Jonathon Clay Davis, Chapter 15, two way anova post, due 2-22-2006

advertisement
Jonathon Clay Davis, Chapter 15, two way anova post, due 2-22-2006
Blogs are becoming an integral part of online classroom experiences…some
instructors use them in place of discussion boards, and some use them in addition to
them. In the field of human computer interaction(HCI), the resistance level of
individuals towards using an interface is often measured to gauge how well the
application is written. This is often referred to as the “user-friendliness” of the
application. The new research opportunities that blogs have afforded HCI is that blogs
have interchangeable “skins” that allow the user to therefore change the interface of the
blog to fit something that better matches his or her preferences.
In an attempt to improve the user-friendliness success of blogs as measured by the
“user-friendliness survey”, a study was conducted considering 3 different blog interface
scenarios: method1(user is only allowed the default skin), method 2(user is allowed the
number one rated skin), or method 3(user is given default skin and told they can choose a
skin he or she prefers) from 3 different blog applications: wordpress, typepad, and
B2evolution. We randomly picked 2 users from each cell for this 3X3 factorial study.
The scores are given below.( A score of 100 means that the user was completely satisfied,
and a score of 0 means that the user was not satisfied.)
Factor A (blog application)
Wordpress
82
79
Typepad
93
91
B2evolution
61
66
method2
87
89
94
90
70
68
method3
95
93
99
100
75
78
method1
Factor B
(skin
choices)
1. Write the null hypothesis in words for the main effects and interaction.
2. Report the cell, column, row, and grand means using the appropriate notation
(i.e.,
= 6.0 )
3. Report the omnibus F-test results for the main effects and interaction effect.
Make a decision and final conclusion for each test.
4. Report and interpret partial 2 for all effects.
5. Use SPSS to generate the post hoc results for all possible contrasts for factor
A. Write the null and alternative hypothesis in symbols. Make a decision and
final conclusion about these comparisons.
6. Use SPSS to report the 95% confidence intervals for factor A using
B2evolution as the reference group. Also, construct these intervals by hand
(use the statistics from the output). Make a decision and final conclusion
about these comparisons.
1.
There is no difference in scores for users using Wordpress, TypePad, and
B2Evolution.
There is no difference in scores for users using the default skin, the most popular
skin, and any skin they want.
There is no interaction between the type of blog and the skin choice.
2.
= 80.50
= 92
= 63.5
= 78.67
s11 = 2.121
s12 = 1.414
s13 = 3.536
s1. = 12.972
n11 = 2
n12 = 2
n13 = 2
n1. = 6
= 88
= 92
= 69
= 83
s21 = 1.414
s22 = 2.828
s23 = 1.414
s2. = 11.1
n21 = 2
n22 = 2
n23 = 2
n2. = 6
= 94
= 99.5
= 76.5
= 90
s31 = 1.414
s32 = .707
s33 = 2.121
s3. = 10.807
n31 = 2
n32 = 2
n33 = 2
n3. = 6
= 87.5
= 94.5
s.1 = 6.189
s.2 = 4.135
s.3 = 6.154
s... = 11.965
n.1 = 6
n.2 = 6
n.3 = 6
N = 18
= 69.67
= 83.89
3.
skin X blog:
Blog :
Skin :
F(4,18)= 2.125, p = .160
Since the p-value is not less than the alpha level of .05,
fail to reject the null hypothesis.
There is no significant interaction between blog skin choice and
blog application user-friendly mean scores.
F(2,18) = 232.987, p = .000
Since the p-value is less than the alpha level of .05,
reject the null hypothesis.
There is a significant difference in user-friendly
mean scores among Wordpress, TypePad, and B2Evolution.
F(2,18) = 46.474, p = .000
Since the p-value is less than the alpha level of .05,
reject the null hypothesis.
There is a significant difference in user-friendly mean scores
among blog skin choice.
4.
Blog:
Skin:
BlogXSkin:
partial 2 = .981
98.1% of variance in user-friendliness score is explained by type of blog.
partial 2 = .912
91.2% of variance in user-friendliness score is explained by type of skin.
partial 2 = .486
48.6% of variance in user-friendliness score is explained by both type of
blog and type of skin.
5.
Comparison One:
wordpress to typepad
H0: w = t
H1: w ≠ t
p = .001
The p-value is less than the alpha-level, so reject the
null hypothesis. There is a significant difference in user-friendliness
mean scores between Typepad and Wordpress. Typepad mean scores
are higher.
Comparison two:
Wordpress to B2evolution
H0: w = b
H1: w ≠ b
p = .000
The p-value is less than the alpha-level, so reject the null hypothesis.
There is a significant difference in user-friendly mean scores between
Wordpress and B2Evolution. Word press mean scores are higher.
Comparison three:
Typepad to B2evolution
H0: t = b
H1: t ≠ b
p = .000
The p-value is less than the alpha-level, so reject the null hypothesis
There is a significant difference in user-friendly mean scores between
Typepress and B2Evolution. Typepad mean scores are higher.
6.
First Confidence Interval:
Wordpress to B2evolution
H0: w = b
H1: w ≠ b
t*(2,9,.05) = 2.61
Confidence Interval by hand:
17.83 +/- (2.61)(1.186)
=14.7345 to 20.925
(14.73, 20.93)
Reject the null hypothesis because zero is not in the interval.
Can be 95% confident that B2Evolution scores are different, statistically, than wordpress
scores.
Wordpress mean scores are 14.73 to 20.93 points higher.
Second Confidence Interval:
Typepad to B2evolution
H0: t = b
H1: t ≠ b
t*(2,9,.05) = 1.186
Confidence Interval by Hand:
24.83 +/- (2.61)(1.186)
= 21.73454 to 27.92546
(21.73, 27.93)
Reject the null hypothesis because zero is in the interval.
Can be 95% confident that B2evolution scores are different, statistically, than Type pad
scores.
Typepad mean scores are 21.73 to 27.93 points higher.
Download