K.Y. Lam (Kar Yin)
http://repub.eur.nl/ppl/7833/
List of Publicationsenhttp://repub.eur.nl/eur_logo.png
http://repub.eur.nl/
RePub, Erasmus University RepositoryEstimating Independent Locally Shifted Random Utility Models for Ranking Data
http://repub.eur.nl/pub/30992/
Mon, 17 Oct 2011 00:00:01 GMT<div>K.Y. Lam</div><div>A.J. Koning</div><div>Ph.H.B.F. Franses</div>
We consider the estimation of probabilistic ranking models in the context of conjoint experiments. By using approximate rather than exact ranking probabilities, we avoided the computation of high-dimensional integrals. We extended the approximation technique proposed by Henery (1981)17. Henery , R. J. 1981 . Permutation probabilities as models for horse races . Journal of the Royal Statistical Society. Series B (Methodological) , 43 : 86 – 91 . Retrieved from http://www.jstor.org/stable/2985154
[Web of Science ®]
View all references in the context of the Thurstone-Mosteller-Daniels model to any independent locally shifted random utility model. In particular, this allowed us to estimate any independent random utility model with common shape (e.g., normal, logistic) and scale. Moreover, our approach also allows for the analysis of any partial ranking. Partial rankings are essential in practical conjoint analysis to collect data efficiently to relieve respondents' task burden. We applied the approach to the reanalysis of the career preference data set described in Maydeu-Olivares and Böckenholt (2005)28. Maydeu-Olivares , A. and Böckenholt , U. 2005 . Structural equation modeling of paired-comparison and ranking data . Psychological Methods , 10 : 285 – 304.
Visualizing attitudes towards service levels
http://repub.eur.nl/pub/26471/
Fri, 30 Sep 2011 00:00:01 GMT<div>K.Y. Lam</div><div>M. van de Velden</div><div>Ph.H.B.F. Franses</div>
To assess the attitudes with respect to the quality of banks’ service levels, we use survey data amongst more than 250 Chief Financial Officers (CFOs) of a range of Netherlands-based companies. These companies range from small to very large (including multinationals as Philips and Shell) companies. The survey was conducted in five subsequent years. In this paper, we explore the evaluations of the service levels of banks where, for all attributes considered, the ratings were accompanied by an importance rating. We propose a visualization method that incorporates the importance weights into correspondence analysis. The resulting maps exhibit the correlation structure of the different service items as well as the variances for each item. Moreover, the results are linked to different banks over time, thus exposing the development of the attitudes over time.Reliability and Rankings
http://repub.eur.nl/pub/22977/
Thu, 14 Apr 2011 00:00:01 GMT<div>K.Y. Lam</div>
Questionnaires are an important way to gather information about large populations for both qualitative and quantitative research. Hence, the value of a good questionnaire design and the quality of questionnaire data cannot be emphasized enough. This thesis discusses some aspects of the statistical analysis of measurement data obtained via questionnaires.
In the first part of this thesis we focus on maximizing scale reliability. We derive the asymptotic distribution of maximal reliability measures to construct confidence intervals in order to assess the adequacy of the measure. We stress the use of confidence intervals accompanying single measures that summarize the parameters to assess the adequacy of the measure. The results can lead to better designs of questionnaires, which in turn lead to more precise survey outcomes.
The second part of this thesis proposes methodologies to perform statistical analysis of stated consumer preferences measured as rankings data, especially in the context of conjoint measurements. Our statistical models allow for the efficient use of partial rankings to collect preference data. As a partial rankings task amount to a smaller burden for respondents than a complete ranking task, they may be more motivated to complete the task and as such the quality of the obtained data may improve. Moreover, we show that our model is able to extract sufficient preference information from partial rankings data to take into account respondents' heterogeneity in their choice and preference behavior, which is generally assumed in marketing. This certainly will help marketers to identify and target consumers by understanding their preference behavior, and to implement a more efficient and optimal marketing strategy.Analyzing preference rankings when there are too many alternatives
http://repub.eur.nl/pub/41214/
Wed, 01 Dec 2010 00:00:01 GMT<div>K.Y. Lam</div><div>A. Koning</div><div>Ph.H.B.F. Franses</div>
Consumer preferences can be measured by rankings of alternatives. When there are too many alternatives, this consumer task becomes complex. One option is to have consumers rank only a subset of the available alternatives. This has an impact on subsequent statistical analysis, as now a large amount of ties is observed. We propose a simple methodology to perform proper statistical analysis in this case. It also allows to test whether (parts of the) rankings are random or not. An illustration shows its ease of application. Ranking Models in Conjoint Analysis
http://repub.eur.nl/pub/20937/
Tue, 12 Oct 2010 00:00:01 GMT<div>K.Y. Lam</div><div>A.J. Koning</div><div>Ph.H.B.F. Franses</div>
In this paper we consider the estimation of probabilistic
ranking models in the context of conjoint experiments. By using
approximate rather than exact ranking probabilities, we do not
need to compute high-dimensional integrals. We extend the
approximation technique proposed by \\citet{Henery1981} in the
Thurstone-Mosteller-Daniels model for any Thurstone order
statistics model and we show that our approach allows for a
unified approach. Moreover, our approach also allows for the
analysis of any partial ranking. Partial rankings are essential
in practical conjoint analysis to collect data efficiently to
relieve respondents' task burden.Confidence intervals for maximal reliability in tau-equivalent models
http://repub.eur.nl/pub/23955/
Sun, 01 Nov 2009 00:00:01 GMT<div>K.Y. Lam</div><div>A.J. Koning</div><div>Ph.H.B.F. Franses</div>
Subjective probabilities play an important role in marketing research, for example where individuals rate the likelihood that they will purchase a new developed product. The tau-equivalent model can describe the joint behaviour of multiple test items measuring the same subjective probability. In this paper we stress the use of confidence intervals to assess reliability, as this allows for a more critical assessment of the items as measurement instruments. To improve the reliability one can use a weighted sum as the outcome of the test rather than an unweighted sum. In principle, the weights may be chosen so as to obtain maximal reliability. We propose two new confidence intervals for the maximal reliability in the tau-equivalent model and we compare these two new intervals with intervals derived earlier in Yuan and Bentler (Psychometrika, 67, 2002, 251) and Raykov and Penev (Multivariate Behavioral Research, 41, 2006, 15). The comparison involves coverage curves, a methodology that is new in the field of reliability. The existing Yuan-Bentler and Raykov-Penev intervals are shown to overestimate the maximal reliability, whereas one of our proposed intervals, the stable interval, performs very well. This stable interval hardly shows any bias, and has a coverage for the true value which is approximately equal to the confidence level. Analyzing preferences ranking when there are too many alternatives.
http://repub.eur.nl/pub/11707/
Tue, 18 Mar 2008 00:00:01 GMT<div>K.Y. Lam</div><div>A.J. Koning</div><div>Ph.H.B.F. Franses</div>
Consumer preferences can be measured by rankings of alternatives. When there are too many alternatives, this consumer task becomes complex. One option is to have consumers rank only a subset of the available alternatives. This has an impact on subsequent statistical analysis, as now a large amount of ties is observed. We propose a simple methodology to perform proper statistical analysis in this case. It also allows to test whether (parts of the) rankings are random or not. An illustration shows its ease of application.Confidence intervals for maximal reliability of probability judgments
http://repub.eur.nl/pub/8842/
Tue, 27 Feb 2007 00:00:01 GMT<div>K.Y. Lam</div><div>A.J. Koning</div><div>Ph.H.B.F. Franses</div>
Subjective probabilities play an important role in marketing
research, for example where individuals rate the likelihood that
they will purchase a new to develop product. The tau-equivalent
model can describe the joint behaviour of multiple test items
measuring the same subjective probability. It improves the
reliability of the subjective probability estimate by using a
weighted sum as the outcome of the test rather than an unweighted
sum. One can choose the weights to obtain maximal reliability.
In this paper we stress the use of confidence intervals to assess
maximal reliability, as this allows for a more critical assessment
of the items as measurement instruments. Furthermore, two new
confidence intervals for the maximal reliability are derived and
compared to intervals derived earlier in \\citet{YuanBentler2002,
RaykovPenev2006}. The comparison involves coverage curves, a
methodology that is new in the field of reliability. The existing
Yuan-Bentler and Raykov-Penev intervals are shown to overestimate
the maximal reliability, whereas one of our proposed intervals, the
stable interval, performs very well. This stable interval hardly
shows any bias, and has a coverage for the true value which is
approximately equal to the confidence level.