On-Line Class for April 27, 2000

Basic illustration of Type I and Type II errors

Imagine a company that has a policy of periodically drug testing all employees.  A given employee, Employee A either is ir isn't a substance abuser

Null Hypothesis (H₀): Employee A is not a substance abuser

Experimental Hypothesis (H₁): Employee A is a substance abuser

A drug test is administered to Employee A, but the test has a margin of error of .05   That is, 5% of the time the drug test produces a false positive or false negative.

Type I Error, alpha risk: false positive; Person A tests positive for drug use, but is not a substance abuser.

Type II Error, beta risk: false negative; Person A tests negative for drug use, but is in fact a substance abuser.

Either type of error is bad, but researchers would perfer to commmit a Type II error over a Type I error (better to find nothing, even if there was something to find, than to shout "Eureka!" when one hasn't really found it).

Assigned reading and learning links

Read the material in the text on Type I and Type II errors (pp. 332-333).

Use the following learning links to find out more about Type I and Type II errors.

links to Type I and Type II errors

http://www.ruf.rice.edu/~lane/hyperstat/A18652.html

Study Questions

1. A researcher thinks he or she has uncovered a significant relationship between two variables, but the results are really a statistical artifact.  The researcher has committed a:

A. Type I error

B. Type II error.

2.If a researcher rejects the null hypothesis, when if fact it is true, the researcher has committed a:

A. Type I error

B. Type II error.

3. Which confidence level below entails the greatest risk of committing a Type I error?

A. p < .05

B. p < .01

C. p < .001

4. Which confidence level entails the greatest risk of committing a Type II error?

A. p < .05

B. p < .01

C. p < .001

5. What is a Type II error?

A. rejecting the null hypothesis when it is true.

B. accepting the null hypothesis when it is true.

C. rejecting the null hypothesis when it is false.

D. accepting the null hypothesis when it is false.

6. The most common probability level used in the social sciences is:

A. p <. 05

B. p < .01

C. p < .001