Efficient paired choice designs with fewer choice pairs

Authors

Aloke Dey, Indian Statistical Institute (Delhi Centre)
Rakhi Singh, IITB-Monash Research Academy
Ashish Das, Indian Institute of Technology Bombay

Article Type

Research Article

Publication Title

Metrika

Abstract

For paired choice experiments, two new construction methods of designs are proposed for the estimation of the main effects. In many cases, these designs require about 30–50% fewer choice pairs than the existing designs and at the same time have reasonably high D-efficiencies for the estimation of the main effects. Furthermore, as against the existing efficient designs, our designs have higher D-efficiencies for the same number of choice pairs.

First Page

309

Last Page

317

DOI

10.1007/s00184-016-0605-9

Publication Date

4-1-2017

Recommended Citation

Dey, Aloke; Singh, Rakhi; and Das, Ashish, "Efficient paired choice designs with fewer choice pairs" (2017). Journal Articles. 2629.
https://digitalcommons.isical.ac.in/journal-articles/2629