fill = 2:4, "topright") # Testing for same distribution in the three populations chisq.test(poll, correct = FALSE) # Expected values chisq.test(poll, correct = FALSE)$expected