Because the number of asteroids in the IMB with absolute magnitude H<16.5 is effectively complete, the distributions of the sizes and the orbital elements of these asteroids must be devoid of observational selection effects. This allows us to state that the observed size-frequency distributions (SFDs) of the five major asteroid families in the IMB, defined by Nesvorný (2015) using the Hierarchical Clustering Method (Zappala et al. 1990), are distinctly different and deviate significantly from the linear log-log relation described by Dohnanyi (1969). The existence of these differences in the SFDs, and the fact that the precursor bodies of the major families have distinctly different eccentricities and inclinations, provides a simple explanation for the observations that the mean sizes of the family asteroids, taken as a whole, are correlated with their mean proper eccentricities and anti-correlated with their mean proper inclinations. While the latter observations do have a simple explanation, we observe that the mean sizes of the non-family asteroids in the IMB are also correlated with their mean proper eccentricities and anti-correlated with their mean proper inclinations. We deduce from this, and from the fact that the SFDs of the non-family and the family asteroids (again taken as a whole) are almost identical, that the family and most of the non-family and asteroids have a common origin. We estimate that ~85% of all the asteroids in the IMB with H<16.5 originate from the Flora, Vesta, Nysa, Polana and Eulalia families with the remaining ~15% originating from either the same families or, more likely, a few ghost families (Dermott et al. 2018).