Although text types (e.g., persuasive and expository), think-aloud procedures, and expertise levels have been the focus of numerous studies, less is known about their effects on metacognitive monitoring and control. The purpose of this study was to investigate these issues during participants' text processing of government and politics texts using log files, think-aloud protocol, and individuals' calibration as data sources. Participants were undergraduates enrolled in human development (n = 38) and government/politics courses (n = 38) presumed to represent different levels of domain expertise. Participants read two passages on judicial review presented via computer while thinking aloud. Trace data on scrolling behaviors were collected during reading and confidence measures after reading. Think-aloud data were analyzed via non-parametric bootstrapping. Significant differences between text types were found for scrolling, calibration, and reading behavior. There was no significant difference for the think-aloud condition on scrolling or calibration. Only scrollback behavior was statistically different between levels of expertise. However, median differences revealed interesting trends between expertise groups in terms of calibration bias.