Digital administration of assessments allows for the collection of process data indices, such as response time, which can serve as indicators of rapid-guessing and examinee test-taking effort. Setting a time threshold is essential to distinguish effortful from effortless behavior using item response times. Threshold identification methods may differ for different item types. This study aimed to evaluate three threshold identification methods: normative thresholds 10% and 15%, NT10 and NT15, and the change in information threshold, ChInf. Using data from different item types in the computer-based PISA 2018 Reading assessment threshold identification methods were compared on five item-level and two global validation criteria. ChInf failed to establish unambiguous thresholds for many items of every item type. The NT10 method yielded more favorable validity results on simple and complex multiple-choice items, while the NT15 showed slightly better results on constructed-response items. Valid threshold methods that can be easily implemented in digital assessments may allow for interventions for promoting effortful behavior and enhance understanding of test-taking strategies linked to better performance on computer-based assessments.