Sources within sources: The dynamic infuence of multiple documents trustworthiness when evaluating and using an embedded source

Abstract

The thoughtful use of information sources to assess the trustworthiness of documents has been highlighted as a metacognitive strategy when reading complex or little-known topics on the internet. Prior studies indicate that readers can use document sources (i.e., meta data about the document) to ponder their evaluation of other sources embedded in those documents (i.e., entities referred to in the text as the origin of an idea). This study aimed to expand on those fndings by examining the dynamics of such evaluations and the use of information provided by a single source when embedded in multiple document sources, which can vary (or not) in trustworthiness. Seventy-seven undergraduates, M=23.84 years, read two texts on a health topic to write an essay. Each participant read the texts in one of four sequences that determined the trustworthiness of the document sources and their order of presentation. High-trust document sources were referred to more often in writ ten productions than low-trust sources, and the information they provided was considered better than information in low-trust document sources. Additionally, the trustworthiness evaluation of the embedded source did not change when both document sources presented similar trustworthiness levels. In contrast, when trustworthiness varied across documents, the evaluation of the embedded source fuctuated: it increased when last mentioned in a high-trust document, and it decreased when last mentioned in a low-trust document. These results are interpreted in terms of the complex dynamics of reader-guided processing and, specifcally, source-guided evaluations.