The influence of prereading and recall instructions on attention and memory for scientific seductive text.

Abstract

The question of how interesting but irrelevant textual information (i.e., seductive details) in a multi-topic scientific text influences the processes and products of comprehension was explored in three experiments. In Experiment 1, participants read a multi-topic informational text on lightning and tornado formation and rated each sentence for importance and interest. In Experiment 2, participants read the text with or without seductive details and completed a free recall. Participants who received the seductive details version of the text demonstrated a seductive details effect (e.g., Harp & Mayer, 1998; Peshkam, Mensink, Putnam, & Rapp, 2011), in which they recalled significantly less important information compared to controls. In Experiment 3, participants read the seductive details version of the text while wearing a head-mounted eye tracker. Prior to reading the text, participants also received prereading questions designed to focus their attention on one topic (e.g., lightning formation) over the other (e.g., tornado formation). In addition, participants completed a test of their working memory capacity (WMC) in the form of a reading span test (RSPAN), and recieved either a general instruction to freely recall the text or two specific recall instructions for each topic. In terms of online processes, the results indicated that participants allocated additional attention to information introduced by the prereading instructions, in the form of increased reinspections and look-backs. However, attention to seductive details was not reduced by these instructions. In terms of offline products, participants recalled significantly more information from the text that was introduced by the prereading instructions, compared to information that was not mentioned in those instructions. Seductive details were also well recalled for participants receiving free recall instructions. Yet, production of seductive details was significantly reduced when participants received specific recall instructions. In addition, participants with high working memory capacities also produced significantly less seductive content as compared to participants with low working memory capacities. These findings suggest that although seductive details are irresistibly alluring during comprehension, that allure might not be similarly demonstrated upon recall-driven reflection.