EPY 6214 Educational and Psychological Statistics Lab

Lab #4 One-Way ANOVA•
•
You have to “write down” your answers and attach the
SPSS output/tables. You will not receive any credit if
only attach the SPSS output/tables without answering
the questions.
Your work needs to be in Word or PDF You receive
your credits by turning your work into Word and PDF
files. You need to attach the SPSS output by copy-paste
the tables in your work.
A researcher examined the effects (if any) of eliminating
either vision, smell, or both, from inclusion in identifying
taste. In order to determine this, he asked participants to
taste Skittles under one of three conditions. Group one
(taste alone) closed their eyes and pinched their noses,
group two (taste and olfaction) closed their eyes, and
group three (taste plus olfaction plus vision) had no
restrictions. Each participant had to taste 10 Skittles and
the researcher recorded the number of correct guesses for
each participant.
Here are the data:
Group 3
Group 1
Group 2
Taste Alone
Taste + Olfaction
Taste + Olfaction
+ Version
Correct guess
Correct guess
Correct guess
2
3
10
5
5
10
4
6
10
8
3
10
7
4
10
2
2
10
6
3
10
10
3
10
7
9
10
8
8
10
7
7
10
2
10
10
5
6
10
4
6
10
6
2
10
5
4
10
3
6
10
4
5
10
4
3
10
3
5
10
6
5
10
1
7
10
5
3
Using alpha = .05.
•
•
•
Is there a statistically significant difference in the mean
number of correct guesses among the three groups?
What can you conclude?
Calculating the effect size for the mean difference
classifies it as being indicative of a “small,” “medium,” or
“large” ES.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~
HINTS:
When reporting the results from Lab #4, here are the steps (which already presented in
the PPT, but I am going to announce it one more time):
1) report Levene’s test result. We would like to have the equal variance assumption
made (when p-value > .05), but it is not always the case. When the p-value of Levene’s
test is less than .05, we violate the equal variance assumption. However, one-way
ANOVA is a robust test. Therefore, we can still go ahead and interpret the result.
2) report the ANOVA test result to see if there is a group mean difference. If the ANOVA
result is not statistically significant (p > .05), just report that there is no group difference.
3) if the one-way ANOVA result is statistically significant ( p < .05), you have to report the effect size (eta-squared, indicate if it is small, .01, moderate, .06, or large, .14, effect size). 4) if you have three or more groups, and your one-way ANOVA result is statistically significant (p