Question

Nordique Products, Inc. is trying to decide which of two available strategies it will implement company-wide in the future – Strategy A or Strategy B.

To identify the best strategy, managers selected two samples of store locations (20 locations in each sample) and implemented one strategy in each sample.

After several months, the firm has compiled data from the different locations within each sample, and has run regression analyses using this data. A variety of independent / explanatory factors are included in the regression models.

The main dependent variable of interest is store profit. (You can assume that lag between the implementation of the strategy and financial results is minimal.)

Relevant information from the regression analysis for stores implementing Strategy A is as follows:

Count = 20

R-squared = 0.692

Adjusted R-squared = 0.651

	Coefficient	Std Error	t-value	p-value
Intercept	101,452	9,546	8.24	0.01
Strategy implementation score	9,659	946	4.10	0.01
Experience score	4,500	321	3.96	0.01
Market size	0.453	0.042	0.97	0.21
Customer satisfaction	25,409	3,422	2.24	0.04

Relevant information from the regression analysis for stores implementing Strategy B is as follows:

Count = 20

R-squared = 0.692

Adjusted R-squared = 0.651

	Coefficient	Std Error	t-value	p-value
Intercept	94,440	8,200	7.04	0.01
Strategy implementation score	10,099	1,031	9.28	0.01
Experience score	3,228	546	4.00	0.01
Market size	0.672	0.100	2.10	0.05
Customer satisfaction	31,094	3,652	2.95	0.03

Required

1. Suppose you are making a prediction about financial profit for a future month, based on the analysis from Strategy A. What would your next step be in relying on regression analysis to assess the effectiveness of Strategy A?

2. Suppose you are making a prediction about financial profit for a future month based on the analysis from Strategy B. What would be your prediction based on a store with the following characteristics:

• Strategy implementation score = 83

• Experience score = 9.2

• Market size = 20,000

• Customer satisfaction rating = 8

3. In interpreting and/or relying on regression analysis (in general), what qualitative concerns should Nordique managers consider?

Answer 1

Sol:

For strategy A regression equation for predicting profit

profit=101452+9659*Strategy implementation score +4500*Experience score+0.453*market size+25403*customer satsifaction

For strategy B regression equation for predicting profit
profit=94440+10099*Strategy implementation score +3228*Experience score+0.672 *market size+31094*customer satsifaction

Prediction scores for strategy A

profit=101452+9659*Strategy implementation score +4500*Experience score+0.453*market size+25403*customer satsifaction

profit=101452+9659*83 +4500*9.2+0.453*20000+25403*8

=1156833

Prediction scores for strategy B

profit=94440+10099*Strategy implementation score +3228*Experience score+0.672 *market size+31094*customer satsifaction

profit=94440+10099*83 +3228*9.2+0.672 *20000+31094*8

=1224547

In interpreting and/or relying on regression analysis (in general), what qualitative concerns should Nordique managers consider?

R sq

here rs qfor both model is

R sq=0.692

=69.2% variation in profit is explained by model

Overall F test for the model

Indivudual t tests for slopes.

For sttargey A

For startegy B

startegy B is best as all are signficant variables at 5% and R sq=69.2%