CONJOINT ANALYSIS

CONJOINT ANALYSIS

Introduction

Many time consumers seem to be confused in determining the relative importance of the attributes used in describing the features of the product or services. In such cases Conjoint analysis is used to measure the perceived values of specific product features on the basis of selected attributes. Conjoint analysis is a multivariate technique used specifically to evaluate preferences of consumers for products or services. It is based on the simple idea that consumers evaluate a product or service by combining a number of attributes. The analysis allows the combination of attributes to determine the relative importance of each attribute in the purchasing decision.

 

CASE ANALYSIS-1


PROBLEM

The manager of a detergent manufacturing unit is interested to conduct a study to determine the features of the product influencing the people in purchase making decision. We first need to identify the attributes of the product which are important to customers and then the levels for each attribute. Suppose three attributes are important:

  1. price of the product
  2. colour
  3. packaging

The levels of the attributes are:

  1. price of the product—Rs.10/kg, Rs.13/kg and Rs.16/kg (3 levels)
  2. colour—Light blue, white and green (3 levels)
  3. Packaging—paper packing A), ordinary plastic bag (B) and designer packing (C).(3 levels)

We have all total 3×3×3=27 combinations of attribute levels. The objective of the conjoint analysis is to find the amount of utility one can derive from the given combination of levels of attributes.

Method of Data Collection

The respondents are then asked to rank these 27 combination levels of attributes. Rank 1 is considered as highest ranking and rank 27 as the lowest ranking in ordinal scale. The following data sheet has been prepared by calculating the average ranking of the respondents. The ranking so collected are to be arranged in reverse order as shown in table-1.

Table-1: Collected Data

Running Conjoint Analysis as Regression Model

It is quite easy to convert the conjoint analysis into an equivalent regression model. The coding of the attributes for regression analysis to be used for this purpose is known as the effect coding. The variables are to be assigned with different coding with the condition that the sum of codes for a particular variable will be zero and the number of variables will be one less than the levels of the variable. The effect coding of different levels of attributes are given in table-2, 3, and 4.

Table-2: Coding for Price

Thus two variables V1 and V2 are used to indicate 3 levels of price. Similarly the coding for 3 levels of colours is shown in table-3.

Table-3: Coding for Colours

The coding scheme for packaging is as follows.

Table-4: Coding for Packaging

Thus six variables V1, V2, V3, V4, V5, V6 are used to represent different levels of chosen attributes.

The set of variables V1, V2, V3, V4, V5, V6 with their respective effect coding and the rating in reverse order (V7) represent the input matrix for the analysis. The variables V1, V2, V3, V4, V5, and V6 are independent variables in a regression run. The rating in reverse order (V7) is to be used as the dependent variable.

Table-5: Input Data coded for Regression Analysis

Performing the Analysis with SPSS

For SPSS Version 11, click on Analyze ⇒ Regression ⇒ Linear

This will bring up the SPSS screen dialogue box as shown below.

After clicking Linear, this will bring up the SPSS screen dialogue box as shown below.

Select V7 and move it into the Dependent box. Select the variables V1, V2, V3, V4, V5, V6 and move them into Independent(s) box. Choose Enter method. This will bring you the following dialogue box.

Now click on OK button. This will bring the outputs.

 SPSS Output

Table-6: Variables Entered/Removed

                             a All requested variables entered.

                             b Dependent Variable: V7

Table-7: Model Summary

                            a Predictors: (Constant), V6, V4, V2, V5, V3, V1

Table-8: ANOVA

                             a Predictors: (Constant), V6, V4, V2, V5, V3, V1

                             b Dependent Variable: V7

Table-9: Coefficients

Now the column titled ‘B’ (Unstandardized Coefficients) provides the part utilities of each level of attributes. The utilities for V1 (Price- Rs.10/kg) and V2 (Price- Rs.13/kg) are 5.44 and 0.778. The utility for the third level (Price- Rs.16/kg) is to be derived from the property of the coding, that all the utilities for a given attribute should sum to zero. Thus the utility for (Price- Rs.16/kg) would be equal to 0-(5.44 + 0.778) = -6.281.

The utilities for all the levels of variables are shown in the following table.

Table-10: Utilities

We can come to several conclusions with these part utilities.

  1. Colour of the detergent is the most important attribute for the customers as it has the highest range of utilities of 17.22 and the second most important attribute is price.
  2. The colour ‘light blue’ has the maximum part utility of 7 and so it is the prominent level of the attribute and the minimum price (Rs.10/ kg) is the next important level of the attribute.

 Combination of Utilities

The different levels of attributes can be combined to give the best combination as given in table-11.

Table-11: Utilities

The best combination is the detergent powder with colour light blue, price of Rs.10/kg and the paper packing.

Limitation of Using Conjoint Analysis

The researchers have to decide carefully about the attributes and its different levels. The attributes relevant for the purpose of the study should be selected. The number of combinations to be ranked by the respondents should not be more than 40.


CASE ANALYSIS-2


PROBLEM

The marketing manager of DEL computer Company wants to identify the product features attracting the customers to choose that DEL laptop. Let us assume that three attributes of computer are important:

  1. Price of the product
  2. Warranty period of use
  3. Dimension of the product-length of the laptop screen measured in inches.

The levels of the attributes are:

  1. Price of the product — Rs.27,000, Rs.32,000 and Rs.40,000 (3 levels)
  2. Warranty period of use — 1 year, 2 years, 3 years (3 levels)
  3. Dimension of the product — 10 inches, 12 inches. (2 levels)

We have all total 3×3×2=18 combinations of attribute levels.

 Method of Data Collection

The respondents are then asked to rank these 18 combination levels of attributes. Rank 1 is considered as highest ranking and rank 18 as the lowest ranking in ordinal scale. The average ranking of the responses arranged in reverse order is as follows.

Table-1: Collected Data

Effect coding

The effect coding of different levels of attributes are given in table-2, 3, and 4.

Table-2: Coding for Price

Table-3: Coding for warranty period of use

Table-4: Coding for Dimension of the product

Thus five variables V1, V2, V3, V4 and V5 are used to represent different levels of attributes.

Table-5: Input Data coded for Regression Analysis

SPSS Output

Table-6: Variables Entered/Removed

                            a All requested variables entered.

                            b Dependent Variable: V6

Table-7: Model Summary

                             a Predictors: (Constant), V5, V4, V1, V3, V2

Table-8: ANOVA

                             a Predictors: (Constant), V5, V4, V1, V3, V2

                             b Dependent Variable: V6

Table-9: Coefficients

                            a   Dependent Variable: V6

The utilities for all the levels of variables are shown in the following table.

Table-10: Part Utilities

  1. Price of the laptop has the highest range of utility of 1.878 and the second most important attribute is Dimension of the product with utility 1.576.
  2. The price ‘Rs.27, 000’ has the maximum part utility of 1.833 and the warranty period of 2 years is the next important level of the attribute with utility 1.402.

Combination of Utilities

The different combination of attributes is given in following table.

Table-11: Utilities

The best combination of laptop is Rs.27, 000 with 2 years warranty period and dimension of 12 inches.

SPSS Command

  1. Click on ANALYZE at the SPSS menu bar (in older versions of SPSS, click on STATISTICS instead of ANALYZE).
  2. Click on REGRESSION followed by LINEAR
  3. Select the dependent variable (variable used for rank) and move it to dependent box. Similarly select the independent variables and move them to independent box.
  4. Choose the method ENTER.
  5. Select OK of the main dialogue box.

 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.