My primary research interest is in the formation and evolution of massive stars and massive star clusters throughout the universe. A theme in much of my recent work has been exploring the connections between massive star formation in the Milky Way, local universe, and more extreme environments typical in starburst galaxies and the earlier universe. In order to probe many stages of massive star evolution, my research has spanned the full range in wavelengths, including the radio, millimeter, near- and mid-infrared (IR), visible, and ultra-violet.

Massive stars are a primary source of energetic phenomena in the universe, and they have a major role in the evolution of galaxies: they are responsible for the ionization of the interstellar medium; their stellar winds and supernovae are main sources of mechanical energy; they are responsible for heating much of the warm dust in galaxies via their strong UV radiation; they are a main driver of chemical evolution in the universe via supernova explosions at the end of their lives; and they may be the sources of gamma ray bursts. Nevertheless, despite the significant role of massive stars throughout the universe, their birth is not well understood.

Relatively nearby starburst galaxies are a crucial tool for understanding the evolution of stars and galaxies. In hierarchical models of structure formation, mergers of dwarf-like galaxies in the early universe are responsible for the distribution of galaxy masses we see today. These merger events, and their resulting starburst episodes, may also play a role in the reionization of the universe at z > 5. Furthermore, metals produced and expelled in these early starbursts may provide an explanation for the heavy element abundances observed in Ly-alpha clouds.

In the local universe, the star formation in many starburst galaxies has been resolved into starburst knots or ``super star clusters'' (SSCs). Recent investigations have suggested that some of these SSCs may in fact be proto-globular clusters. With the aid of the Hubble Space Telescope, a large number of galaxy systems are now known to host these compact adolescent SSCs. However, once these massive star clusters have emerged to become visible in optical light, the epoch of their formation is long past. We have recently discovered a sample of natal SSCs that are still deeply embedded in their birth material. We may be witnessing the birth of the globular clusters.

Compact groups of galaxies provide a rich environment in which to study galaxy interactions and merger events. Compact groups of galaxies are among the densest concentrations of galaxies known, comparable to the centers of rich galaxy clusters. However, unlike galaxy clusters, compact groups have relatively low velocity dispersions, increasing the likelihood of gravitational interactions between group members. Because of their relative proximity, compact groups provide us with a unique environment to study the possible conditions in which a substantial amount of galaxy formation took place at high redshift.

In order to interpret the wealth new observations on embedded star forming regions becoming available from infrared to the radio wavelengths, it is crucial to have appropriate physical models available. In order to propertly model natal star formation, one of the main difficulties that must be overcome is that the interstellar medium is not smooth, but rather fractal and clumpy. Therefore, 1-D models cannot reproduce physical structures that are consistent with the observed ISM. We have produced a suite of models of natal SSCs using the 3-D Monte Carlo radiation transfer code described by Whitney et al. (2003). These models allow for arbitrary density distributions, and we have invoked a clumpy fractal structure, consistent with observations of the ISM.