New Ways to Measure Consumers’ Judgments GREEN, PAUL E. AND YORAM WIND (1975), “NEW WAYS TO MEASURE CONSUMERS’ JUDGMENTS, “HARVARD BUSINESS REVIEW, JULY-AUGUST,

New Ways to Measure Consumers’ Judgments

GREEN, PAUL E. AND YORAM WIND (1975), “NEW WAYS TO MEASURE CONSUMERS’ JUDGMENTS, “HARVARD BUSINESS

REVIEW, JULY-AUGUST, 73-184.

PLEASE CONSIDER READING MY HANDOUT “ADVANCE RESEARCH METHODS” WHILE STUDYING THIS ARTICLE. POWER POINTS FOR THE HANDOUT ARE IN BLACKBOARD AND ON THE

WEB

Authors

Wharton School Professor of Marketing Expert in mathematics and statistics 12 years of Industry experience Author of 160 books and 1400 articles

Wharton School Professor of Marketing Expert in buyer behavior and market

research Editorial Board Member: Journal of

Marketing, Journal of Business Research Author of 22 books and 250 research papers

Yorum Wind

Paul Green

Outline of the Article

The article discusses conjoint analysis and other multivariate techniques used in business/marketing. Conjoint analysis Multiple regression Factor analysis Perceptual mapping Cluster analysis Discriminant analysis

This power point presentation also includes materials from my handout “Advance Research Methods.”

Conjoint Analysis

Conjoint analysis is “a new measurement technique from the fields of mathematical psychology and psychometrics that can aid the marketing manager in sorting out the relative importance of a product’s multidimensional attributes.”

It starts with the consumers’ overall judgments about a set of complex alternatives, and, then decomposes consumers’ original evaluations into separate and compatible utility scales from which the original global judgments can be reconstituted.

How Consumer Measurement Works

Consumers rate three designs on the basis of 5 factors. Package design, brand name, and price have 3 levels. Good housekeeping seal and money back guarantee has 2 levels. A total of 18 consumer cards are used giving a combination of 108 (3x3x3x2x2) alternatives. Consumers are asked to rank the cards. Each card is a combination, an orthogonal array, a point in a multidimensional space; it is a choice based on utilities.

Importance of attributes

•Utility combination 18 is 3.6•Package: C (0.6)•Brand: Bissell (0.5)•Piece: $ 1.19 (1.0)•Good Housekeeping: Yes (0.3)•Money-back: yes (0.7)

•All utility scale have same unit•Relative importance of one factor depends on the levels included in the design•Packed design B highest utility

Air Carrier Study Example

Factors to evaluate Aircraft/Carrier Departure time relative to ideal Punctuality of arrival Passenger load Number of stops en route Attitudes of flight attendants Entertainment

Findings of the study Two factors comprised the biggest utility differences: punctuality

of arrival and attitudes of flight attendants Utility difference between the two jet planes is very small Other factors to consider: departure time and number of stops

Replacement Tire Study Example

Design to protest TV commercials for sponsor’s brand tires

Brand name did not play important roleTread mileage & price are important

Potential Use and Limitations

Can be useful in evaluating consumer judgments for….- New product formulations- Package design, brand and promotion- Pricing and brand alternatives- Verbalized descriptions of new products- Alternative service designs

Limitations include…- Number of attributes is too large- Too new method- Not fit all products and services- Not all product can be decomposed

Multiple Regression

Compile explanatory variables to identify a response variable.

Sales forecasting for cable subscribers (Y). Y=number of subscribers

X1=Advertising rate for one minute of prime time X2=Kilowatt power X3=Number of families in living area of coverage X4=Number of competing stations

Factor Analysis

Data reduction technique

Used to determine the underlying dimensions/factors defining the attributes/variables

Attributes provided by researcher

Objective is to find commonalities in factors

Perceptual Mapping

Use consumer judgments

Find objects that are similar

Very helpful in brand positioning

Cluster Analysis

Used to determine homogeneous groups

Used in segmentation and profiling

Clusters are based on similarity

Discriminant Analysis

Develop criterion variables which are categorical Based on criterion develop predictor variables Criterion variables must fall into two groups

Identify lung cancer: Yes or no outcome Run tests (predictor variables) Based on predictor variables we get results to validate

one option

Can we predict how many people will go bankrupt in Texas next year?

Discussion Questions

What is conjoint analysis? What are the common applications and limitations of conjoint analysis? How can we overcome its limitations?

Why should marketers study multivariate analysis techniques? Define/elaborate various techniques we discussed in the class and explain how they help marketers make “informed decisions.”