Provides stated preference games to your experimental design
Discrete choices are sometimes obtained from stated preference or stated choice experiments which can require complex statistical designs. Visual Choice can generate these designs to your specification. The graphical interface helps you set up your choice specification. You then select the type of design (e.g. orthogonal or efficient design), interactions etc. and Visual Choice will give you the design. If you supply a template (or use one of ours), Visual Choice will even output the games for you.
The SP Design pane is where everything is done (figure 1). Set up your design specification with the wizard (figure 2) including the alternatives and attributes with their different levels and values. Select the type of design from:
- Orthogonal full factorial design
- Orthogonal fractional factorial design
- Efficient design with unknown coefficient values
- Efficient design with known coefficient values
- Efficient design with Bayesian coefficient values
- Efficient design with known coefficient ranges
Press the Run button to produce your design. If you wish you can change the optimising algorithm and the evaluation criteria and other features to give you control of your design. This includes some innovatory features for optimisation and to ensure you produce balanced designs
You can change your optimising algorithm from:
- Modified Federov
- Simulated annealing
- PDC algorithm
And your evaluation criteria from:
Deals with up to 20 alternatives at up to 20 levels each. You can set the range of coefficient values you want the design to recover, change the number of blocks in a block design. For the Bayesian design select your number of random draws and type from one of the following:
- Halton number generator
- Randomised Halton
- Pseudo random numbers
Estimate your stated preference coefficients using Visual Choice.
The SP Design tab under the Estimation tab where everything is done.
Set up your design with the SP Design wizard.