The ability to accurately recall rich episodic memories and the decision-making strategies used in memory-based tasks are critical in many real-world contexts, such as legal testimony, healthcare, and education. Criterion shifting is a decision-making strategy that adjusts decision thresholds based on individual task demands and has been found to be a uniquely stable cognitive trait. The goal of this research was to systematically explore criterion-shifting behavior in ecologically valid scenarios, focusing on recollection processes and the influence of social external factors. The research begins with an ecologically valid scenario in Experiment 1 (Chapter 2), where criterion manipulations were introduced during an interview to examine how participants adjusted their recall strategies after a stressful autobiographical episodic event. The findings revealed that conservative instructions led to stricter decision criteria and more selective recall, whereas liberal instructions did not significantly lower the already liberal baseline criterion. Given these results, the subsequent experiments aimed to systematically examine how criterion shifting functions in both controlled laboratory settings and more ecologically valid scenarios, providing insights into how individuals adapt their decision-making strategies across different contexts. Experiments 2 and 3 (Chapter 3), in a controlled laboratory setting, isolate the components of social context and criterion shifting. Experiment 2 investigates how competition and public rankings influence criterion shifting during recognition memory tasks, while Experiment 3 considers how collective outcomes influence performance in a group-based task. Neither experiment found significant impacts from social pressure on criterion shifting tendencies, supporting the view that such tendencies are stable cognitive traits. However, Experiment 3 revealed that collective outcomes can enhance discriminability when tasks are relatively easy, suggesting that social influences may improve memory performance under favorable conditions. Experiment 4 (Chapter 4) focused on understanding recollection processes through a within-subject design that systematically increased reliance on recollection across image recognition and word cue tasks. This study tested predictions from the Dual Process Signal Detection (DPSD) model, which views recollection as a thresholded process, and the Unequal Variance Signal Detection (UVSD) model, which treats recollection as a continuous process with Gaussian memory strength distributions. The results revealed consistent criterion shifting across both tasks, indicating that participants employ a generalized decision-making strategy that adapts to varying cognitive demands. The linearity of the z-transformed ROC curves further supports the UVSD model’s predictions, which challenges the DPSD model’s assumption that recollection is a discrete process. Overall, this dissertation contributes to a broader understanding of how criterion shifting operates in diverse contexts, providing insights into the stability of this cognitive trait across different memory tasks and the extent to which social factors and recollection processes influence decision-making. The findings underscore the need for future research to further explore criterion shifting with ecological validity in mind.