Dr.
David Pillow
Practice Problem Set #3
1. An educator tests a sample of 12 nursery-school children to determine
how well they grasp the concept of conservation of volume. This is tested
by determining whether they realize that an amount of clay remains the
same regardless of whether it is rolled in shape of a long sausage or into a
compact ball. The educator then institutes a program designed to give the
children experiences that should improve their understanding of
conservation of volume. After the training session, the same test that was
initially given to the children is repeated. Is the program effective, or is any
difference observed from pretest to posttest due to chance sampling-
fluctuations?
(a) Complete the following table so you can get all the info needed to find
the value of the t-statistic. (Find the sums and sum of squared values
needed for the rest of the problem.)[3]
Subject# PRETEST POSTTEST
S1 5 6
S2 6 7
S3 4 3
S4 4 6
S5 2 2
S6 1 3
S7 2 4
S8 5 4
S9 5 7
S10 5 4
S11 7 9
S12 3 6
(b) State the null hypothesis and the alternative hypothesis.[3]
(c) What are the degrees of freedom for this t-test? [3]
(d) Draw a figure showing the t-distribution under the null hypothesis.
Color in the area for a two-tailed test with α = .05. Label the x-axis using
the critical values of t, and designate the regions of acceptance and
rejection.[3]
(e) Find the value for the t-statistic (show all work)[6]
(f) What number from part (e) represents the estimated standard error? [1]
(g) Do you retain or reject the null hypothesis? Also, state the conclusion in
your own words.[3]
2. If the population average income for owners of Mamma Mia Pizza Shops
is $40,000, does a special course on pizza store management given to a
sample of 121 owners have any effect in increasing revenues. The average
for those given the special course is $40,800, with a standard deviation of
$5,500 across the sample of 121 owners. Test the hypothesis that the
course has a significant effect using an alpa of .05, with a directional one-tail
test.
(a) State the null hypothesis and alternative hypothesis.[4]
(b) How many degrees of freedom are there in this study?[3]
(c) Make a graph of the regions of rejection and acceptance, and show the
critical value of t which must be surpassed in order to reject the null. Label
the x-axis using the critical values of the t statistic.[4]
(d) Calculate the t statistic.[6]
(e) Compare the t statistic with the critical value of t obtained in part d and
reject or retain the null hypothesis. State the conclusion in your own words.
[3]
3. An experimenter wishes to test the hypothesis that a special vitamin
supplement increases memory. To do this the experimenter randomly
assigns subjects to either receive or not receive the vitamin supplement.
After allowing a time period for the vitamin to work, the subjects are then
given a memory test. The number of nonsense syllables correctly recalled
serves as the dependent variable. The data for the two groups (with 10
subjects in each group) are as follows:
Control Experimental
10 100 14 196
8 64 10 100
12 144 13 169
15 225 19 361
17 289 16 256
11 121 18 324
7 49 8 64
8 64 11 121
11 121 10 100
11 121 11 121
ΣX1 = 110 ΣX2 = 1298 ΣX2 = 130 ΣX2 = 1812
(a) State the null and alternative hypotheses:[3]
(b) How many degrees of freedom are there for testing the hypothesis?[3]
(b) Diagram the regions of acceptance and rejection for the test. Label the
horizontal axis using the critical values of the test statistic, using an alpha
equal to .01, two-tailed.[4]
(c) What is the value of the estimated standard error? (show all work). Hint
#1: Find the sum of squared deviations first, and then find the pooled
standard error. [9]
(d) What is the value of the standard error from part (c). [2]
(e) Based on the value of the test statistic you calculated in part (c), do you
retain or reject the null hypothesis? State the conclusion in your own words.
[4]
For multiple choice part of test:
1. When do you use independent groups vs. matched pairs vs. one-sample?
What are advantages and disadvantages of each?
2. How does your choice of t-test affect sample size requirements?
3. What are the assumptions for each test?
4. How does changing N, the standard deviation, and the difference
between the means affect the size of t (which is related to power)?
5. When do you retain the null versus reject the null?
6. What is the expected value of t under the null hypothesis?
(i.e., what do you expect t to equal if the null hypothesis is true?)
7. How does the t distribution differ from the normal distribution?
8. An important assumption for the t-test that could affect the validity of
the test is that
a. the population distribution of score is normal when the sample size
is small.
b. the sample size is less than 30
c. the value for σ is known
d. the population distribution of scores is normal when calculating t-
values using
very large samples.
9. One clue to identifying a situation calling for an independent-groups t is
a. the value for σ should be known.
b. the mean for a treated group of subjects is compared to a known
population mean.
c. one sample is used to test a hypothesis about one population.
d. there are two samples containing different subjects.
10. Which of the following studies uses a matched-subjects or repeated
measures design?
a. a group of twins is tested for visual acuity on one day and then
tested again the following day
after receiving a drug.
b. a sample of children is tested for vocabulary skills at age six and
then again at age ten.
c. subjects are assigned to one of two treatment groups so that each
subject in one group has
the same IQ as a subject in the other group.
d. none of the above.
e. all of the above.
