Cosmic ray transport theories that incorporate the effect of adiabatic focusing into the usual pitch angle scattering formalism suggest that the commonly assumed first-order (cosine of pitch angle) anisotropy should be replaced by an exponential function in which the scattering mean free path &lgr; and the index q, which characterized the dependence of the scattering rate upon pitch angle, appear as parameters. The solar cosmic ray event of February 16, 1984, which exhibited an unusually large, persistent anisotropy, provides an ideal observational basis for testing this theoretical prediction. Analysis of energetic particle data from the International Cometary Explorer and from eight high-latitude neutron monitors indicates that an exponential with &lgr;>2 AU and with q in the range 1.3--1.7 does indeed provide a significantly better description of particle anisotropies than the first-order form. Because the local interplanetary mean free path is determined purely from the form of the angular distribution, that is without reference to the shape of the temporal density profile, the technique utilized here allows the relative influences of coronal versus interplanetary transport processes to be more cleanly separated. For the February 16 event the time profile was shaped primarily by coronal processes. Specifically, coronal diffusion with a rigidity-independent diffusion coefficient ~1018 cm2 s-1 and with velocity-dependent escape provides a good fit to both the ICE and the neutron monitor data.