Auditing Section Research Summaries Space

A Database of Auditing Research - Building Bridges with Practice

This is a public Custom Hive  public

research summary

    Patterns of Language Use in Accounting Narratives and Their...
    research summary posted July 20, 2015 by Jennifer M Mueller-Phillips, tagged 09.0 Auditor Judgment, 09.06 Adequacy of Disclosure 
    Patterns of Language Use in Accounting Narratives and Their Impact on Investment-Related Judgments and Decisions.
    Practical Implications:

    This study makes several important contributions. This study extends and complements prior research by looking at an additional characteristic of narrative disclosures not accounted for in prior research: language categories. In examining language categories, the authors contribute to the voluntary disclosure literature by: (1) introducing the LCM as a framework for classifying and organizing the language in accounting narratives; (2) validating that the LCM is indeed descriptive of the language included in accounting narratives; and (3) demonstrating that how prior events are construed in a narrative (using different predicate forms) has a predictable effect on investor judgments and decisions. This study is the first to take a meta-semantic approach to the content of accounting narratives and to apply the LCM in an accounting context.


    Riley, T. J., Semin, G. R., & Yen, A. C. 2014. Patterns of Language Use in Accounting Narratives and Their Impact on Investment-Related Judgments and Decisions. Behavioral Research In Accounting 26 (1): 59-84.

    accounting narratives, earnings press release, investor judgment and decision making, linguistic category model (LCM)
    Purpose of the Study:

    Recent research has examined the role of accounting narratives on investors’ judgments and decisions. This study extends this line of inquiry by examining the effects of language categories on investors’ judgments and decisions—the notion that narratives written with different predicates (verbs versus adjectives/nouns) will have a differential effect on investors. In examining the effects of language categories, the research focus shifts away from semantic manipulations and word choice (e.g., choosing words to convey either an optimistic or pessimistic tone), which can alter a narrative’s overt meaning. Instead, the language category differences examined in the current study can manifest themselves even as a narrative’s overt meaning is held constant, with the use of parallel/similar words in different language categories, or with the use of variations of the same root word (e.g., ‘‘demonstrating our ability to broaden and diversify’’—verbs, versus ‘‘demonstrating our breadth and diversity’’—nouns). Understanding the effects of this more subtle, as yet unexplored, source of variation in accounting narratives is the objective of this study.

    Design/Method/ Approach:

    The authors use a language classification system, the Linguistic Category Model (LCM), to identify linguistic categories that vary on the dimension of abstractness/ concreteness. The sample in this test includes the narrative sections of 553 quarterly earnings press releases of fifty-one randomly selected S&P 500 publicly traded companies, between the years 2002 and 2004. The sample includes only companies whose earnings press releases are available on Lexis-Nexis and published by PR Newswire or BusinessWire. The experiment manipulated linguistic construal and valence resulting in a 2 x 2 between-subjects design to investigate the effect of language categories (contained in an earnings press release) on nonprofessional investors’ investment judgments and decisions. Seventy-nine graduate business students from several major Northeastern universities who had completed their core courses served as participants for this study. Sixty-seven were accounting majors, two were finance majors, and ten were M.B.A. students.


    The first part of this study introduces a psycholinguistic model used in social psychology, the Linguistic Category Model, and proposes it as an organizing framework for the content of accounting narratives. The analysis of the earnings press releases shows that positive information is more likely to be written in a concrete construal than negative information, validating the model’s use in this new context. 

    The second part manipulates orthogonally the construal and valence of the narrative section of an earnings press release and finds that investors are influenced by these construal-related differences. Specifically, investors reading a concretely written negative (positive) narrative are least (most) likely to find value in the firm as an investment, with the judgments of those with abstractly written narratives of either valence in between. 

    The analysis of earnings press releases show a distinct pattern, where the language in press releases is more concrete (abstract) when the associated financial information is positive (negative). These results validate the use of the LCM in this new context.

    Auditor Judgment
    Adequacy of Disclosure