Agriculture Reference
In-Depth Information
can be specified including the alpha level, power, sample sizes, and
whether it is a one-sided or two-sided test, to name a few.
For example, a manufacturer of rolling greenhouse benches is
thinking about changing its supplier of roller tubes. The new vendor
says he can deliver 1 5/16-inch-diameter galvanized steel tube that
is within 1/64 inch of this diameter. How large a sample would be
needed to have a 95% confidence estimate with the mean diameter
within these tolerances? Past data supplied by the vendor have the
standard deviation at 1/32 inches. To answer this, enter
sampsi
1.3125 1.328125,
sd1(
0.03125
)
alpha(
0.05
)
power(
0.5
)
onesample
The first number (1.3125) is the decimal form of 1 5/16 inches (pipe
diameter) and the second number (1.328125) is 1 5/16 inches plus
1/64 inch in decimal form. The 0.03125 is the 1/32-inch standard
deviation. The 95% confidence is entered as the alpha level (1 - 0.95
= 0.05) and the power (0.5) is the value entered to ignore the power.
Many textbooks ignore the power when presenting this subject mat-
ter, which can be confusing. The results of this command are
Estimated sample size for one-sample comparison of
mean to hypothesized value
Test Ho: m = 1.313, where m is the mean in the population
Assumptions:
alpha = 0.0500 (two-sided)
power = 0.5000
alternative m = 1.32813
sd = .03125
Estimated required sample size:
n = 16
The estimated sample size in this case is 16. If you were to enter a
power of, say, 0.80, the result would be 32. In addition, this command
can be used with two samples as well as with proportions.
This command also can be used to estimate the power of the test. It
may be appropriate at this time to discuss some basic concepts in these
Search WWH ::
Custom Search