Automobile manufacturers have a responsibility to consumers to remedy product defects through recalls. This research studies consumer product recall decision making in the automotive industry. The recall decisions considered include: defect identification and investigation, uninfluenced recalls, proactive recalls and recall timing. Textual, statistical and analytical analysis methods are used to address three major decision-making scenarios. First, the antecedents and consequences of the recall decisions, uninfluenced and proactive, are empirically modeled and analyzed. Next, the recall timing decision is empirically and analytically modeled, with an emphasis on the timing choice within the framework of a cost minimization problem. Finally, a consideration of the choice between two text-based modeling alternatives, supervised and unsupervised, is presented which aims to provide a foundation for a framework for consumer product defect trend monitoring and identification.