Measurements of anisotropy in ~1 MeV corotating events at 1 AU indicate an average direction more or less perpendicular to the interplanetary magnetic field and outwards from the sun. In a general diffusion-convection model the steady-state anisotropy is determined by spectral index &ggr;, diffusion coefficient &kgr;ij, and particle velocity v (and possibly shock reflection coefficient &bgr;). If the anisotropy mimics that produced by the E¿B drift it must be fortuitous (scatter-free propagation does not predict the right behavior). Several models are examined, but when fitted to three comparable sets of anisotropy observations, there is a diversity of best-fit &lgr;∥ and &kgr;⊥/&kgr;∥ values. First, a discrepancy in &kgr;⊥/&kgr;∥ that may exist between 6-hour and 1-hour observations would be consistent with the concept of random walking of field lines. Second, anisotropies measured parallel to the field probably underestimate those pertinent to pure corotating events owing to (1) a solar-injected component, (2) magnetospheric effects, and (3) spiral field geometry variations. The last seems particularly important. Thus the best fit value of &lgr;∥ would be an underestimate too, and this may account for the discrepancy in &lgr;∥ deduced from the different sets of data. After considering these effects, out best estimate of &lgr;∥ is 0.05--0.13 AU and &kgr;⊥/&kgr;∥<0.25 at 1 AU. Our best fit estimate of &lgr;∥ agrees with model independent values obtained independently from corotating events and is also consistent with the range &lgr;∥=0.08--0.3 AU at 1 AU obtained from impulsive flare events.