I am trying to determine if results of a survey question for two different departments is significant. The question is "Were you fully satisfied by the service you received today?" Answer is Yes / No.
Department A: 52% of 141 customers were satisfied with their service
Department B: 67% of 233 customers were satisfied with their service
Is the difference statistically significant? (I am using excel and don't have access to the more fancy stats packages.)
Building upon Michael Chernick's comment, if they are independent, you can test this with a standard test comparing two proportions. In R:
prop.test( c(73, 156), c(141, 233) )
2-sample test for equality of proportions with continuity correction
data: c(73, 156) out of c(141, 233)
X-squared = 7.8991, df = 1, p-value = 0.004946
alternative hypothesis: two.sided
95 percent confidence interval:
-0.25971681 -0.04387799
sample estimates:
prop 1 prop 2
0.5177305 0.6695279
The p-value is 0.004946
