Reconceptualizing metacomprehension calibration accuracy

Accurate judgment of text comprehension is compulsory for learners to effectively self-regulate learning from text. Unfortunately, until relatively recently the literature on text comprehension judgment, termed metacomprehension, has shown learners to be inaccurate in their judgments. Over the last decade researchers have discovered that when learners use delayed summaries of text to make judgments metacomprehension accuracy increases. In contrast, when learners use individual differences (e.g., knowledge and interest) to make judgments they are less accurate. Traditionally metacomprehension accuracy has been construed as the average correlation of judgments and comprehension assessments across multiple texts. In the current study multiple alterations to the delayed summarization paradigm were evaluated. Specifically, the difference between learners’ comprehension assessments and assessment scores were calculated within text to assess text specific differences in judgment cue use and accuracy. Second, pre-reading prompts were provided to focus learners on connections within the text (graphic organizer) and specific factual information (a list of facts). Third, the relative influence of individual differences (i.e., interest, knowledge, GPA, understanding of university text and Need-For-Cognition), text, pre-reading prompts, and summary delay on comprehension judgments were evaluated. Finally, experimental influences on judgment/assessment score differentials were considered. Results indicated that the two obscure texts in this experiment, with similar structures, and different levels of interest resulted in statistically detectably different judgments, scores, and accuracy. Delayed summarization did not improve metacomprehension accuracy for either text. This is a departure from the current literature, and may indicate that within-text measures of metacomprehension accuracy react differently to delayed summarization. Individual differences also affected judgments differently across texts and groups. Finally, participants used different cues to make judgments at the detailed and explanation levels of understanding