SpAM is convenient but also satisfying: Reply to Verheyen et al. (2016).

Hout, Goldinger, and Ferguson (2013) critically examined the spatial arrangement method (SpAM), originally proposed by Goldstone (1994), as a fast and efficient way to collect similarity data for multidimensional scaling. We found that SpAM produced high-quality data, making it an intuitive and user-friendly alternative to the classic "pairwise" method. Verheyen, Voorspoels, Vanpaemel, and Storms (2016) reexamined our data and raised 3 caveats regarding SpAM. In this reply, we suggest that Verheyen et al. mischaracterized our reported data as representing the entire range of potential SpAM data. SpAM results might appear more nuanced with modified instructions or stimuli. By contrast, the pairwise method is inherently limited because of its laborious, serial nature. We also demonstrate that, when the methods are equated in terms of required data-collection time, SpAM is clearly superior in terms of predicting classification data. We agree that caution is required when adopting a new method but suggest that fair assessment of SpAM requires a richer data set.