What test to compare males vs females?

Toivo

New Member
#1
Hi.

I have a sample of children with a condition, lets say asthma. so I have n=100, but I have 75 girls and only 25 boys.

Can I use a statistical test to show a significant difference? i.e. That there is more significantly more girls in the sample. If I know the ratio for births is 50:50 for girls and boys in our population, can I say that significantly more girls develop asthma than boys?
 

trinker

ggplot2orBust
#4
I believe so an you should get similar responses:

Code:
prop.test(75, 100)

        1-sample proportions test with continuity correction

data:  75 out of 100, null probability 0.5
X-squared = 24.01, df = 1, p-value = 0.0000009584
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
 0.6516159 0.8288245
sample estimates:
   p 
0.75 


binom.test(75, 100)

        Exact binomial test

data:  75 and 100
number of successes = 75, number of trials = 100, p-value =
0.0000005636
alternative hypothesis: true probability of success is not equal to 0.5
95 percent confidence interval:
 0.6534475 0.8312203
sample estimates:
probability of success 
                  0.75
 

Toivo

New Member
#5
good good. I think I'll side with the binomial test as I can alter the expected proportion size. Don't think I can do that with a chi-square.