Ann Lynn, Ph.D.
Psychology 207
Introduction to Hypothesis Testing: The Z test
I. Hypothesis testing: the procedure for doing inferential statistics
A. The z-test compares a sample mean (with treatment) to a
known population mean (without treatment), when we also know the
population standard deviation
NOTE: this research design rarely happens in real life!!
II. Language and steps of Hypothesis testing
A. State statistical hypothesis
Ho : Null hypothesis (statement that there is no effect)
H1 : Alternate hypothesis/Research hypothesis (statement that there is an effect)
B. Find Critical area
1. specify the cutoff sample mean needed in order to conclude
that the sample mean was not likely to be obtained from the population
2. alpha level = α = the size of the critical area
a. Convention is that this is 5% or less of the distribution of sampling means
3. How to do it:
a. Find the Z-critical. This is the cutoff Z score that defines the critical area
Zc for Common α levels (taken from Normal Table)
α = .05 α = .01
One tailed 1.64 2.33
Type of H1
Two tailed ± 1.96 ± 2.58
C. Collect data and Calculate test statistic
1. Convert the sample data to a standardized score in order to find
out if it is beyond the cutoff specified in Step B.
2. How to do it: Convert sample mean to Z- observed
Zo = M-μ where: σM = σ
------- ---
σM √n
Ann Lynn, Ph.D.
Psychology 207
D. Make decision and state conclusion
1. Decision will either be:
a. The null hypothesis is a good explanation (Sample data not beyond
cutoff)
* "We do not reject the null hypothesis" (statistically correct terminology)
* NOT statistically significant
* The observed differences between the M and μ are probably due to
error/chance
* The null hypothesis is a reasonable explanation of the results
* The IV had no effect on the DV
* p > .05
b. The researcher's hypothesis is a good explanation (Sample data IS
beyond the cutoff)
* "We reject the null hypothesis" (statistically correct terminology)
* Statistically significant
* The observed differences between the M and μ are probably due to
the IV
* The null hypothesis is wrong
* The IV had an effect on the DV
* The probability of no real difference between the M and μ is less than
5%
* p < .05
3. State conclusion
a. Was there a sig. effect?
b. Include the names of the IV and DV
c. Include the sample mean and SD
d. Include the string of numbers:
If reject Null: Z = obs. Z score, p < alpha level
If do not reject null: Z = obs. Z score, p > .05
example:
“A Z-test revealed that the new stress reduction technique significantly
reduced the stress level of the students in the sample (M = 10.5, SD = .28)
from the stress level reported by the population (μ=12, SD=.3),
Z = 2.0, p< .05.”
Ann Lynn, Ph.D.
Psychology 207
Z-test Example Problems
Demonstration example
A local school district recently implemented an experimental program for science education.
After 1 year, the 36 children in this program obtained an average score of 61.8 on a national
science achievement test. This national test is standardized so that the national average is 60
with a standard deviation of 6. Did the students in the special program score above the
national average? Test at α = .05.
Practice Example
A recent national survey reported that the general population gave the president an average
rating of μ = 60 with a standard deviation of 3.5 on a scale of 1 to 100 with 1 being most
unfavorable, and 100 most favorable. A researcher at IC hypothesizes that IC students are
more critical of the president than the general population. She collects data on 10 IC students
(below). At α = .05 test the one-tailed hypothesis that IC students are more critical than the
population.
Sample Ratings of President
44 50 24 45 39 57 25 90 78 54
