Decision Analyst releases free ChoiceModelR software for choice modeling

Decision Analyst has made its ChoiceModelR analytics package available for free via its website and CRAN. The R-based tool analyzes conjoint, choice modeling and MaxDiff experiments and is aimed at users working with large, complex datasets.

Why it matters: - ChoiceModelR gives researchers and analysts a no-cost option for estimating choice models without buying commercial software. - The package is aimed at large datasets and computationally intensive projects, where speed and model flexibility matter. - The release may lower barriers for teams that already collect choice data and need a free analysis tool.

What happened: - Decision Analyst announced ChoiceModelR, an R-language choice-modeling package developed and supported by the company’s software engineers. - The package is available as a free download from the Decision Analyst website and from CRAN: CRAN package page. - ChoiceModelR analyzes conjoint, choice modeling and MaxDiff experiments. - The software is an analytics package, not a data collection system.

The details: - Chris Hammack, an Advanced Analytics team member, said the program is ideal for large datasets with complex variables. - Beth Horn, Ph.D., head of Advanced Analytics at Decision Analyst, said the software is ideal for large datasets where computational speed is a major need. - Horn said ChoiceModelR replicates the functionality of commercial choice-modeling programs. - Horn said the software assumes the choice-modeling experiment has already been conducted and the data are ready to analyze. - John Colias, Ph.D., the primary architect of the software, said ChoiceModelR uses a Markov chain Monte Carlo algorithm to estimate a hierarchical multinomial logit model with a normal heterogeneity distribution. - Colias said the software uses a hybrid Gibbs sampler with a random-walk Metropolis step for multinomial logit coefficients. - Colias said the dependent variable can be discrete, either nominal or ordinal, or continuous, with share values between 0 and 1. - Colias said constraints may be imposed on model parameters. - Colias said the number of choice observations per respondent can vary. - Colias said the number of choice alternatives per observation is flexible. - Horn said users need a basic understanding of choice modeling and the R language, and must download R software to run ChoiceModelR. - Horn said both ChoiceModelR and R are open source and free. - Decision Analyst provided contact information for Beth Horn at bhorn@decisionanalyst.com and Chris Hammack at chammac@decisionanalyst.com for more information. - Decision Analyst is based in Dallas-Fort Worth and describes itself as a global marketing research and analytical consulting firm focused on advanced modeling, market segmentation, strategy research and new product development research. - The firm says it serves clients in CPG, technology, home improvement, automotive, healthcare, durable goods and retail. - Decision Analyst also listed its website as www.decisionanalyst.com and shared its LinkedIn page at Decision Analyst on LinkedIn.

Between the lines: - The release positions ChoiceModelR as a free alternative to commercial tools, which could matter most for researchers with existing data but limited software budgets. - The emphasis on open-source access and R compatibility suggests Decision Analyst is targeting users who already work in statistical programming environments. - The technical description signals that the package is designed for experienced practitioners rather than casual users.

What's next: - Users interested in the software can download ChoiceModelR from CRAN or the Decision Analyst website. - Researchers needing help with implementation can contact Decision Analyst directly for more information. - Wider adoption will likely depend on whether the package performs reliably on the large, complex datasets it is designed to handle.