6_2 : Oppenheimer - Consequences of Erudite Vernacular Flashcards
If you want people to read your writing and think you’re smart, this article suggests that you should
Avoid long words when short ones will do
Q2: The studies suggest at least part of the cause of reduced rating of intelligence is
Difficulty of comprehension
Q3: Which of the following was/were the limitation(s) of the paper you read?
All stimuli were written language. Results may not generalize to oral language
Processing fluency leads to
- higher judgment of truth, confidence, liking
- Someone reading your paper would think you’re more truthful bc you’re using words that are more true and real to your daily life and true self
What are the key concepts in what we read?
- diminish processing fluency with big words
- Word use
What question(s) did the author(s) set out to answer?
- Does increasing the complexity of a text 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?
How did the author(s) go about studying their research question(s)? (study 1)
Study 1
N = 71 Stanford UGS
Stimuli:
- 6 personal statements for grad school
- varied on content and writing quality
Condition (between subjects) : (3 different versions of EACH 6 ps)
- unaltered : original personal statement
- moderately complex: **every third noun, verb, or adjective was replaced
- highly complex: every noun, verb, and adjective replaced with the longest applicable equivalent
DVs
Read 1 essay
- would you accept
- Rate confidence in decision (1-7)
- Rate difficulty to understand (1-7)
Manipulation check (it worked)
- more complex text were more difficult to understand
How did the author(s) go about studying their research question(s)? (study 2)
Study 2
Q’s remaining from Study 1
- Were the replacement words misused? (like joeys rec letter in friends)
- Application essays may have put raters on the alert to attempts by author to look smart
- what if there’s no expectation that the author is particularly intelligent?
Procedure
N = 39 Stanford UG
- Text: 2 translations of Descartes’ Meditation IV
- Same content, comparable length, differing complexity
- More natural stimuli (not concoted by researchers)
- Authorship: Descartes vs. anonymous
- (to test prior experience of intelligence)
- 2X2 factorial design (translation x authorship)
- Between subjects design
- Rate
- Intelligence of author 1-7
- difficulty of understanding 1-7 (manipulation and fluency)
Manipulation
- Complex translation is perceived as more complex
How did the author(s) go about studying their research question(s)? (study 3)
Study 3 - Procedure
N = 85 Stanford UGS
- Dissertation abstract with high proportions of 9-letter words
- Start with someone complex
- Simplified: 9-letter words replaced with shorter version
- Ratings
- Intelligence of author (1-7)
- Difficulty of passage (1-7)
How did the author(s) go about studying their research question(s)? (study 4)
Study 4 - Fluency has yet to be manipulated
N = 51 Stanford UGs
- Unedited essay from study 1 (personal statement from grad school)
- 2 versions of essay:
- Times new roman - 12 point font
- Juice it 12 point font
- Font was attributed to experimenter, not author
- Rate authors intelligence (1-7)
What did they find (study 1)
In 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
- “Admission confidence” = Confidence rating (1 - 7) x (+1 - accept) or (-1 reject)
- Yield score of -7 to +7
- Original > moderate
- Moderate > high
- Higher complexity → Lower Acceptance
- Higher complexity → Higher difficulty of comprehension
- Higher difficulty of comprehension → Lower acceptance
- Lower complexity → Higher Acceptance
- Lower complexity → Lower difficulty of comprehension
- Lower difficulty of comprehension → Higher acceptance
P < .5 so, R is significant
- Along the bottom path (LoC → AR), P>.10 which means that it is not significant. So a high level of complexity will not be significant in getting a high acceptance rate
- When controlling for difficulty of comprehension, the relationship between complexity and acceptance was reduced
What did they find (study 2)
Study 2: Intelligence Rating
summary
- Complexity negatively influences rater’s assessment, regardless of expectations of the authors intelligence
- The process is mediated by fluency
- NO interaction
- There was Main effects in
- Complexity
- expectation
- Higher complexity → Lower intelligence rating
- Higher complexity → Higher difficulty of comprehension
- Higher difficulty of comprehension → Lower intelligence rating
- Lower complexity → Higher intelligence rating
- Lower complexity → Lower difficulty of comprehension
- Lower difficulty of comprehension → Higher intelligence rating
When controlling for the difficulty of comprehension(path on top) , the relationship between complexity and intelligence was reduced
- High complexity is lower intelligence rating
P > .05 not significant
What did they find (study 3)
- Ss who read the simplified version rated the author as more intelligent than those who read the original
- Level of difficulty partially mediated the relationship between complexity and intelligence (testing fluency hypothesis)
- They improved grad students’ dissertation abstracts
What did they find (study 4)
- The author of Juice it 12 point font version was rated as less intelligent than the author of TNR version
- i.e. “non-fluent” less intelligent than “fluent”
RESULTS OVERALL ACROSS ALL 4 STUDIES
- Needless complexity leads to negative evaluations
- Across domains (personal statements, philosophical essays, dissertations)
- Across types of judgments (acceptance decisions, intelligence)
- Across paradigms (Higher word replacement, Lower translations)
- Regardless of the quality of the original (multiple personal statements)
- Regardless of expectations of a text’s quality (authorship)
- ^^^robust effect^^^
- Effect is at least partially due to lowered processing fluency
