Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Conjoint Analysis is a quantitative research method used to understand the relationship between consumers' preferences and the attributes of products. the concept, assumptions, data requirements, and process of Conjoint Analysis, including the definition of factors and factor values, survey design, estimation of utility values, and aggregation of utility values. It also provides an example of Conjoint Analysis for attitudes towards dishwashing products.
Typology: Study notes
1 / 39
This page cannot be seen from the preview
Don't miss anything!
Research question: To what extent does eachcomponent (factor) contribute to the total utility of aproduct?
Ā
Assumption:
Total utility = Sum of all partial utilities
Ā
Data base of the Conjoint Analysis are preferences ofthe interviewed subjects
Ā
Important application: Design of a new productaccording to the requirements of the market
Factors and their values are defined by the researcherin advance
Ā
The various combinations of the factor values yieldfictive products that are being ranked by theinterviewed persons
Ā
With Conjoint Analysis it is possible to derive metricpartial utilities from the ranking results
Ā
The summation of these partial utilities thereforeresults in metric total utilities
Ā¾
Relevance
Ā¾
Interference
Ā¾
Independence
Ā¾
Realisability
Ā¾
Compensatory relationships of the various factorvalues
Ā¾
They do not constitute exclusion criteria
Ā¾
Terminableness
Example: Attitudes towards dishwashing products
Clean: glass/dishes clean
Shiny: glass/dishes shiny
Smell: Non-perfumed/lemon fresh/intensive lemonfresh
Quantity: small/medium/x-large
Packaging: loose in box/tab in plastic/tab in dissolvingplastic
Design: single/multi-colored/multi-colored + ball
Number of incentives Ā¾
Can increase very fast depending on the numberof factors and their values. It is advisable tochoose a subset (reduced design) out of alltheoretically possible incentives (full design).There shouldnāt be more than 20 fictive products.
Ā¾
Choice of the number can be based on randomselection (not very common) or systematicselection
Ā¾
Possible are symmetric (= all factors have thesame number of values) and asymmetric(divergent number of values) designs
Ā¾
Addelman developed several āBasic Plansā forthe construction of reduced asymmetric designs
Ā
Example (cont.): Attitudes towards dishwashingproducts
Ā¾
The full design consists of 2x2x3x3x3x3=324different incentives
Ā¾
The reduced design consists of 16 incentives/cards,therefore we get a ranking from 1ā¦16 for eachperson
Ā¾
Ranking
Ā¾
Rough classification into groups of different utilitywith succeeding ranking within these groups.Aggregation of these results leads to a totalranking. Used when there is a large number ofincentives.
Ā¾
Rating scales
Ā¾
Paired comparison
Ā¾
Conjoint Analysis is used to determine partialutilities (āpartworthsā)
for all factor values based
upon the ranked data
Ā¾
Furthermore, with this partworths it is possible tocompute the metric total utilities
of all incentives
and the relative importance of the single objectattributes
Ā¾
Individual Conjoint Analysis: For each person utilityvalues are computed
Ā¾
Combined Conjoint Analysis: Only one value foreach factor category
y
Ā¾
Target criterion for the determination of the partialutilities: The resulting total utilities should yield agood representation of the empirically ranked data
Ā¾
Related procedure for the determination of thepartial utilities: monotonous analysis of variance
Ā¾
Metric solution
metric ANOVA
Non-metric solution
monotonous ANOVA (not
covered here)
Ā¾
Model for the metric solution Steps needed to get the solution:1.
Determination of
(= average of all ranks)
Determination of the mean rank for each factorvalue (= which ranks did the interviewed personassign on the average in conjunction with thisattribute?)
1
j 1
M
J
j
m
jm
k
j
m
y
x
Ī¼
Ī²
=
=
=
ā
Ā¾
For the comparability of the individual analyses ofthe single persons it is necessary to carry out astandardisation:
The factor value that yields the lowest utilitycontribution is set to zero
The differences between the single partial utilityvalues and the lowest partial utility value arecomputed:
M
in
jm
jm
j
The total utility value of the incentive that showsthe strongest preference among all persons isset to one. The partial utility values are computedaccordingly:
Ā¾
The importance of an attribute for the preferencechange results from the utility range of the severalvalues of the attribute
1
Ė
max
jm
jm
J
jm
m
J
Ī²
Ī²
Ī²
=
Ā¾
The largest standardised partial utility values allowto make a statement about the relative importanceof the attributes at the same time
Ā¾
Aggregation always implies the loss of information
Ā¾
If the aggregated utility structures are highlyheterogeneous, this loss of information issubstantial
Ā¾
In this case, more homogeneous subgroups shouldbe formed by performing a Cluster Analysis
Ā
Example (cont.): Attitudes towards dishwashingproducts
Ā¾
There is no graphical user interface available inSPSS that would allow the performance of aConjoint Analysis. The SPSS-Syntax has to be usedin order to retrieve the required procedureCONJOINT.
Ā¾
After the Syntax-window has been opened, thefollowing commands should be entered:
