Types of Conjoint in Research


A choice-based design approach to larger conjoint designs, the ACBC employs choice-based design principles. It is more robust for pricing tasks and follows a series of stages, starting with a build-your-own section and then moving into choice tasks similar to CBC. ACBC allows for more complex attribute sets and a more intelligent feel for respondents. However, the differences between ACBC and CBC are slight and often not worth mentioning. Find out more at https://www.qualtrics.com/experience-management/research/types-of-conjoint/.

Choice-Based Conjoint (CBC) analysis

CBC analysis is an approach to conjoint decision-making in which respondents rank several options for a single outcome. This approach mimics the decision-making process in humans, resulting in more accurate findings. The choice-based conjoint method is most commonly used to evaluate consumer product attributes.

CBC analyses use ranking and choice procedures to determine the combination of features and benefits a customer wants. In contrast to ranking or rating procedures, CBC studies require large sample sizes and complex algorithms to derive reliable parameter estimators. As a result, choice-based experiments may be more expensive than conventional conjoint analysis. In addition, the complexity of choice-based experiments and their need for specialized software may not be suitable for small-scale experiments.

Another type of conjoint analysis uses the Choice-Based Conjoint method. This method is considered to be more life-like than ratings. Respondents are presented with a list of potential products and asked to select one, two, or three. The resulting data is analyzed to identify which attribute levels respondents value most. CBC is widely used in consumer research. However, it is not as robust as other approaches.

Multidimensional scaling (MDS) analysis

Multidimensional scaling is the mathematical process of reducing a dataset to a single set of variables. This method reduces the number of dimensions to a minimum while maintaining the maximum degree of fidelity to the original relationships. The key to nonmetric MDS is that distances in X are ordered as close to each other as possible. These distances are expressed as curves, which depend on ranks alone. The curves form a rising trend from left to right.

MDS analysis has two primary uses. First, it is used to measure the proportions of variance and is often used as a confirmatory tool. In this case, researchers can use linear regression to test the hypothesis that a given variable is similar to another. In the example above, the data relating to domesticated animals were regressed over the coordinates derived from each dimension. The regression result will give the researchers a picture of the level of similarity between the two groups.

Menu-based conjoint (MBC) analysis

The concept of menu-based conjoint (MBC) analysis is an evolution of the traditional conjoint and was developed to model complex and customized outcomes. Consumers face dozens of choices, making capturing each decision’s nuances difficult. However, MBC is a potent tool in digital commerce, where consumers are increasingly juggling multiple options. Hence, Menu-Based Conjoint is an increasingly relevant research methodology.

You developed the MBC approach in the early 2000s to address a common problem of survey visualization and analysis. It enabled respondents to view a menu of bundles and choose independent channels. Different bundles were presented with varying prices to determine the respondent’s preference. MBC is used to improve existing or develop new offerings. In addition, the research results are used to inform strategic decisions and help achieve the company’s profit and revenue goals.

Hierarchical Bayes (HB) analysis

When using HB, you must first define the FWER or the frequency of a given phenomenon. HB models have many advantages, including the ability to account for many variables. Furthermore, they don’t require full-rank design matrices, which is often a problem with extensive questionnaires. Hence, you can use the individual-level coefficients to calculate preference shares and other purposes. The proposed PPI provides an appropriate method to evaluate the likelihood of an FWER.

One example of an HB analysis is the estimation of the state variables and parameters of a parsimonious hydrological model. The model’s parameters and state variables were estimated using a hierarchical Bayesian framework. The estimated soil moisture levels were correlated with daily streamflow. The analysis showed that the model estimated the parameters of the parsimonious hydrological model, GR4J.



Please enter your comment!
Please enter your name here