In: Statistics and Probability
For each analysis, write out the null hypothesis, explain what the analysis is looking at, calculate anything you think needs calculating (like expected cell counts), identify whether you think the null is right, and then tell the manager what the results mean.
11. Correlations - I have put the following correlations into a matrix form. Each r is the correlation between the row variable and the column variable. The p value for each r is located immediately below the r.
Age |
$ spent in category |
$ spent on our brand |
Household size |
|
Age |
--- |
r = .33 p = .05 |
r = .15 p = .57 |
r = .18 p = .23 |
$ Spent in category |
r = .33 p = .13 |
--- |
r = .48 p = .01 |
r = .67 p = .01 |
$ spent on our brand |
r = .15 p = .57 |
r = .48 p = .01 |
--- |
r = .04 p = .74 |
Household size |
r = .18 p = .23 |
r = .67 p = .01 |
r = .04 p = .74 |
--- |
12. Multiple Regression
Source |
Variation |
F |
p |
Predicted |
735 |
4.75 |
.003 |
Error |
4265 |
||
Total |
5000 |
||
Dependent Variable = $ spent on product category
R2 = 735/5000 = 14.7%
b |
Beta |
p |
|
Household size |
3.5 |
.245 |
.02 |
Household Income |
257.83 |
.475 |
.005 |
13. Multiple Regression
Source |
Variation |
F |
p |
Predicted |
132 |
2.32 |
.34 |
Error |
3155 |
||
Total |
3287 |
||
Dependent Variable = $ spent on product category
R2 = ?
b |
Beta |
p |
|
Commuting Distance |
.356 |
.044 |
.974 |
Value of Home |
1.83 |
.0052 |
.835 |
14. Multiple Regression
Source |
Variation |
F |
p |
Predicted |
575 |
12.87 |
.01 |
Error |
2000 |
||
Total |
2575 |
||
Dependent Variable = $ spent on product category
R2 = ?
b |
Beta |
p |
|
Hours spent on recreation |
5.75 |
.762 |
.000 |
Hours spent listening to music |
8.66 |
.003 |
.961 |
Source |
Variation |
F |
p |
Main Effect (A) |
27.5 |
2.11 |
.24 |
Main Effect (B) |
58.9 |
6.75 |
.03 |
A x B Interaction |
45.7 |
4.98 |
.05 |
Within |
325 |
||
Total |
457.1 |
Dependent Variable = Purchase Intention
Means
Ad 1 |
Ad 2 |
|
Price 1 |
7.5 |
3.5 |
Price 2 |
4.0 |
8.2 |
if p value < 0.05, there is correlation between two variables
p value >0.05 no correlation
12)
p = 0.003
conclusion : p-value<α , reject null hypothesis
Model is significant.
Both Household size and Household Income have less p value means both variable are significant in model.
13)
p value =0.34
p value > 0.05
not significant
so, not a good modal
also, commuting distance and value of home are not significant , we find that the fitted model is not significant. The goodness of fit of the model can be determined using the Coefficient of Determination measure R2 which gives the proportion of explained variation in the model:
%
We find that the predictors in the model could explain only about 4% variation in $ spent on product category. This supports the conclusion that the fitted model is not significant and the model cannot be used effectively for predicting the amount spent.
Looking at the fitted regression equation,
Predicted $ spent on product category = 0.044 ( Commuting Distance ) + 0.0052 ( Value of Home )
We find that both the estimated slope coefficients 0.044 (p-value = 0.974> 0.05) and 0.0052 (p-value = 0.835 > 0.05) fail to contribute significantly in predicting the $ spent on product category.
14.
Here, using Multiple regression, we may establish a causal relationship between the probable predictors and the dependent variable - $ spent on product category (at say, 5% level of significance). Looking at the test results, F = 12.87, p-value = 0.01 < 0.05, we find that the fitted model is significant at 5% level. The goodness of fit of the model can be determined using the Coefficient of Determination measure R2 which gives the proportion of explained variation in the model:
%
We find that the predictors in the model explain about 22.3% variation in $ spent on product category. Looking at the fitted regression equation,
Predicted $ spent on product category = 0.762 ( Hours spent on recreation ) + 0.003 ( Hours spent listening to music )
We find that the estimated slope coefficient of Hours spent on recreation, 0.762 (p-value = 0.000 < 0.05) contribute significantly in predicting the $ spent on product category but not Hours spent listening to music, 0.0052 (p-value = 0.961 > 0.05).
Although this model is significant, since this explains only 22.3% variation, rerunning the model by eliminating the insignificant predictor (Hours spent listening to music) we might get a more efficient model.
MULTI-FACTOR ANOVA:
We test:
Factor A has no effect on a significant effect on Purchase Intention Factor A has not effect on Purchase Intention There is no interaction between factors A and B
Vs
Factor A has a significant effect on Purchase Intention Factor A has a significant effect on Purchase Intention There is a interaction between factors A and B
From the F test of overall significance obtained from ANOVA table, we find that effect of Factor B (F = 6.75, p-value = 0.03) and Interaction AB (F = 4.98, p-value = 0.05) on Purchase Intention is significant at 5% level.
Multi factor ANOVA A = type of ad; B = price level
Main Effect (A) |
27.5 |
2.11 |
.24 |
p value > α , Insignificant |
Main Effect (B) |
58.9 |
6.75 |
.03 |
p value < α , significant |
A x B Interaction |
45.7 |
4.98 |
.05 |
p value = α , insignificant |
If you have any doubts please comment and please don't dislike.
PLEASE GIVE ME A LIKE. ITS VERY IMPORTANT FOR ME.