The connection between theory and data is an iterative one. In principle, each is informed by the other: data provide the basis for theory that in turn generates the need for new information. This circularity is reflected in the notion of abduction, a concept that focuses on the space between induction (generating theory from data) and deduction (testing theory with data). Einstein, in the 1920s, placed scientific creativity in that space. In the field of social network analysis, some remarkable theory has been developed, accompanied by sophisticated tools to develop, extend, and test the theory. At the same time, important empirical data have been generated that provide insight into transmission dynamics. Unfortunately, the connection between them is often tenuous and the iterative loop is frayed. This circumstance may arise both from data deficiencies and from the ease with which data can be created by simulation. But for whatever reason, theory and empirical data often occupy different orbits. Fortunately, the relationship, while frayed, is not broken, to which several recent analyses merging theory and extant data will attest. Their further rapprochement in the field of social network analysis could provide the field with a more creative approach to experimentation and inference. 1. Introduction Theory and empirical data are in principle intimately interwoven. Yet in the practice of social network analysis, there appears to be a disconnect: theorizing and empiricism often seem to occupy separate orbits, and these separate discussions may be difficult to relate to each other. The root of the problem may lie in the different skill sets required by each, or perhaps in the substantial obstacles to collection of human network data. The following exploration of the distance between theory and empiricism suggests that a rapprochement would be of considerable benefit to the field. The mid-19th Century American philosopher Charles Peirce coined the term “abduction” (which he also called “retroduction”) to fill a gap he perceived in the territory occupied by induction and deduction. As distilled by Professor Burch , Peirce used syllogisms to explain this term, substituting Rule, Case, and Result for the more familiar Major Premise, Minor Premise, and Conclusion. But perhaps more interesting to epidemiologists and social network analysts, he related this logical process to sampling. As Professor Burch explains it, a standard valid syllogism would progress as follows.Rule: All balls in this urn are red.Case: All balls in this particular random sample are
S. N. Goodman and D. R. Bellhouse, “Hypothesis tests, and likelihood: implications for epidemiology of a neglected historical debate,” American Journal of Epidemiology, vol. 137, no. 5, pp. 485–501, 1993.
A. C. Ghani, C. A. Donnelly, and G. P. Garnett, “Sampling biases and missing data in explorations of sexual partner networks for the spread of sexually transmitted diseases,” Statistics in Medicine, vol. 17, no. 18, pp. 2079–2097, 1998.
R. B. Rothenberg, J. J. Potterat, D. E. Woodhouse, W. W. Darrow, S. Q. Muth, and A. S. Klovdahl, “Choosing a centrality measure: epidemiologic correlates in the Colorado Springs study of social networks,” Social Networks, vol. 17, no. 3-4, pp. 273–297, 1995.
C. P. Hudson, A. J. M. Hennis, P. Kataaha et al., “Risk factors for the spread of AIDS in rural Africa: evidence from a comparative seroepidemiological survey of AIDS, hepatitis B and syphilis in Southwestern Uganda,” AIDS, vol. 2, no. 4, pp. 255–260, 1988.
L. Sawers and E. Stillwaggon, “Concurrent sexual partnerships do not explain the HIV epidemics in Africa: a systematic review of the evidence,” Journal of the International AIDS Society, vol. 13, article 34, 2010.
M. Morris, A. E. Kurth, D. T. Hamilton, J. Moody, and S. Wakefield, “Concurrent partnerships and HIV prevalence disparities by race: linking science and public health practice,” American Journal of Public Health, vol. 99, no. 6, pp. 1023–1031, 2009.