This dissertation addresses three questions about administrative rulemaking: • Do comments submitted on proposed rules vary in identifiable ways? • Do these differences directly relate to the likelihood that recommendations will be associated with changes to regulations? • Can these characteristics be incorporated into a model that accurately predicts whether or not suggestions will coincide with changes to administrative rules? Using data collected from the Commonwealth of Virginia's Regulatory Town Hall, I analyze 2,534 comments that address 67 regulations proposed by state agencies during an 11-year period. I find that submissions do differ in meaningful ways. I also find statistically significant evidence that those differences are related to the probability that a requested change coincides with a subsequent modification to a rule. The principal result of this research is a model that predicts with a high degree of accuracy the outcome of participants' recommendations to alter proposed regulations. I also demonstrate the implications of these results and how failure to account for these differences undermines the legitimacy of conclusions that can be drawn from studies of notice-and-comment rulemaking. The primary contribution of this dissertation is methodological, but the empirical evidence presented here also raises questions about the value of citizen participation in notice-and-comment rulemaking in its current form. As a result, it challenges contentions that participation contributes to the democratic legitimacy of bureaucracy, serves as a safeguard against the influences of organized interests, or improves the substantive quality of administrative decisions. / Ph. D.