Module 10c Excel Testing Hypotheses Chi2 & Annova (1)

Ch11 VIDEO: ASSUMPTIONS (a) Analysis of two categorical variables (nominal or ordinal) (b) Contingency tables (c) Two independents samples EXAMPLE 1 Step 1 Observed (raw data) Step 2 Female Male total Without 6 7 13 (E21*C25)/E25 College 13 16 29 (E22*C25)/E25 Bachelor 16 15 31 (E23*C25)/E25 Master 8 11 19 (E24*C25)/E25 43 49 92 EXAMPLE 2 Step 1 Observed (raw data) Step 2 smoker non-smoker female 29 71 100 (E34*C36)/E36 male 16 84 100 (E35*C36)/E36 45 155 200 STATISTICS BOOK: https://openstax.org/details/books/introductory-business-statistics https://www.youtube.com/watch?v=NDhmMH25AC4&list=PLxHjLUbDs-WGyyRD1VrSAMbpv2z3oK

FORMULA Calculate the expected Step 3 Calculate the Chi squa (row total * colum total)/overall total (Observed -Expected Female Male Female Without 6.08 6.92 (E21*D25)/E25 ((C21-J21)^2)/J21 Without 0.00 College 13.55 15.45 (E22*D25)/E25 ((C22-J22)^2)/J22 College 0.02 Bachelor 14.49 16.51 (E23*D25)/E25 ((C23-J23)^2)/J23 Bachelor 0.16 Master 8.88 10.12 (E24*D25)/E25 ((C24-J24)^2)/J24 Master 0.09 SUM(Q21:R24) x2 0.50 (Number of rows -1)*(Number of columns -1) df 3 CHISQ.DIST.RT(Q26,Q27) pvalue 0.918 Calculate the expected Step 3 Calculate the Chi squa smoker non-smoker smoker female 22.50 77.50 (E34*D36)/E36 ((C34-J34)^2)/J34 female 1.88 male 22.50 77.50 (E35*D36)/E36 ((C35-J35)^2)/J35 male 1.88 4.85 x2 4.85 (Number of rows -1)*(Number of columns -1) df 1 CHISQ.DIST.RT(Q39 pvalue 0.0277123 KfZy&index=24

are Step 4 Make your decision d)2 /Expected Male 0.00 ((D21-K21)^2)/K21 0.02 ((D22-K22)^2)/K22 0.14 ((D23-K23)^2)/K23 0.08 ((D24-K24)^2)/K24 Our expected pvalue < 0.05 P value (x2 test) p value (0.918) > 0.05; therefore, Ho is true Ho is supported There is not correlation between gender and are Step 4 Make your decision non-smoker 0.55 ((D34-K34)^2)/K34 0.55 ((D35-K35)^2)/K35 Our expected pvalue < 0.05 P value (x2 test) p value (0.0277) < 0.05; therfore, Ho is false Ho is rejected There is a correlation between gender and hi Ho: there is no relationship between gender and educati Ha: There is a correlation between gender and the educ Ho: there is no relationship between gender and smokin Ha: There is a correlation between gender and smoking