In: Statistics and Probability
An education researcher is interested in students’
performance on a mathematics test for a novel---and generally
unfamiliar---concept. The researcher believes that if students
participate in an extracurricular training session, then they will
be more likely to perform better on the mathematics test.
This researcher is testing two hypotheses. The first is that the
training session provides specific instruction and has an effect on
students’ mathematics test scores. The second is that this
treatment effect is moderated by gender. Past research has
suggested that there should not be a main effect for gender.
However, to control for this possibility, a 2-way (2x2) factorial
ANOVA is being conducted to account for any influence from gender,
the treatment, and the interaction of the two main effects.
To test the hypotheses, the research has obtained 28 participants:
14 boys and 14 girls. From each group of 14, 7 students were
randomly chosen for the treatment group. After the treatment, all
students were given the same test. This test has been used in
previous research, has a range of 0 to 100, and measures students’
ability to correctly perform the novel mathematics exercise. (The
higher the test score, the more positive the result.)
Use a 2x2 ANOVA with α=0.02α=0.02 to test the the data and evaluate
the hypotheses. Results are provided below for each group of
students.
Factor B: Treatment | |||
Control: No Training |
Treatment: Attend Training |
||
Factor A: Gender |
Male | 67.6 60.0 65.4 67.3 57.6 43.1 61.3 |
61.6 69.3 70.5 56.6 64.2 60.5 62.6 |
Female | 52.8 51.5 66.5 62.7 55.8 48.1 58.6 |
71.0 73.6 76.4 62.4 74.4 86.9 71.2 |
( 1a ) What is the F value for the
treatment effect?
(Report answer accurate to 2 decimal place.)
( 1b ) What is the p-value for the
F value for the treatment effect?
(Report answer accurate to 4 decimal places.)
( 1c ) Does this support the researcher's
hypothesis that the treatment has an effect on ability to solve the
mathematics exercise?
( 2a ) What is the F value for the gender
effect?
(Report answer accurate to 2 decimal place.)
( 2b ) What is the p-value for the
F value for the gender effect?
(Report answer accurate to 4 decimal places.)
( 2c ) Does this support the researcher's
assumption that gender does NOT have an effect on
ability to solve the mathematics exercise?
( 3a ) What is the F value for the
interaction effect?
(Report answer accurate to 2 decimal place.)
( 3b ) What is the p-value for the
F value for the interaction effect?
(Report answer accurate to 4 decimal places.)
( 3c ) Does this support the researcher's
hypothesis that the treatment effect is moderated by gender?
Anova: Two-Factor With Replication | ||||||
SUMMARY | NO TRANING | ATTEND TRANING | Total | |||
MALE | ||||||
Count | 7 | 7 | 14 | |||
Sum | 422.3 | 445.3 | 867.6 | |||
Average | 60.32857 | 63.61429 | 61.97143 | |||
Variance | 72.08571 | 24.01143 | 47.25912 | |||
FEMALE | ||||||
Count | 7 | 7 | 14 | |||
Sum | 396 | 515.9 | 911.9 | |||
Average | 56.57143 | 73.7 | 65.13571 | |||
Variance | 42.09238 | 53.87667 | 123.2825 | |||
Total | ||||||
Count | 14 | 14 | ||||
Sum | 818.3 | 961.2 | ||||
Average | 58.45 | 68.65714 | ||||
Variance | 56.49808 | 63.33495 | ||||
ANOVA | ||||||
Source of Variation | SS | df | MS | F | P-value | F crit |
Sample | 70.08893 | 1 | 70.08893 | 1.459683 | 0.238751 | 6.210859 |
Columns | 729.3004 | 1 | 729.3004 | 15.18852 | 0.000683 | 6.210859 |
Interaction | 335.3432 | 1 | 335.3432 | 6.983909 | 0.014255 | 6.210859 |
Within | 1152.397 | 24 | 48.01655 | |||
Total | 2287.13 | 27 |
F value for treatment = 15.188
p value = 0.000683
p value < 0.02 , significant
researcher's hypothesis that the treatment has an effect on ability to solve the mathematics exercise?
yes
.................
F stat for gender effect = 1.4596
p value = 0.2387
p value > 0.02 , not significant
( 2c ) Does this support the researcher's assumption that gender does NOT have an effect on ability to solve the mathematics exercise?
YES
...............................
F stat for interaction = 6.983909
p value = 0.014255
p value < 0.02 , significant
Does this support the researcher's hypothesis that the treatment effect is moderated by gender?
yes
.............................................
Please revert back in case of any doubt.
Please upvote. Thanks in advance.