A test of homogeneity versus a specified heterogeneity in an epidemic model

A two-parameter epidemic model allowing any specified heterogeneous contact structure is studied. With the use of recursive formulas for the final size distribution, as in (Addy, Longini and Haber, Biometrics, 47:961-974, 1991), the score test of the hypothesis that the heterogeneous structure is nonsignificant is derived. The test may be used if there is uncertainty about the spreading mechanism of an infectious disease and a known heterogeneous structure, such as geographical or social structure or both, would be apparent if the disease spread through person-to-person contacts.