Before you unfold a fan, as shown in Figure 1, you can see a compressed set of images stretched out on a line from the center to the periphery. These images may appear as nothing more than ordered blotches. You can imagine what you might see when you unfold the fan, but almost certainly the real image will confound your imagination. Liking and other hedonic measures, expressed as ordered means, are like the blotches on an unopened fan. We will not know what the drivers of liking space looks like and how the items are arranged in it until we unfold the data.
In a previous paper, we reviewed some of the more common methods for generating spatial maps of hedonic data and considered the extent to which they are based on a well-defined process. We concluded that the use of a model based on the process that respondents use to generate hedonic data, rather than relying on models that contain no such process considerations, is important to obtaining a meaningful interpretation of hedonic data. In addition, by following a process-based approach, researchers can evolve their thinking about what their data means by testing and improving their models. One of our recommendations was to consider ideal point ideas in hedonic models, particularly those that incorporate uncertainty into the location of items and ideals.
This technical report appears as:
Ennis, D. M. (2014). Unfolding. IFPress, 17(3) 3-4.
Not a Colleague? Click here to join for free!