Post on 14-Apr-2017
Experimental TestingIzmir VodinajMKMR 310Fall 2015
SUMMARY
OBJECTIVEDetermining if coupon usage
increases the sales or not.
Sample Categories• No coupon usage • Newspaper coupons• Store mailing coupons• Newspaper and store mailing
coupons.
Customer Shopping Intention • Self• Self and Spouse• Self and Family
Key Procedures• Recoding• ANOVA• GLM• Graphing
KEY DATA
Before RecodingUse of Coupons
After RecodingUse of Coupons
Question # 1
Does coupon usage have an effect on amount spent by the customer for in-store purchases?
H0: Coupon Usage does not affect the amount spent by the customer on in-store purchases.
Ha: Coupon Usage does affect the amount spent by the customer on in-store purchases.
Question # 2
INDEPEDENDENT VARIABLE COUPON USAGE
DEPENDENT VARIABLE AMOUNT SPENT
Coupons do affect the amount of purchases made by customers!
Question # 2 A
Lavene’s Test is significant. Therefore, variances are significantly different at different levels, and we will use Games-Howell Post Hoc.
It can be seen that when coupons are given, customers spent more, yet one must check the significance level. Furthermore, the highest amount seems to be when both coupons are used.
1
3
2The results are significant, and they indicate that there is a difference between the groups, but one does not know where the significance lies. F(2,1401) = 7.768; p < 0.001
4 We can see here that the difference between no coupon and both coupons is the only significant value. Therefore, we can determine that the most impact is made when both coupons are used, thus rejecting the null hypothesis.
Does the purpose of shopping have an effect on amount spent by the customer for in-store purchases?
H0: The purpose of shopping does not affect the amount spent by the customer on in-store purchases.
Ha: The purpose of shopping affects the amount spent by the customer on in-store purchases.
Question # 3
INDEPEDENDENT VARIABLE SHOPPING PURPOSE
DEPENDENT VARIABLE AMOUNT SPENT
Purpose of shopping affects the amount of purchases made by customers!
Lavene’s Test is significant. Therefore, variances are significantly different at different levels, and if we had multiple factors will would Games-Howell Post Hoc.It can be seen that when customers shop
for their spouses and their family, their amount spend is higher, yet one must check the significance level.
1
3
2The results are significant, and they indicate that there is a difference between the groups. Thus we conclude that the purpose of shopping makes a difference on the amount spent on in-store purchases, and we reject the null hypothesis. F(1,1402) = 66.912; p < 0.001
Question # 3 A
Does shopping purpose influence the effect of coupon usage on amount spent by the customer for in-store purchases?
H0: The purpose of shopping does not influence the effect of coupon usage on amount spent by the customer on in-store purchases.
Ha: The purpose of shopping influences the effect of coupon usage on amount spent by the customer on in-store purchases.
Question # 4
INDEPEDENDENT VARIABLE SHOPPING PURPOSE
DEPENDENT VARIABLE AMOUNT
SPENT
INDEPEDENDENT VARIABLE
COUPON USAGE
SHOPPING PURPOSE &
COUPON USAGE
Question # 4A
The Interaction is more than .01, thus we fail to reject the null hypothesis. This concludes that the shopping purpose does not influence the effect of coupon usage on the amount spent by the customer. After doing a cost-benefit analysis, the manager should target all of its segments and have both mail and newspapers coupons sent.
Shopping purpose does not influence the effect of coupon usage on amount spent by the customer for in-store purchases!
Highest amount spent
Lowest amount spent
It would be helpful to note here that even though the interaction is not significant, the model accounts for 5.6% of the variance in amount spent. Obviously, we would like to
be able to explain much more of the variance, yet we know many factors influence spending and we would need to compare to another model to see how good this is.
Does shopping style influence the effect of coupon usage on amount spent by the customer for in-store purchases?
Bonus
The Interaction is more than .01, thus we fail to reject the null hypothesis. This concludes that the shopping style does not influence the effect of coupon usage on the amount spent by the customer.
Once again looking at the R square value tells us quite a bit. In this case only .9% of the variance in amount spent is explained by the model. So, not as
useful as the last model.