| Accounting researchers have studied extensively the influence on investors of numerical information in firms' financial disclosure. In this dissertation, I examine the influence of the non-numerical information---the tone and other stylistic attributes of press releases that accompany announcements of earnings. I use elementary computer-based content analysis to measure the tone and other stylistic aspects of press releases. Tone is measured simply as the number of positive or negative words, scaled by total words. Other stylistic aspects include the overall length of the press release, the overall percentage of numbers versus words, and the complexity of the words used.; To validate the measure of tone, I use a classroom experiment in which fictitious press releases containing different verbal descriptions accompany the same financial data. This experiment confirms that readers are influenced by the tone of earnings press releases and shows the relationship between tone and readers' judgments.; Having confirmed that the tone of press releases can influence the judgment of advanced as well as beginning students, I turn to studying its influence on investors, using actual earnings press releases and archival capital markets data. I use two techniques: a standard event study and a predictive data mining algorithm. Both techniques produce evidence suggesting that tone and other stylistic attributes influence investors' decisions. |
