Approximate and exact optimal designs for paired comparisons of partial profiles when there are two groups of factors

Großmann, Heiko; Graßhoff, Ulrike; Schwabe, Rainer

doi:10.1016/j.jspi.2008.07.006

Cited by 16 publications

(25 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It would be interesting to investigate whether we can still improve the estimation performance of the partial profile designs using other types of master designs. Grossmann et al , presented master designs that are

scriptD

‐optimal for very specific utility‐neutral design situations. We could use these master designs as a basis to find partial profile designs that perform better in terms of the

{scriptD}_{B}

‐optimality criterion.…”

Section: Summary and Discussionmentioning

confidence: 99%

“…The most recent utility‐neutral partial profile designs created by Grossmann et al . only allow for two or three groups of attributes where the number of attribute levels is fixed in every group. There are also limitations on the allowable numbers of choice sets in all cases considered and on the allowable numbers of varying attributes when there are two groups of attributes.…”

Section: Introductionmentioning

confidence: 99%

“…Holding the level of the dominant attribute constant in part of the choice sets, in a partial profile design, forces the respondents to evaluate the remaining attributes, resulting in useful information concerning these attributes. Obviously, information concerning the non-dominant attributes is as important as information concerning the dominant attribute, because successfully competing products and items often have the exact same key characteristics (i.e., the same level for the dominant attribute), and a wise choice of secondary features may result in a significant increase in general attractiveness.So far, most of the literature for DCEs using partial profiles has focused on utility-neutral designs involving choice sets with two profiles [17][18][19][20][21]. These designs are optimal for one specific set of parameter values, namely the set of all zero values.…”

mentioning

confidence: 99%

See 2 more Smart Citations

An improved two‐stage variance balance approach for constructing partial profile designs for discrete choice experiments

Kessels

Jones

Goos

2014

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

In a discrete choice experiment, each respondent chooses the best product or service sequentially from many groups or choice sets of alternative goods. The alternatives, called profiles, are described by level combinations from a set of predefined attributes. Respondents sometimes make their choices on the basis of only one dominant attribute rather than making trade-offs among all the attributes. For example, in studies involving price as an attribute, respondents may always choose the profile with the lowest price. Also, a choice task including many attributes may encourage respondent decisions that are not fully compensatory. To thwart these behaviors, the investigator can hold the levels of some of the attributes constant in every choice set. The resulting designs are called partial profile designs. In this paper, we construct D-optimal partial profile designs for estimating main-effects models. We use a Bayesian design algorithm that integrates the D-optimality criterion over a prior distribution of likely parameter values. To determine the constant attributes in each choice set, we provide three alternative generalizations of an approach that makes use of balanced incomplete block designs. Each of our three generalizations constructs partial profile designs accommodating attributes with any number of levels and allowing flexibility in the numbers of choice sets and constant attributes. We show results from an actual experiment in software development performed using one of these algorithms. Finally, we compare the algorithms with respect to their statistical efficiency and ability to avoid failures due to the presence of a dominant attribute.Keywords: discrete choice experiments, Bayesian D-optimal design, partial profiles, lexicographic choice behavior, balanced incomplete block design, coordinateexchange algorithm 2

show abstract

scriptD

‐optimal for very specific utility‐neutral design situations. We could use these master designs as a basis to find partial profile designs that perform better in terms of the

{scriptD}_{B}

‐optimality criterion.…”

Section: Summary and Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

An improved two‐stage variance balance approach for constructing partial profile designs for discrete choice experiments

Kessels

Jones

Goos

2014

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

show abstract

“…We refer to Burgess and Street (2005); Grasshoff et al (2003Grasshoff et al ( , 2004; Grossmann et al (2006Grossmann et al ( , 2009and Street and Burgess (2007) for D-optimal designs derived under that assumption. Under the assumption that the model parameters are zero, the information matrix for the multinomial logit model is proportional to the information matrix for a linear regression model, in case the experiment is run in blocks (Kessels et al, 2011b).…”

Section: Literature Reviewmentioning

confidence: 99%

Bayesian D-Optimal Choice Designs for Mixtures

2014

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Abstract Consumer products and services can often be described as mixtures of ingredients. Examples are the mixture of ingredients in a cocktail and the mixture of different components of waiting time (e.g., in-vehicle and out-of-vehicle travel time) in a transportation setting. Choice experiments may help to determine how the respondents' choice of a product or service is affected by the combination of ingredients. In such studies, individuals are confronted with sets of hypothetical products or services and they are asked to choose the most preferred product or service from each set. However, there are no studies on the optimal design of choice experiments involving mixtures. We propose a method for generating an optimal design for such choice experiments. To this end, we first introduce mixture models in the choice context and next present an algorithm to construct optimal experimental designs, assuming the multinomial logit model is used to analyze the choice data. To overcome the problem that the optimal designs depend on the unknown parameter values, we adopt a Bayesian D-optimal design approach. We also consider locally D-optimal designs and compare the performance of the resulting designs to those produced by a utility-neutral (UN) approach in which designs are based on the assumption that individuals are indifferent between all choice alternatives. We demonstrate that our designs are quite different and in general perform better than the UN designs. Terms of use: Documents in

show abstract

“…As one of the main goals of paired comparison studies is to determine the part-worths, the optimal design approach is appropriate to design them. Applications of optimal design to paired comparison experiments can be found in Offen and Littell (1987), van Berkum (1987van Berkum ( , 1989, El-Helbawy, Ahmed, and Alharbey (1994), Schwabe (2003, 2004), Sándor and Wedel (2001), Kessels, Goos, and Vandebroek (2006), Kessels, Jones, Goos, and Vandebroek (2009) Burgess (2004, 2007) and Großmann, Graßhoff, and Schwabe (2009). It should be pointed out that some of the publications in this enumeration do not restrict attention to paired comparison experiments only, but deal with choice experiments in general (where respondents evaluate sets of two or more profiles).…”

Section: Introductionmentioning

confidence: 99%

Optimal design of factorial paired comparison experiments in the presence of within-pair order effects

Goos

Großmann

2011

Food Quality and Preference

Self Cite

View full text Add to dashboard Cite

Approximate and exact optimal designs for paired comparisons of partial profiles when there are two groups of factors

Cited by 16 publications

References 10 publications

An improved two‐stage variance balance approach for constructing partial profile designs for discrete choice experiments

An improved two‐stage variance balance approach for constructing partial profile designs for discrete choice experiments

Bayesian D-Optimal Choice Designs for Mixtures

Optimal design of factorial paired comparison experiments in the presence of within-pair order effects

Contact Info

Product

Resources

About