conjoint analysis obtain the respondent’s estimated evaluation of profile 1, one sums the part-worths:

https://www.qualtrics.com/support/research-resources/conjoint-analysis-theory/

Loading...

by DV THOMPSON - ‎2005 - ‎Cited by 408 - ‎Related articles

significantly, empirical evidence indicates that consumers ... as conjoint analysis or discrete choice analysis, model each ... the utility of a product is based on its potential benefits to .... the high level included the 21 most important features.

 

 

The following hypothetical example is for a manufacturer of over-the-counter allergy medication. Allergy sufferers are recruited and presented with the 16 alternatives. Each respondent is asked to rate each of the alternatives on a 0–10 scale, where 0 indicates absolutely no interest in purchasing and 10 indicates extremely high interest in purchasing.

 

 

Conjoint Analysis Theory

About Conjoint Analysis

Conjoint analysis is one of the most widely used advanced techniques in marketing research and allows the researcher to predict choice share for evaluated stimuli such as competitive brands. A Conjoint Analysis can be programmed using standard question types, such as the MaxDiff variation of the Matrix Table question.
When using conjoint analysis, the researcher is concerned with the identification of utilities—values used by people making trade-offs and choosing among objects having many attributes and/or characteristics.
The typical sequence that one goes through to implement a conjoint study involves seven steps:

1. Identification of the problem, along with dimensions of the product to be studied. How many attributes are considered and what are the levels of each attribute?
2. Develop the study protocol including all contact, sampling and follow-up protocols. Also develop the survey and associated visual aids, products, graphics, etc. that are to be used.
3. Develop the questionnaire and then pretest the survey and data collection activity. Evaluate the process and revise until you are satisfied with the approach, instrument, and the methodology.
4. Using one of a variety of data collection procedures described below, collect the data.
5. Process the data to derive at the individual respondent level estimates of the part-worths of each person’s utility function.
6. Segmentation Analysis: The matrix of respondent by attribute-level part-worths may then be related to other subject background data in an effort to identify possible market segments based on similarities in part-worth functions.
7. Build and Run the Choice Simulator using a set of product configurations that represent feasible competitive offerings. These product profiles are entered into a consumer choice simulator, along with the earlier computed individual utility functions. Choice simulators differ, in the simplest case each respondent’s individual part-worth function is used to compute the utility for each of the competing profiles.

There are many methodologies for conducting conjoint analysis, including two-factor-ata- time trade-off, full profile, Adaptive Conjoint Analysis (ACA), choice-based conjoint, self explicated conjoint, hybrid conjoint, and Hierarchical Bayes (HB). In this chapter, two of the most popular methodologies are discussed: the full-profile and self-explicated models.
Conjoint analysis is a methodology for the measurement of psychological judgments, such as consumer preferences. Stimuli (product configurations, advertisements, movie endings, etc.) are presented to the respondent according to some statistical design (factorial structure).
A respondent will be presented with a set of alternative product descriptions (such as automobiles). The automobiles are described by their stimulus attributes (level of gas mileage, size of engine, type of transmission, etc.). The respondent views selected alternatives and choice or preference evaluations are made.
From this information, the researcher determines the respondent’s utility for each stimulus attribute (i.e., what is the relative value of an automatic versus a five-speed manual transmission). Once the utilities for all attributes are determined for all respondents the analysis of the utility data can begin.
Preference curves are identified for each attribute so as to show how the different levels are valued. This analysis may be conducted for all respondents or for market segments.
Simulations are then run to determine the relative choice share of a competing set of new or existing products. In full-profile conjoint analysis, the objective is to decompose a set of overall responses to a set of stimuli (product or service) so that the utility of each attribute describing the stimulus can be inferred from the respondent’s overall evaluations of the stimuli.
Conjoint analysis models are constrained by the amount of data required in the data collection task. Managers demand models that define products with increasingly more stimulus attributes and levels within each attribute. Because more detail increases the size, complexity, and time of the evaluation task, new data collection methodologies and analysis models are continually being developed. Online data collection of conjoint data has greatly eased this researcher burden.

Conjoint Analysis Data Collection Methodologies

Two Attribute Tradeoff Analysis

One early conjoint data collection method presented a series of attribute-by-attribute (two attributes at a time) tradeoff tables where respondents ranked their preferences of the different combinations of the attribute levels. For example, if each attribute had three levels, the table would have nine cells and the respondents would rank their tradeoff preferences from 1 to 9. The two-factor-at-a-time approach makes few cognitive demands of the respondent and is simple to follow but it is both time-consuming and tedious. Moreover, respondents often lose their place in the table or develop some stylized pattern just to get the job done. Most importantly, however, the task is unrealistic in that real alternatives do not present themselves for evaluation two attributes at a time.

Full-Profile Conjoint Analysis

Full-profile conjoint analysis has been a popular approach to measure attribute utilities. In the full-profile conjoint task, different product descriptions (or even different actual products) are developed and presented to the respondent for acceptability or preference evaluations. Each product profile is designed as part of a fractional factorial experimental design that evenly matches the occurrence of each attribute with all other attributes. By controlling the attribute pairings, the researcher can estimate the respondent’s utility for each level of each attribute tested.

Adaptive Conjoint Analysis

Adaptive Conjoint Analysis was developed to handle larger problems that required more descriptive attributes and levels. A unique contribution of ACA was to adapt each respondent’s interview to the evaluations provided by each respondent. Early in the interview, the respondent is asked to eliminate attributes and levels that would not be considered in an acceptable product under any conditions. The attributes are then presented for evaluation, followed by sets of full profiles, two at a time, for evaluation. The choice pairs are presented in an order that increasingly focuses on determining the utility associated with each attribute.

Choice-Based Conjoint

Choice-based conjoint requires the respondent to make a choice of their preferred full-profile concept. This choice is made repeatedly from sets of 3–5 full profile concepts. This choice activity is thought to simulate an actual buying situation, thereby mimicking actual shopping behavior.

Self-Explicated Conjoint Analysis

Self-explicated conjoint analysis offers a simple but surprisingly robust approach that is very simple to implement and does not require the development of full-profile concepts. First, factors and levels are presented to respondents for elimination if they are not acceptable in products under any condition.
The attribute levels retained in the analysis are then evaluated for desirability. Finally, the relative importance of attributes is measured using a constant sum scale to allocate 100 points between the most desirable levels of each attribute.
The attribute level desirabilities are then weighted by the attribute importances to provide utility values for each attribute level. This approach does not require regression analysis or aggregated solution required in many other conjoint approaches. This approach has been shown to provide results equal or superior to full-profile approaches, and places fewer demands on the respondent.

Hierarchical Bayes Conjoint Analysis (HB)

Hierarchical Bayes Conjoint Analysis (HB) is similarly used to estimate attribute level utilities from choice data. HB is particularly useful in situations where the data collection task is so large that the respondent cannot reasonably provide preference evaluations for all attribute levels. The HB approach uses averages (information about the distribution of utilities from all respondents) as part of the procedure to estimate attribute level utilities for each individual. This approach again allows more attributes and levels to be estimated with smaller amounts of data collected from each individual respondent.

Examples of Conjoint Analysis

Full Profile Conjoint Analysis

Full Profile Conjoint Analysis requires only that the respondent provide preference ratings of each item in a set of full profile product description. Once this data is collected, metric conjoint analysis uses dummy variable regression analysis to derive the respondent’s preference utilities.
The respondent’s preference ratings for the product descriptors (full profile descriptors of the service or other item) are used as the dependent (criterion) variable in the analysis. The independent (predictor) variables are the various factorial levels making up each stimulus.
In the nonmetric conjoint analysis, the dependent (criterion) variable represents a ranking of the alternative profiles and is only ordinal-scaled. The full-profile methods for collecting conjoint analysis data will be illustrated to show how conjoint data are obtained.
The example illustrated below shows sixteen cards that profile the product that is the object of the study. The set of cards represents a “factorial design” that balances how many times each attribute appears with each other attribute. The details of each card are shown on the left side of the figure. If we are using a non-metric conjoint analysis we would ask the respondent to group the 16 cards into three piles (with no need to place an equal number in each pile) described in one of three ways:

1. Definitely like
2. Neither definitely like nor dislike
3. Definitely dislike

The criterion task is usually to rank based on preference or purchase likelihood rating. Following this, the respondent takes the first pile and ranks the cards in it from most to least liked, and similarly so for the second and third piles. By means of this two-step procedure, the full set of 16 cards is eventually ranked from most liked to least liked.

Again, the analytical objective is to find a set of part-worths or utility values for the separate attribute (factor) levels so that, when these are appropriately added, one can find a total utility for each combination or profile. The part-worths are chosen so as to produce the highest possible correspondence between the derived ranking and the original ranking of the 16 cards. While the approach just described assumes only ranking-type data, one could just as readily ask the respondent to state preferences on an 11-point equal-interval ratings scale, ranging from like most to like least, or a 0-to-100 rating scale, representing likelihood of purchase.
If more than six or seven attributes are involved the number of profiles to be evaluated becomes large and the full profile design must be modified to handle specific subsets of interlinked factors across two or more evaluation tasks.
The following hypothetical example is for a manufacturer of over-the-counter allergy medication. Allergy sufferers are recruited and presented with the 16 alternatives. Each respondent is asked to rate each of the alternatives on a 0–10 scale, where 0 indicates absolutely no interest in purchasing and 10 indicates extremely high interest in purchasing.
Illustratively, the 16 items are to be evaluated by each respondent. Although a total of 4 × 4 × 4 × 4 = 256 combinations of attribute levels could be made up, the respondent needs to evaluate only 16 of these. However, this specific set of 16 should be selected in a particular way (using a fractional factorial experimental design).

The following table shows utility values for each of the attribute levels derived for one respondent. These values can be obtained from an ordinary multiple regression program using dummy-variable coding. All one needs to do to estimate the respondent’s utility score for a given concept profile is to add each separate value (the regression coefficient) for each component of the described combination (the regression’s intercept term may be added in later if there is interest in estimating the absolute level of purchase interest). For example, to obtain the respondent’s estimated evaluation of profile 1, one sums the part-worths:


In this instance we obtain an almost perfect prediction of a person’s overall response to profile 1. Similarly, we can find the estimated total evaluations for the other 15 options and compare them with the respondent’s original evaluations. The regression technique guarantees that the (squared) prediction error between estimated and actual response will be minimized.
The derived attribute level utilities also permit the researcher to find estimated evaluations for all combinations, including the 256 – 16 = 240 options never shown to the respondent. Moreover, all respondents’ separate part-worth functions (as illustrated for the average of all respondents in Table 19.6) can be compared in order to see if various types of respondents (e.g., high-income versus low-income respondents) differ in their separate attribute evaluations.
In summary, full profile conjoint analysis requires the respondent to evaluate complete bundles of attributes. The approach is then to solve for a set of part-worths—one for each attribute level. These part-worths can then be combined in various ways to estimate the evaluation that a respondent would give to any combination of interest. The full-profile approach requires a knowledge of experimental designs to develop the profiles and software to perform the regression analyses that derives the utilities. We will now consider the self-explicated model as an approach that provides superior accuracy, but does so with a much easier design and data collection task.

Self-Explicated Conjoint Analysis

The self-explicated model provides a simple alternative producing utility score estimates equal to or superior to that of other conjoint analysis methods. The self-explicated model is based theoretically on the multi-attribute attitude models that combine attribute importance with attribute desirability to estimate overall preference. This model is expressed as
where Ij is the importance of attribute j and Djk is the desirability of level k of attribute j. In this model, Eo, the evaluation of product or service o, is formed by summing the importance weighted desirabilities of the attributes and attribute levels that make up the profile.

The Self-Explicated Data Collection Task

Step 1: All attribute levels are presented to respondents for evaluation to eliminate any levels that would not be acceptable in a product under any conditions. Step 2: All attribute levels are presented to the respondent and each level is evaluated for desirability (0 –10 scale). Step 3: The most desirable level of every attribute (as reported by the respondent) is evaluated in a constant sum question to assign relative attribute importances.
Using this self reported information, the attribute importance scores are used to weight the standardized attribute level scores. This process produces what are called “self-explicated utility values” for each attribute level. Utility values are computed for each respondent. The self explicated conjoint analysis does not require a fractional factorial design or regression analysis.
As with all conjoint models, attribute level utility values can be summed and simulations run to obtain a score for any profile of interest. This simple self-reporting approach is easier for the respondent to complete and straightforward in terms of determining the importance or desirability of attributes and attribute levels (Srinivasan, 1997). An easy to use online implementation of the self-explicated model is part of our online survey software. For this implementation, the conjoint analysis is automatically developed after the attribute level descriptors are entered into the question builder.
Online conjoint analysis is especially appealing because graphic images, video and audio clips can be presented to add realism and authenticity to the choice decision. Furthermore, blocks of questions and individual attribute levels can be randomized to control for presentation order bias.