Some material from ppt slides for Statistics for Managers Using Microsoft Excel, 5e © 2008 Pearson Prentice-Hall, Inc.

The Standard Normal Distribution • ‘Bell Shaped’ & Symmetric • The mean, μ, is always 0 • The standard deviation, σ , is always 1 To find P(z 30 1.

Define the NULL & ALT hypotheses

ONE-SAMPLE HYPOTHESIS TESTING For the population mean, n > 30 1.

Define the NULL & ALT hypotheses

2.

Calculate the test statistics

One- or two-tailed? 2.

Calculate the test statistics zOBS = STANDARDIZE(x, , / n)

One- or two-tailed?

zOBS, zCRIT, p-value

zOBS = STANDARDIZE(x, , / n)

zCRIT = NORMSINV(1- /# of tails)

zOBS, zCRIT, p-value

zCRIT = NORMSINV(1- /# of tails)

p-value =(# of tails) * (1-NORMSDIST(|zOBS|))

p-value =(# of tails) * (1-NORMSDIST(|zOBS|))

Make a decision

3.

|zOBS| > zCRIT? p-value < α-level?

…then REJECT the NULL

4

6/19/2009

ONE-SAMPLE HYPOTHESIS TESTING For the population mean, n < 30 Define the NULL & ALT hypotheses

1.

ONE-SAMPLE HYPOTHESIS TESTING For the population proportion Define the NULL & ALT hypotheses

1. One- or two-tailed?

One- or two-tailed?

NULL: = 0.a ALT: ≠ 0.a ----------------------------------------------------------------NULL: ≥ 0.a NULL: ≤ 0.a ALT: < 0.a ALT: > 0.a

Calculate the test statistics

2.

tOBS, tCRIT, p-value

Use t-test! (watch the video lecture & video example)

Make a decision

3.

|tOBS| > tCRIT? p-value < α-level?

…then REJECT the NULL

ONE-SAMPLE HYPOTHESIS TESTING

ONE-SAMPLE HYPOTHESIS TESTING For the population proportion

1.

Define the NULL & ALT hypotheses

For the population proportion 1.

Define the NULL & ALT hypotheses

2.

Calculate the test statistics

One- or two-tailed? 2.

Calculate the test statistics (1

zOBS = STANDARDIZE(p, ,

One- or two-tailed?

zOBS, zCRIT, p-value

)

(1

n )

zOBS = STANDARDIZE(p, ,

zCRIT = NORMSINV(1- /# of tails)

zOBS, zCRIT, p-value

) n )

zCRIT = NORMSINV(1- /# of tails)

p-value =(# of tails) * (1-NORMSDIST(|zOBS|))

p-value =(# of tails) * (1-NORMSDIST(|zOBS|))

Make a decision

3.

|zOBS| > zCRIT? p-value < α-level?

…then REJECT the NULL

5