Chapter 10--Hypothesis Testing--Categorical Data

# 6812 pnorml806812 observedc57330 21324584

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: C Contingency Tables Chi-Square Goodness-of-Fit Test The Kappa Statistic Chi-Square Goodness-of-Fit Test: Example Example: Assess the goodness of ﬁt of the normal distribution to the DBP data. The sample mean and standard deviation computed from the row data are: x = 80.68 and s = 12. ¯ To compute the probabilities of the classes, we use pnorm. for the the third class P (60 ≤ X < 70) = pnorm(70, 80.68, 12)−pnorm(70, 80.68, 12). for the ﬁrst class, P (X < 50) = pnorm(50, 80.68, 12). Since we are estimating 2 parameters and we have eight classes in the frequency table df = 8 − 1 − 2. Chapter 10: Hypothesis Testing: Categorical Data Stat 491: Biostatistics Introduction Two-Sample Test for Binomial Proportions McNemar’s Test Estimation of Sample Size and Power R × C Contingency Tables Chi-Square Goodness-of-Fit Test The Kappa Statistic Chi-Square Goodness-of-Fit Test: R-Codes L<- c(-Inf,seq(50,110,10)) U<- c(seq(50,110,10),Inf) prob<-pnorm(U,80.68,12)-pnorm(L,80.68,12) observed=c(57,330, 2132,4584, 4604,2119,659,251) n<-sum(observed) expected <- n*prob X<- sum((observed- expected)^2/expected) df <- length(observed) - 1 - 2 p_value <- 1 - pchisq(X,df) Chapter 10: Hypothesis Testing: Categorical Data Stat 491: Biostatistics Introduction Two-Sample Test for Binomial Proportions McNemar’s Test Estimation of Sample Size and Power R × C Contingency Tables Chi-Square Goodness-of-Fit Test The Kappa Statistic Measuring the Strength of Association We are interested in the degree of association between two categorical variables. When the same variable is measured more than onc...
View Full Document

## This note was uploaded on 02/03/2014 for the course STAT 491 taught by Professor Solomonharrar during the Fall '12 term at Montana.

Ask a homework question - tutors are online