asypow (version 2012.04-1)

asypow.sig: Asymptotic Significance

Description

Calculates the significance level of a test via likelihood ratio methods.

Usage

asypow.sig(asypow.obj, sample.size, power)

Arguments

asypow.obj

The object returned from asypow.noncent.

sample.size

The sample size of the test.

power

The power of the test.

Value

• Returns the significance level of the test.

concept

signficance

References

Cox, D.R. and Hinkley, D.V. (1974). Theoretical Statistics Chapman and Hall, London.

See Also

asypow.noncent, asypow.n, asypow.power

Examples

# Single Group Binomial Example:
#    Consider testing the null hypothesis that the binomial
#    probability is p = .2 when the actual probability is .4.
#    What significance level corresponding to a sample
#    size of 47 and power of .8?
# Step 1: Find the information matrix
info.binom <- info.binomial.kgroup(.4)
# Step 2: Create the constraints matrix
constraints <- c(1, 1, .2)
# Step 3: Find the noncentrality parameter and
#         degrees of freedom for the test
binom.object <- asypow.noncent(.4, info.binom, constraints)
# Step 4: Compute the power of a test with
#         sample size of 47 and a significance level 0.05
sig.binom <- asypow.sig(binom.object, 47, 0.8)
print(sig.binom)