Conveying Intelligence psych and edu

IG Nobel Award 

• Ignoble definition: not honorary 

• “First make people laugh, then make them think” 

• Ceremony begins with a paper airlane deluge

Why might you think to do those things? 

• Intelligence–vocab size 

• If the writing is complex, the ideas expressed must be complex 

Evidence of the Opposite 

• Processing fluency: the ease with which information is processed 

• Simpler writing is easier to process 

• An example is right in the title: 

  • Consequences of erudite vernacular utilized irrespective of necessity 

  • Problems with using long words needlessly 

• Ease in processing 

  • Higher judgment of truth 

  • Higher judgment of confidence 

  • Higher judgment of liking 

• 5 studies–we’ll look at the first 4 

• Loss of fluency→ negative impact on ratings of the author 

  • When due to needless complexity

Research Questions 

• Does increasing the complexity of the test succeed in making the author appear more intelligent? 

• To what extent does the success of this strategy depend on the quality of the original, simpler writing? 

• If the strategy is unsuccessful, is the failure of the strategy due to loss of fluency?

Study 1–Procedure 

• N = 71 Stanford UGs 

• Stimuli: 

  • 6 personal statements for grad school 

• Varied in content and writing quality 

Conditions (between subjects) 

• Highly complex version: every noun, verb, and adjective replaced with the longest applicable equivalent 

• Moderately complex version: every third noun, verb, or adjective was replaced 

• Unaltered version

Study 1–DVs 

• Read 1 essay

  • Accept 

  • Rate confidence in decision (1-7)

    • Combined the first 2 

  • Rate difficulty to understand (1-7)

Study 1–Manipulation Check

More complex texts were more difficult to understand

Study 1–Results

• “Admission confidence” = confidence rating (1 to 7) + 1 (accept) or -1 (reject)

• Yields a score between -7 and +7

• Original had the highest score 

• Moderate had the second-highest 

• High had the lowest

• When controlling the difficulty of comprehension, the relationship between complexity and acceptance was reduced 

• Mediation 

  • Mediator = difficulty of comprehension 

  • Direct effect = the level of complexity’s impact on acceptance rating

    • Significant relationship between the level of complexity and the acceptance rating  

  • The lower the level of complexity, the higher the acceptance rating and vice versa 

Study 1–Summary 

• Simple texts were given higher ratings than complex texts, regardless of the quality of the original 

• Complex texts are difficult to read, which leads to lower ratings

Study 1–Remaining Questions 

• Were the replacement wors misused? 

• Application essays may have put raters on the alert to attempts by the author to look smart 

• What if there’s no expectation that the author is particularly intelligent? 

Study 2–Procedure 

• N = 39 Stanford UGs 

• Text: 2 translations from Latin of Descartes’ Meditation IV

  • Same content, comparable length, differing complexity 

  • More natural stimuli (not concocted by researchers)

• Authorship: Descartes vs. anonymous 

  • Tests the expectation of intelligence 

  • Descartes was expected to be intelligent 

• 2 x 2 factorial design (translation x authorship)

• Between-subjects design

• Rate 

  • Intelligence of the author (1-7) 

  • Difficulty to understand (1-7) (manipulation check and fluency) 

Study 2–Manipulation Check 

• Complex translation was complex 

Study 2–Results: Intelligence Rating

• No interaction 

• Main effects: 

  • Complexity 

  • Expectation 

• Simple translation made the author look more intelligent 

• When people knew it was Descartes, they rated the author smarter 

• Mediation 

  • Direct effect = The more complex, the lower the intelligence rating 

  • When controlling for difficulty of comprehension, the relationship between complexity and intelligence was reduced 

  • Significant relationship for the lower the complexity, the lower the difficulty of comprehension, the higher the intelligence rating

    • Vice versa 

Study 2–Summary 

• If you expected an author to be intelligent, you rated them as more intelligent than if you had no expectation 

• Complexity negatively influences raters’ assessments, regardless of expectations of the author’s intelligence

• The process is mediated by fluency 

Study 3–Procedure 

• N = 85 Stanford UGs 

• Dissertation abstract with a high proportion of 9-letter words 

  • Start with something complex 

• Simplified: 9-letter words replaced with shorter versions 

• Ratings 

  • Intelligence of the author (1-7)

  • Difficulty of passage (1-7) 

Study 3–Results 

• Students who read the simplified version rated the author as more intelligent than those who read the original 

• Level of difficulty partially mediates the relationship between complexity and intelligence 

  • Test fluency hypothesis 

  • They improved grad students’ dissertation abstracts

Study 3–Remaining Question 

• Fluency has yet to be directly manipulated 

Study 4–Procedure 

• N = 51 Stanford UGs 

• Unedited essay from study 1 (personal statement for grad school 

• 2 versions of the essay: 

  • Times New Roman 12 point font 

  • Juice ITC 12 point font 

  • Font was attributed to the experimenter, not the author

• Rate the author’s intelligence (1-7)

Study 4–Results 

• The author of the Juice ITC 12-point font version was rated as less intelligent than the author of the Times New Roman version 

  • “Non-fluent” is less intelligent than “fluent” 

In Sum

• Needless complexity leads to negative outcomes 

  • Across domains (personal statements, etc.) 

  • Across types of paradigms (word replacements)

  • Across types of judgments 

  • Regardless of the quality of the original 

  • Regardless of expectations of a text’s quality 

• Robust Effect 

• The effect is at least partially due to lowered processing fluency

Limitations 

• Generalizability from Stanford students? 

• Generalizability to oral language? 

• Long words are sometimes more appropriate than shorter versions (studied needless complexity) 

• Jargon can display in-group status