Variance and Dissent: "Spydergrams" for Short Form 36 Questionnaire Results
- M. Boers in his discussion of the use of “spydergrams” to simultaneously display the Medical Outcomes Survey Short Form-36 (SF-36) domain scores makes a number of cogent observations about both the heuristic utility of this graphical representation and raises cautions about the possibility of their misinterpretation. Some of his concerns are generic, in that they could apply to any graphical representation of data—but specific reservations about spydergrams in particular are expressed. Given that spydergrams have proven a powerful tool for depicting and comparing the complex sets of self-reported health-related quality of life (HRQOL) data, it is important to clarify their strengths as graphical representations and clarify recommendations for their legitimate interpretation.
- In a recent article, Strand et al.  have suggested radar graphs (also termed radial graphs) to plot the results of the Short Form-36 questionnaire (SF-36). They have coined the term “spydergram” which relates to the web-like appearance of the graph but is slightly confusing because the term “spidergram” has been used for connected text boxes such as in a mind map . Spydergrams constitute one of the attempts to graphically depict more than two dimensions on a flat surface. Tufte , an authority on scientific graphs, has termed this as “escaping flatland.” To escape flatland, compromises are inevitable, and the challenge is to compromise without distorting the story in the data.