|
Further, using the same positive and negative word lexicons on both debate wordfiles:the negative and positive words that were used gave rather similar results.
Dem
130/287
Positive .45296167247386759581
157/287
Negative .54703832752613240418
Rep
114/259
Positive .44015444015444015444
145/259
Negative .55984555984555984555
I then tested that same methodology on 2 articles:
The first a NYTIMES article titled"In ISIS Strategy, U.S. Weighs Risk to Civilians"
20/55
Positive .36363636363636363636
35/55
Negative .63636363636363636363
Then a stock report titled "Forget the BBRY Stock Bears; Here’s Why BlackBerry Limited Could Soar"
22/28
Positive .78571428571428571428
6/28
Negative .21428571428571428571
the above test returned the expected results ...
The positive/negative word ratios of those debates were little more than 1% different with more negative words used. Scare politics ...
The time duration length and number of debate participants in each debate would not have influenced the positive/negative word ratios of those debates.
|