Table 2: 
Debate-level features used in estimating skill levels. Aside from Elo, the features are a part of the user’s linguistic profile. The third column represents statistical significance levels in comparing winners and losers’ features (independently) with Bonferroni correction: ↑ is p < 0.05, ↑↑ is p < 0.01, ↑↑↑ is p < 0.001.
FeatureDescription
Elo Score Traditional Elo score calculated and updated. Updated traditionally, not averaged as in §4.3. n/a 
 
Length Number of words this user uttered in the debate. ↑↑↑ 
 
Part of speech Count of each noun, verb, adjective, preposition, adverb, or pronoun from the participant in the entire debate. Nouns (↑↑↑) Adjectives (↑↑↑) 
 
Flesch reading ease Measure of readability given the number of sentences in a document and the number of words in each sentence Kincaid et al. (1975). ↑↑ 
 
Emotional words Cues that indicate a positive or negative emotion (Tausczik and Pennebaker, 2010). Pos (↑↑↑) Neg (↑↑↑) 
 
Links Links to external websites outside of Debate.org. This feature operationalizes the number of sources a debater used.  
 
Questions The number of questions the user asked in the debate. ↓↓↓ 
 
Quotations The number of quotations the user included in the debate.  
 
Hedging The number of phrases that soften a statement by adding uncertainty (Hyland, 1996; Hanauer et al., 2012).  
 
Fightin’ words The number of instances of words most strongly associated with either debater (Monroe et al., 2008). ↑↑↑ 
 
H∧FW The number of cooccurences of hedging and fightin’ words, described in §4.2. ↑↑ 
FeatureDescription
Elo Score Traditional Elo score calculated and updated. Updated traditionally, not averaged as in §4.3. n/a 
 
Length Number of words this user uttered in the debate. ↑↑↑ 
 
Part of speech Count of each noun, verb, adjective, preposition, adverb, or pronoun from the participant in the entire debate. Nouns (↑↑↑) Adjectives (↑↑↑) 
 
Flesch reading ease Measure of readability given the number of sentences in a document and the number of words in each sentence Kincaid et al. (1975). ↑↑ 
 
Emotional words Cues that indicate a positive or negative emotion (Tausczik and Pennebaker, 2010). Pos (↑↑↑) Neg (↑↑↑) 
 
Links Links to external websites outside of Debate.org. This feature operationalizes the number of sources a debater used.  
 
Questions The number of questions the user asked in the debate. ↓↓↓ 
 
Quotations The number of quotations the user included in the debate.  
 
Hedging The number of phrases that soften a statement by adding uncertainty (Hyland, 1996; Hanauer et al., 2012).  
 
Fightin’ words The number of instances of words most strongly associated with either debater (Monroe et al., 2008). ↑↑↑ 
 
H∧FW The number of cooccurences of hedging and fightin’ words, described in §4.2. ↑↑ 
Close Modal

or Create an Account

Close Modal
Close Modal