Displayr

The incrementality plot breaks when alternatives from the previous portfolio are not included. This request is to add an option to TURF that forces the previous portfolio alternatives to be included, so this issue doesn’t arise.

My client needs a preference share output for all N combinations in the model, without using a simulator. This feature would help in making final decisions on reporting by providing a comprehensive view of preference shares across all combinations.

My client wants the simulator to output the total utility score for the options chosen, in addition to the preference share. This feature would allow us to have both the preference share and total utility score versions of dashboard simulators for future conjoint studies.

I am interested in running unsupervised random forests in Displayr, where no outcome variable is selected, focusing on understanding the structure of the data or relationships between observations. It would be beneficial if Displayr could support this feature, allowing for more exploratory data analysis.

Users can coerce the inclusion of a single alternative in a TURF portfolio. But it's not possible to coerce the inclusion of "at least one" from a set of alternatives. The feature request is to suggest the inclusion of more advanced constraints like the above in the TURF feature.

Currently a conjoint simulator is created given a number of alternatives. Would be great if you could include/exclude these alternatives as needed say if people wanted to run a 3 alternative simulation, and also a 4 alternative simulation using the same simulator. this would be pretty easy to implement in the widget/R code, I’ve already written a little custom code to do it, but ideally it would be built-in.

To compute standardized coefficients, a user needs to first standardized the input variables, then run the model. Suggest including standardized coefficients are an optional statistic output when running regressions in Displayr.

Allow option to use Sawtooth constraints on conjoint models

As a user, I may run a regression output and decide to exclude outliers from the analysis. At the moment, there is no mechanism to see exactly which respondents were removed due to the outlier removal. Although it's possible to back into this using some R code, it should be easier for users to see which cases were removed.

Currently, exploratory factor analysis is only available in Displayr indirectly by accessing relevant R packages. This feature request suggests adding EFA as part of the built-in interface menu.