Business Research Methods

BUSINESS RESEARCH METHODS

ASSIGNMENT 1

Submitted by :

Amol Bankar (10BM60008)

Anuradha Chakraborty ( 10BM60014)

Mayank Mohan (10BM60048)

Niloy Ghosh (10BM60054)

Conjoint Analysis

Conjoint analysis is a statistical technique used in market research to determine how people

value different features that make up an individual product or service. The objective of

conjoint analysis is to determine what combination of a limited number of attributes is most

influential on respondent choice or decision making. A controlled set of potential products or

services is shown to respondents and by analyzing how they make preferences between these

products, the implicit valuation of the individual elements making up the product or service

can be determined.

In the carpet cleaner example, we are evaluating 5 attributes and observing which attributes

are more valuable to the consumer. The following 18 combinations are presented to the

consumer along with 4 holdout cases which are used to test our analysis.

Card List

Card ID package design brand name price

Good

Housekeeping

seal

money-back

guarantee

1 1 A* Glory $1.39 yes no

2 2 B* K2R $1.19 no no

3 3 B* Glory $1.39 no yes

4 4 C* Glory $1.59 no no

5 5 C* Bissell $1.39 no no

6 6 A* Bissell $1.39 no no

7 7 B* Bissell $1.59 yes no

8 8 A* K2R $1.59 no yes

9 9 C* K2R $1.39 no no

10 10 C* Glory $1.19 no yes

11 11 C* K2R $1.59 yes no

12 12 B* Glory $1.59 no no

13 13 C* Bissell $1.19 yes yes

14 14 A* Glory $1.19 yes no

15 15 B* K2R $1.39 yes yes

16 16 A* K2R $1.19 no no

17 17 A* Bissell $1.59 no yes

18 18 B* Bissell $1.19 no no

19a 19 A* Bissell $1.59 yes no

20a 20 C* K2R $1.19 yes no

21a 21 A* Glory $1.59 no no

22a 22 A* Bissell $1.19 no no

a. Holdout

The 5 attributes that are tested and their apparent relation to the value of the product are given

as:

Model Description

N of Levels

Relation to Ranks

or Scores

package 3 Discrete

brand 3 Discrete

price 3 Linear (less)

seal 2 Linear (more)

money 2 Linear (more)

All factors are orthogonal.

The response of 10 individuals is taken and the part worth of the various factors is

determined as:

Utilities

Utility Estimate Std. Error

package A* -2.233 .192

B* 1.867 .192

C* .367 .192

brand K2R .367 .192

Glory -.350 .192

Bissell -.017 .192

price $1.19 -6.595 .988

$1.39 -7.703 1.154

$1.59 -8.811 1.320

seal no 2.000 .287

yes 4.000 .575

money no 1.250 .287

yes 2.500 .575

(Constant) 12.870 1.282

The regression coefficients for the linear factors come out to be:

Coefficients

B Coefficient

Estimate

price -5.542

seal 2.000

money 1.250

The percentage importance of each factor is determined as:

Importance Values

package 35.635

brand 14.911

price 29.410

seal 11.172

money 8.872

Averaged Importance

Score

Thus, we can see that package and price are the two most important factors for the consumer

while a money-back guarantee has the least importance. The producer may now concentrate

on the appropriate factors so as to provide maximum value to the consumers.

The validity of the analysis may be determined by the correlation between observed and

estimated preferences as shown by the following statistics.

Correlationsa

Value Sig.

Pearson's R .982 .000

Kendall's tau .892 .000

Kendall's tau for Holdouts .667 .087

a. Correlations between observed and estimated

preferences

CLUSTER ANALYSISCLUSTER ANALYSISCLUSTER ANALYSISCLUSTER ANALYSIS

Cluster Analysis can be of two types:

A. Hierarchical Cluster Analysis: It helps in revealing natural groupings. Therefore, here, the

no. of clusters is automatically found out by the software itself.

B. Non-hierarchical : Here, the required fixed no. of clusters is given.

Here, we tried both the analyses on the data file dmddata.sav.

Hierarchical Cluster Analysis

1. MODEL SUMMARY

2. THE VARIABLES:

A. Categorical Variables:

• Income

• Marriage

• Gender

B. Continuous Variables:

• Age

• Children

• Year of Residence

K-Means Cluster Analysis

1. MODEL SUMMARY

3. THE VARIABLES:

.

A. Categorical Variables:

• Income

• Married

• Response to previous survey

B. Continuous Variables:

• Age

• Children

Factor Analysis

The analysis given below was done for data sample file car_sales.sav in PASW 18.0. Following variables were considered for

1) Price in thousands

2) Engine size

3) Horsepower

4) Width

5) Wheelbase

6) Length

7) Fuel capacity

8) Fuel efficiency

The eigen values used for extraction were greater than 1 using principal components analysis. The rotation method used was varimax ( Kaiser Normalization) and regression analysis method was made for relating variables. The Scree Plot for the same is given below where the components with steeper slopes are

recognized as important factor while making a decision. In the plot given below, the slope is steeper for components 1 and 2 and these account for almost 88 % of variance of other factors.

Total Variance Explained

Component Initial Eigenvalues Extraction Sums of Squared

Loadings

Rotation Sums of Squared

Loadings

Total % of

Variance

Cumulative

%

Total % of

Variance

Cumulative

%

Total % of

Variance

Cumulative

%

1 4.987 62.341 62.341 4.987 62.341 62.341 3.386 42.329 42.329

2 1.506 18.822 81.163 1.506 18.822 81.163 3.107 38.834 81.163

3 .584 7.303 88.466

4 .332 4.155 92.621

5 .241 3.019 95.640

6 .150 1.880 97.520

7 .127 1.589 99.109

8 .071 .891 100.000

Extraction Method: Principal Component Analysis.

The rotation component matrix is shown below :

Rotated Component Matrix

Component

1 2

Price in thousands -.008 .925

Engine size .497 .776

Horsepower .228 .920

Width .782 .385

Wheelbase .935 .052

Length .895 .121

Fuel capacity .707 .480

Fuel efficiency -.546 -.637

The above matrix shows the correlation of the factors with principal components. The above

matrix shows that wheelbase, horsepower and price in thousands are the main factors when

decision for choosing a new car is made.