Hypotheses
Normal
H0: An increase of the PV does not lead to an increase in the OV
H1: An increase of the PV does lead to an increase in the OV
Moderator
H0: The relationship between PV and OV is not moderated by/depends on …
H1: The relationship between PV and OV is moderated by/depends on
Mediator
H0: The effect of the PV on the OV is not mediated by…
H1: The effect of the PV on the OV is mediated by…
, V3 Q3
D.
- Sales, Business B2C, and countries have a significant effect on the odds of the outcome
variable being 1.
- When CI does not contain a zero, then the variable is significant.
- Exp (B) must be below 1 or above 1.
E.
Sales: Exp (B) 1.727 > 1 so this is a positive effect. For every additional unit of Sales (million),
the odds whether or not a firm retained the CEO increases with 1.727 on its scale.
Business is a dummy variable. Exp (B) for B2C is 1.727, which is above 1 so significant. This
means that the odds whether a firm retained the CEO is 1.727 larger in B2C than in B2B.
V1 Q4 (Q3)
A.
- N= 439 which is good
- OV = quantitative, namely income
- PV’s are categorical, namely psychological items
- Bartlett’s Test – Chi-square 6093.838, significance level of 0.000 < 0.05 so significant. This
means that the data is correlated.
- KMO: .914 which is close to 1, so strong shared correlation. Should be 0.5 < KMO< 1.0
B.
Kaiser: All Eigen Values > 1 should be extracted. So based on this: 8 components.
Cumulative %: should normally be between 50 – 60% so based on this: 5 to 9 components.
Cattell criterion: big drop at component 5, so we keep 4 components.
Advice: extract 8 components because two out of three criteria suggest that.
C.
Rotation makes the matrix better to interpretate. Rotating maximizes the loadings on the
one hand and minimizes on the other hand.
As can be seen in the unrotated matrix, some items load on two components, which makes
it difficult to interpret. Besides, the topics in the unrotated matrix are mixed, which makes it
difficult to interpret.
It’s better to use the rotated matrix, because then can Tom interpret the items better, here
he has less difficult items and they are organized.
D.
1. These loadings indicates that they are more correlated on this topic, than on the one
which was expected.
2.
Normal
H0: An increase of the PV does not lead to an increase in the OV
H1: An increase of the PV does lead to an increase in the OV
Moderator
H0: The relationship between PV and OV is not moderated by/depends on …
H1: The relationship between PV and OV is moderated by/depends on
Mediator
H0: The effect of the PV on the OV is not mediated by…
H1: The effect of the PV on the OV is mediated by…
, V3 Q3
D.
- Sales, Business B2C, and countries have a significant effect on the odds of the outcome
variable being 1.
- When CI does not contain a zero, then the variable is significant.
- Exp (B) must be below 1 or above 1.
E.
Sales: Exp (B) 1.727 > 1 so this is a positive effect. For every additional unit of Sales (million),
the odds whether or not a firm retained the CEO increases with 1.727 on its scale.
Business is a dummy variable. Exp (B) for B2C is 1.727, which is above 1 so significant. This
means that the odds whether a firm retained the CEO is 1.727 larger in B2C than in B2B.
V1 Q4 (Q3)
A.
- N= 439 which is good
- OV = quantitative, namely income
- PV’s are categorical, namely psychological items
- Bartlett’s Test – Chi-square 6093.838, significance level of 0.000 < 0.05 so significant. This
means that the data is correlated.
- KMO: .914 which is close to 1, so strong shared correlation. Should be 0.5 < KMO< 1.0
B.
Kaiser: All Eigen Values > 1 should be extracted. So based on this: 8 components.
Cumulative %: should normally be between 50 – 60% so based on this: 5 to 9 components.
Cattell criterion: big drop at component 5, so we keep 4 components.
Advice: extract 8 components because two out of three criteria suggest that.
C.
Rotation makes the matrix better to interpretate. Rotating maximizes the loadings on the
one hand and minimizes on the other hand.
As can be seen in the unrotated matrix, some items load on two components, which makes
it difficult to interpret. Besides, the topics in the unrotated matrix are mixed, which makes it
difficult to interpret.
It’s better to use the rotated matrix, because then can Tom interpret the items better, here
he has less difficult items and they are organized.
D.
1. These loadings indicates that they are more correlated on this topic, than on the one
which was expected.
2.
