Resumen de Self-explanation versus question-answering: differences in processing and effectiveness for learning complex conceptual knowledge
Learning complex conceptual knowledge by reading expository texts (e.g., science texts) is a difficult task because it requires understanding text ideas and generating elaborations from the ideas of the text. To help students with these processes, teachers often ask them to answer questions from the text (question-answering, QA) or to self-explain key text sentences while reading (self-explanation, SE). These two learning activities are inspired by different theoretical approaches to reading. Whereas QA activity inspires by task-oriented reading models because of the use of relevant information for answering the questions, SE inspires by models of reading comprehension, which emphasize the construction of a coherent mental representation of the text. So far, no study has compared the students’ text processing and the effectiveness of these two learning activities, which has theoretical and applied interest in educational settings. Therefore, the main goal of this dissertation is to fill in this gap. Accordingly, we designed three experiments, that recorded the online students’ actions while performing the two learning activities, and then the students’ final learning was measured with a delayed posttest. The students' prior knowledge was also tested to QA and SE.
Study 1 was aimed at examining the variability in the use of text information and inferences in the students’ responses depending on the learning activity; while Study 2 was designed to explore the reading strategies used by students depending on the type of question to be answered (i.e., low-level, high-level) and the type of target sentence to be self-explained (i.e., local, global), as well as to examine the influence of the students’ prior knowledge in the generation of elaborations in QA and SE. Main findings suggested that QA and SE induced different processing, but no differences between these two learning activities were found in effectiveness for learning. Although high-level questions can guide the student to inferential cognitive processes, QA seems to induce a partial and piecemeal text representation as it focuses the student’s attention on the relevant information to answer the questions. However, SE encourages the understanding of the text as a whole depending on the student's mental model and their prior knowledge.
Study 3 included three experimental conditions: answering inserted questions, question-answering after reading the text, and self-explanations. The objectives were similar to Study 2, but focusing on the effect of inserting the questions into the text. Although the processing of text information is closely-defined by the questions posed by the teachers, the results showed that inserting the questions made students focus on relevant information, helped them to generate more accurate elaborations, and consequently contributed to learning more compared to QA and SE. The short delay between reading the relevant information and asking the question facilitates the processes for accessing and retrieving relevant information in the memory resources. Inserted questions also ease the identification and selection of relevant information when students need to search for information in the text. These cognitive processes facilitated by inserting the questions in the text are especially beneficial for low-knowledge students to generate fewer incorrect elaborations when answering high-level questions compared to students in QA.
Overall, this thesis contributes to understanding how the two learning activities and the timing of presenting questions influence the student’s text processing and final learning of complex conceptual knowledge. The studies presented in this thesis have important theoretical and practical implications to design instructional activities and help students learn complex conceptual knowledge.
