Incremental Theories of Intelligence Predict Multiple Document Comprehension

• In Event: Situating the “Who” in Multiple Source Use: What Do Learners’ Characteristics Reveal About Online Processing?

Abstract

Objectives and Theoretical Frame
The current information age brings about new challenges for multiple document comprehension (Alexander, 2012). Due to the relaxed requirements of publishing, many contend that readers should be more vigilant when interpreting parameters defining a "source" (e.g., the author’s topic knowledge, a publication venue’s interests and biases; Britt, Rouet, & Braasch, 2013). Readers must also integrate information across multiple documents as germane to their research question (Britt & Rouet, 2012). We investigated readers’ beliefs about the nature of intelligence as a relatively stable characteristic or as a malleable quality that one can flexibly develop and change in relation to their multiple document evaluation and integration.

Methods
Fifty-nine Norwegian upper-secondary students completed five individual differences assessments: word recognition skill, working memory capacity, prior topic knowledge, and two implicit theories of intelligence measures (endorsements of entity and incremental theories). Eleven days later, students read six complex, variably-reliable documents about a weather topic. The core scientific concepts were distributed across three documents with reliable source features (e.g., an oceanographer). Three less reliable documents reported on common misconceptions (e.g., an astrologer claims planetary alignments cause El Niños). Then, students produced essays on the topic from memory, rank-ordered the documents, justified their rankings, and validated potential inferences across the reliable documents (IIVT). Four measures were derived: (a) the number of accurate essay concepts, (b) rank-order discrimination, (c) spontaneous trustworthiness evaluations, and (d) the number of inferences accurately validated.

Results
The five individual differences predictors together explained a statistically significant amount of variance in all four outcome measures, Fs(5, 52) > 2.62, ps < .05 (R²s: .20-.38). Among the control variables, prior knowledge uniquely predicted the number of correct essay concepts (β = .57) and trustworthiness assessments (β = .44); working memory predicted ranking discrimination (β = .41) and IIVT performance (β = .35). Most importantly, endorsement of an incremental theory of intelligence was a unique positive predictor for the number of correct essay concepts (β = .41), spontaneous trustworthiness assessments (β = .35), and IIVT performance (β = .43). Ranking discrimination did not reach conventional levels of statistical significance (β = .31, p = .055). Endorsement of entity theories produced no statistically significant results.

Conclusions and Scholarly Significance
Like other work on the role of students’ implicit theories of intelligence in challenging learning contexts (Blackwell Trzesniewski, & Dweck, 2007; Hong et al., 1999; Robins & Pals, 2002), beliefs in intelligence as a malleable quality appeared to promote the use of reading strategies to meet the challenges of comprehending multiple documents. Incremental theorists considered trustworthiness as a basis for document discrimination, and constructed deeper, intertextual understandings of the information found in reliable documents by drawing bridging inferences. Thus, incremental theorists may have sustained their cognitive engagement in the face of uncertain outcomes, if merely to take advantage of the opportunity to gain new, coherent understandings (Dweck, 1999). Future directions include potential longer-term benefits of incremental theories of intelligence when students must accrue knowledge from multiple documents over prolonged periods of time (e.g., semesters).

Authors

• Jason Lawrence Braasch, The University of Memphis

• Ivar Braten, University of Oslo

• Helge I. Stromso, University of Oslo

• Oistein Anmarkrud, University of Oslo