## Python, Statistics, Educational Measurement: Cohens kappa interrater agreement coefficient

Cohen's kappa coefficient is a measure of interrater agreement for qualitative or categorical variables. It is computed normally only for two raters or judges . Input data is usually a 2 by 2 table. Consider the following:

Yes No
Yes a b
No c d

and a specific example(from en.wikipedial):

 20 5 10 15

The following Python program computes Cohen's kappa

 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  def kappa(a,b,c,d): tot = a + b + c + d Pa = float(a + d)/tot PA1 = float(a + b)/tot PA2 = 1.0- PA1 PB1 = float(a + c) /tot PB2 = 1.0 -PB1 Pe = PA1 *PB1 + PA2*PB2 print Pa, PA1, PB1, PA2, PB2 return (Pa -Pe)/ (1.0 -Pe)   def Test(): a = 20 b = 5 c = 10 d = 15 print "kappa(20,5,10,15)=", kappa(a,b,c,d)   if __name__ == "__main__": Test()