Scribd
Upload
49 views1 page

Statistical Tests for Proportions and Means

This document outlines statistical tests for comparing two proportions, means, and variances. It defines the null and alternative hypotheses, test statistics, and assumptions for one-tailed and two-tailed left, right, and two-tailed tests comparing two proportions using a z-test, two means using a z-test, t-test, or Welch's t-test, and two variances using an F-test. Normality assumptions and required sample sizes are provided for each test.

Uploaded by

Nguyễn Trang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
49 views1 page

Statistical Tests for Proportions and Means

This document outlines statistical tests for comparing two proportions, means, and variances. It defines the null and alternative hypotheses, test statistics, and assumptions for one-tailed and two-tailed left, right, and two-tailed tests comparing two proportions using a z-test, two means using a z-test, t-test, or Welch's t-test, and two variances using an F-test. Normality assumptions and required sample sizes are provided for each test.

Uploaded by

Nguyễn Trang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Test for Equality of Two Proportions

Left-Tail Test Two-Tail Test Right-Tail Test


H0: 1  2 H0: 1 = 2 H0: 1  2
H1: 1 < 2 H1: 1  2 H1: 1 > 2
p1  p2 x1  x2 combined number of successes
Test statistic: z  where pc  
1 1  n 1  n2 combined sample size
pc (1  pc )   
 n1 n2 
Normality may be assumed if each sample has at least 10 “successes” and 10 “failures,” i.e., if
n1p1  10, n1(1p1)  10, n2p2  10, n2(1p2)  10 (some textbooks say 5 instead of 10).

Test for Equality of Two Means (Independent Samples)


Left-Tail Test Two-Tail Test Right-Tail Test
H0: 1  2 H0: 1 = 2 H0: 1  2
H1: 1 < 2 H1: 1  2 H1: 1 > 2
Case 1: Known Variances or Very Large Samples (z-test)
x1  x2 x1  x2
zcalc  or zcalc  Use z-table for critical values.
1  2
2 2 2
s1 s2 2
 
n1 n2 n1 n2

Case 2: Unknown Variances Assumed Equal (pooled samples t-test)


x1  x2 (n  1) s12  (n2  1) s2 2
tcalc  with pooled variance s p 2  1
1 1  n1  n2  2
s p2   
 n1 n2 
Use d.f. = (n11)+(n21). If both samples are large (n1  30 and n2  30) see Situation 1.
Case 3: Unknown Variances Assumed Unequal (Welch’s t-test)
2
x1  x2  s12 n1  s2 2 n2 
tcalc  Use d.f.   
s   s 
2 2
s12s 2 2
n1 2
n2
 2 1 2
n1 n2 n1  1 n2  1
Welch’s d.f. formula is too hard for hand calculations, but MINITAB and Excel use it. The test
statistics in Case 2 and Case 3 will be the same if n1 = n2. (but decisions may differ due to d.f.).

Test for Equality of Two Variances


Left-Tail Test Two-Tail Test Right-Tail Test
H0: 1  2
2 2
H0: 12 = 22 H0: 12  22
H1: 12 < 22 H1: 12  22 H1: 12 > 22
or, equivalently,
H0: 12/22  1 H0: 12/22 = 1 H0: 12/22  1
H1: 1 /2 < 1
2 2
H1: 12/22   H1: 12/22 > 1
Test statistic is Fcalc  s12 s22 with df1 = n11 (numerator) and df2 = n21 (denominator). The F
test is sensitive to population skewness. If you put the larger variance in the numerator (to force a
right-tail test) then use  when you look up the critical value in the F table.

Revised 11/15/11

You might also like