Anda di halaman 1dari 2

Testing Category Probabilities: One Way Table Example

A large firm has established what it hopes is an objective system of deciding on annual
pay increases for its employees. The system is based on a series of evaluation scores
determined by the supervisors of each employee. Employees with scores above 80
receive a merit pay increase, those with scores between 50 and 80 receive the standard
increase, and those below 50 receive no increase. The firm designed the plan with the
objective that, on the average, 25% of its employees would receive merit increases, 65%
would receive standard increases, and 10% would receive no increase. After 1 year of
operation using the new plan, the distribution of pay increases for the 600 company
employees was:
None
42

Standard
365

Merit
193

Test at the = 0.01 level to determine whether these data indicate that the distribution of
pay increases differs significantly from the proportions established by the firm
a) Write down the null and alternative hypothesis to test this.

b) Compute the test statistic.

c) Specify the rejection region using = 0.05, or calculate the p-value. What are your
conclusions?

d) Construct a 95% confidence interval for the proportion of the companys employees
who will receive merit increases under the new system.

Testing Category Probabilities: One Way Table Example Using SAS


To do the example using SAS, use the following program:
DATA pay;
INPUT type $ employees;
DATALINES;
None 42
Standard 365
Merit 193
;
RUN;
PROC FREQ DATA=pay ORDER=data;
WEIGHT employees;
TABLES type / TESTP=(.1,.65,.25);
RUN;
To calculate the confidence intervals from part d) using SAS, add a BINOMIAL
(level="None") option at the end of the TABLES statement. Repeat for the other
two levels.

Testing Category Probabilities: One Way Table Example Using R


chisq.test(c(42,365,193),p=c(.1,.65,.25))

Anda mungkin juga menyukai