Statistical tests, Continued) – Sharp EL-9900 User Manual
Page 115
18
Advanced Keyboard/STATISTICS USING THE SHARP EL-9900
Copyright © 2002, Sharp Electronics Corporation. Permission is granted to photocopy for educational use only.
Steps for performing a two-small-sample hypothesis test for
the population means
1.
Access the data entry screen and enter the additional data for L2.
(Use the Bears data in L1.):
L2 Packers
8 5 5 8
8 8
8
4 5 4 10 6
4 9 9 9 11
13
13
11
8 9 12
2.
Check the data you have entered and correct any errors you may find.
The data shown above reflects the number of wins for the Green Bay
Packers in each complete regular season since 1978. Test the alternate
hypothesis that "da Bears" are a better (on the average win more games)
football team than the Packers during these recent years.
3.
Set up the statistical test by pressing 2ndF QUIT STAT E (TEST)
0 4 (Ttest2samp).
4.
Set the alternate hypothesis to µ1 > µ2 by pressing ENTER .
Pool the standard deviations in the calculation by pressing
▼
ENTER .
Set the List1 to L1 by pressing
▼ 2ndF L1 ENTER . Set List2 to L2
by pressing 2ndF L2 ENTER . Press 2ndF EXE to compute the
statistical test.
5.
The first item on the screen, µ1>µ2, is the alternate hypothesis of the statistical
test. The second item is the observed statistic from the sample . The third
item is decision statistic or p value. In our problem, the p value is .3119 which
is more than .05. Our decision is to support the null hypothesis that µ1=µ2.
This test shows that "da Bears" average about the same number of wins a
season as the Packers. However, the fifth and sixth values show, on the
average, "da Bears" win approximately 8.56 games a season, whereas the
Packers win 8.13. The Bears win more on the average, but statistically, they
are not significantly different.
STATISTICAL TESTS
(continued)
▼
▼
▼