# Why isn't this code for prop.test giving me the right answer for test statistics?

I have used the following line of code but it doesn't give me the right answer for (chi-squared) test statistic while it gives the right answer for p-value regarding the following question: "Consider testing H_0: p = 0.5 against H_1: p ≠ 0.5, given that there were 38 successes in a sample of size 56. Find the value of the (chi-squared) test statistic. (Do not use the continuity correction.)"

> prop.test(38,n=56,p=1/2)
1-sample proportions test with continuity correction

data:  38 out of 56, null probability 1/2
X-squared = 6.4464, df = 1, p-value = 0.01112
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
0.5390880 0.7935168
sample estimates:
p
0.6785714


Can you give some explanation in your answers? I am not coming from a statistics background.

• What does the last line of the question say? What does the first line of your output say? What does ?prop.test say about continuity corrections? – Glen_b -Reinstate Monica Feb 14 '14 at 4:10
• If this is a homework question you should apply the self-study tag. – dmartin Feb 14 '14 at 4:25

The problem was that I was using proportions test with continuity correction while in the question it was mentioned to use proportions test without continuity correction. So we just need to add correct=FALSE parameter to prop.test function.

> prop.test(38,n=56,p=1/2,correct=FALSE)

1-sample proportions test without continuity correction

data:  38 out of 56, null probability 1/2
X-squared = 7.1429, df = 1, p-value = 0.007526
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
0.5482264 0.7859901
sample estimates:
p
0.6785714