
markowitzdavidm
Author of What Words Are Worth: National Science Foundation Grant Abstracts Indicate Award Funding
Works by markowitzdavidm
Tagged
Common Knowledge
There is no Common Knowledge data for this author yet. You can help.
Members
Reviews
What Words Are Worth: National Science Foundation Grant Abstracts Indicate Award Funding by markowitzdavidm
PDFMA | Highlight Abstract | Can word patterns from grant abstracts predict National Science Foundation (NSF)
funding? In an analysis of over 7.4 million words covering 19,569 proposals, this
article presents evidence that the writing style of NSF grant abstracts corresponds to
the amount of money received for the award. The data describe a clear relationship
between word patterns and funding magnitude: Grant abstracts that are longer than
the average abstract, contain fewer common words, and are show more written with more
verbal certainty receive more money from the NSF (approximately $372 per one word
increase). While such language patterns correspond to award amount, they
largely contradict the NSF’s call to communicate science plainly, suggesting
an inconsistency between the injunctive norms of the NSF and the descriptive norms
of science writing. Broadly, the results support a tradition of research that uses big
text data to evaluate social and psychological dynamics | http://crossmark.crossref.org/dialog/?doi=10.1177/0261927X18824859&domain=pd... | DOI: 10.1177/0261927X18824859 | http://journals.sagepub.com/home/jls |
Contents
1. Abstract pg. 1
2. Language Patterns and Financial Funding pg. 2
3. Predictions pg. 3
-- Discourse and Thinking Style Complexity
-- Confidence in the Science Proposal
4. Method pg. 6
-- Database Descriptive Statistics
-- Automated Text Analysis
-- Language Predictors: Discourse and Thinking Style Complexity
-- Word Count
-- Words per Sentence
-- Common Words
-- Analytic Speech
5. Language Predictors: Confidence in the Science Proposal pg. 8
-- Certainty Terms and Tentativeness
-- Causal Terms
6. Results pg. 8
-- Table 1. Bivariate Correlation Matrix of Primary Dependent and Independent Variables
-- Figure 1. Scatterplots of word count and log-transformed award amount
7. Discourse and Thinking Style Complexity pg. 10
-- Table 2. Mixed Effect Regression Results Predicting NSF Award Funding From Language Patterns.
-- Confidence in the Science Proposal
-- Exploratory Robustness Checks
8. Discussion pg. 13
9. Limitations and Future Research pg. 15
10. Conclusion pg. 16
11. Acknowledgments pg. 16
12. Declaration of Conflicting Interests pg. 16
13. Note pg. 16
14. References pg. 16
15. Author biography pg. 19
SA - https://www.librarything.com/work/32134500/book/262584061 | https://www.librarything.com/work/31507604/book/255801740 | https://www.librarything.com/work/32130292/book/262528448 |
RT - Linguistics
BT - Awards
NT - Correlation
UF - The journal article is about a study that examines the relationship between language patterns in grant abstracts and the amount of funding received from the National Science Foundation (NSF).
SN - PDF downloaded from the website/internet. (This entry does not reference a hierarchical list) show less
funding? In an analysis of over 7.4 million words covering 19,569 proposals, this
article presents evidence that the writing style of NSF grant abstracts corresponds to
the amount of money received for the award. The data describe a clear relationship
between word patterns and funding magnitude: Grant abstracts that are longer than
the average abstract, contain fewer common words, and are show more written with more
verbal certainty receive more money from the NSF (approximately $372 per one word
increase). While such language patterns correspond to award amount, they
largely contradict the NSF’s call to communicate science plainly, suggesting
an inconsistency between the injunctive norms of the NSF and the descriptive norms
of science writing. Broadly, the results support a tradition of research that uses big
text data to evaluate social and psychological dynamics | http://crossmark.crossref.org/dialog/?doi=10.1177/0261927X18824859&domain=pd... | DOI: 10.1177/0261927X18824859 | http://journals.sagepub.com/home/jls |
Contents
1. Abstract pg. 1
2. Language Patterns and Financial Funding pg. 2
3. Predictions pg. 3
-- Discourse and Thinking Style Complexity
-- Confidence in the Science Proposal
4. Method pg. 6
-- Database Descriptive Statistics
-- Automated Text Analysis
-- Language Predictors: Discourse and Thinking Style Complexity
-- Word Count
-- Words per Sentence
-- Common Words
-- Analytic Speech
5. Language Predictors: Confidence in the Science Proposal pg. 8
-- Certainty Terms and Tentativeness
-- Causal Terms
6. Results pg. 8
-- Table 1. Bivariate Correlation Matrix of Primary Dependent and Independent Variables
-- Figure 1. Scatterplots of word count and log-transformed award amount
7. Discourse and Thinking Style Complexity pg. 10
-- Table 2. Mixed Effect Regression Results Predicting NSF Award Funding From Language Patterns.
-- Confidence in the Science Proposal
-- Exploratory Robustness Checks
8. Discussion pg. 13
9. Limitations and Future Research pg. 15
10. Conclusion pg. 16
11. Acknowledgments pg. 16
12. Declaration of Conflicting Interests pg. 16
13. Note pg. 16
14. References pg. 16
15. Author biography pg. 19
SA - https://www.librarything.com/work/32134500/book/262584061 | https://www.librarything.com/work/31507604/book/255801740 | https://www.librarything.com/work/32130292/book/262528448 |
RT - Linguistics
BT - Awards
NT - Correlation
UF - The journal article is about a study that examines the relationship between language patterns in grant abstracts and the amount of funding received from the National Science Foundation (NSF).
SN - PDF downloaded from the website/internet. (This entry does not reference a hierarchical list) show less
Statistics
- Work
- 1
- Member
- 1
- Popularity
- #2,962,639
- Reviews
- 1
