Question 1:
I wanted to see if missing data was completely at random, I ran Little MCAR’s test and the result is
as below:
This is not significant therefore (>0.05), missing values are completely at random. Even though
there were about 1% of the observations missing data, I went ahead to impute the missing data.
After computing mahalanobis probabilities for all of the observations and found 02 outliers at 860
and 204 (participant numbers) that have mahalanobis probabilities of 0.00  I went ahead and
deleted them out.
Question 2:
Reliability Statistics
Cronbach's Alpha
.825
N of Items
13
Reliability test is shown with Cronbach’s Alpha = .825  exceeds threshold of .7
Questions 3 and 4:
I did Ward’s method and Squared Euclidean distance, there seem to be three clusters together.
Question 5:
I ran the analysis and the outputs are as below and QCL_1 is computed.
Question 6, 7, and 8:
There is no significant difference between clusters for perceived opinion whereas for importance,
there is significant difference between the clusters.
Question 9:
Question 10, 11, and 12:
Nodes
Importance Importance
importance_video_production
0.1212
0.1212
importance_creator
0.1308
0.1308
importance_duration
0.2984
0.2984
importance_recently
0.4819
0.4819
importance_recommended
0.5277
0.5277
importance_thumbnail
0.5466
0.5466
importance_ad
0.6666
0.6666
importance_colors_thumbnail
0.7508
0.7508
V4
V5
importance_matrix - The
0.1212
quality of the video production
importance_matrix - Who the
0.1308
content creator is
importance_matrix - How long
0.2984
the video is
importance_matrix - How
0.4819
recently the video was released
importance_matrix - Whether
someone recommended the 0.5277
video
importance_matrix - How
attractive the video Thumbnail 0.5466
is
importance_matrix - Whether
0.6666
you saw an ad about the video
importance_matrix - The
colors used in the video
0.7508
Thumbnail
importance_likes
0.8683
0.8683
importance_views
1
1
importance_matrix - How
many "likes" the video has
importance_matrix - How
many views the video has
0.8683
1
Colors of thumbnail, how many likes videos have and how many views videos have, seems to be the
most important predictors.
Cluster sizes are shown below:
Means of each attribute in each cluster are reported in V6 column.
VV
cellNu me
V V1
V3
V6
V7
V8
V11
V12
1 2
meric1 an
9 0
1
Me 3. 0.06610443
3.
0.06610443
1 Inputs no
importance_title
importance_title
4
an 43 28107574
43
28107574
1
Me 3. 0.06610443
3.
0.06610443
2 Inputs no
importance_title
importance_title
4
an 89 28107574
89
28107574
1
Me 4. 0.02604158 importance_abo
4. importance_abo 0.02604158
1 Inputs no
6
an 55 10066024
ut
55
ut
10066024
1
Me 4. 0.02604158 importance_abo
4. importance_abo 0.02604158
2 Inputs no
6
an 37 10066024
ut
37
ut
10066024
2 1 Label yes
1
Me 2. 0.48188893 importance_rece
2. importance_rece 0.48188893
1 Inputs no
0
an 07 2834321
ntly
07
ntly
2834321
1
Me 3. 0.48188893 importance_rece
3. importance_rece 0.48188893
2 Inputs no
0
an 35 2834321
ntly
35
ntly
2834321
1 1 Label
1
1 Inputs
2
1
2 Inputs
2
yes
2 5 Inputs
no
1 5 Inputs
no
2 9 Inputs
no
1 9 Inputs
no
2 7 Inputs
no
1 7 Inputs
no
Descri
ption
Descri
2 2
ption
1
2 Inputs
5
1
1 Inputs
3
1
1 Inputs
5
1
2 Inputs
1
1
2 Inputs
3
1
1 Inputs
1
1 2
no
no
Me 2. 0.13079436 importance_crea
an 68 9200504
tor
Me 3. 0.13079436 importance_crea
an 43 9200504
tor
Me 2. 0.86832755
importance_likes
an 89 7725445
Me 1. 0.86832755
importance_likes
an 40 7725445
Me 3. 0.52766622 importance_reco
an 12 1446234
mmended
Me 1. 0.52766622 importance_reco
an 83 1446234
mmended
Me 2. 0.66661471
importance_ad
an 52 000987
Me 1. 0.66661471
importance_ad
an 24 000987
2. importance_crea 0.13079436
68
tor
9200504
3. importance_crea 0.13079436
43
tor
9200504
2.
0.86832755
importance_likes
89
7725445
1.
0.86832755
importance_likes
40
7725445
3. importance_reco 0.52766622
12
mmended
1446234
1. importance_reco 0.52766622
83
mmended
1446234
2.
0.66661471
importance_ad
52
000987
1.
0.66661471
importance_ad
24
000987
Me 4. 0.05652003 importance_audi
an 12 06442166
o
Me 3. 0.12123165 importance_vide
an 45 0557421
o_production
Me 3. 0.05652003 importance_audi
an 77 06442166
o
Me 3. 0.29836396 importance_dura
an 77 1181772
tion
Me 3. 0.12123165 importance_vide
an 99 0557421
o_production
Me 2. 0.29836396 importance_dura
an 85 1181772
tion
Me 1.
importance_vie
1.0
an 50
ws
Me 2. 0.75076200 importance_colo
an 85 0527798
rs_thumbnail
Me 3.
importance_vie
1.0
an 17
ws
Me 2. 0.54659933 importance_thu
an 03 8783934
mbnail
4. importance_audi 0.05652003
12
o
06442166
3. importance_vide 0.12123165
45 o_production
0557421
3. importance_audi 0.05652003
77
o
06442166
3. importance_dura 0.29836396
77
tion
1181772
3. importance_vide 0.12123165
99 o_production
0557421
2. importance_dura 0.29836396
85
tion
1181772
1. importance_vie
1.0
50
ws
2. importance_colo 0.75076200
85 rs_thumbnail
0527798
3. importance_vie
1.0
17
ws
2. importance_thu 0.54659933
03
mbnail
8783934
yes
yes
no
no
no
no
no
no
1 4 Inputs
no
2 6 Inputs
no
2 4 Inputs
no
1 8 Inputs
no
1 3 Size
yes
1 6 Inputs
no
2 3 Size
yes
2 8 Inputs
no
(6
58
65.9%
)
Me 1. 0.75076200 importance_colo
an 44 0527798
rs_thumbnail
(3
40
34.1%
)
Me 3. 0.54659933 importance_thu
an 34 8783934
mbnail
(6
58
65.9%
)
1. importance_colo 0.75076200
44 rs_thumbnail
0527798
(3
40
34.1%
)
3. importance_thu 0.54659933
34
mbnail
8783934
The below graphs show one cell distribution for cluster 1 and cluster 2, respectively for variable
“The importance of how many views videos have”.