0

I have this data set where I calculate the percentage of fall given an age. Total Members is the whole date by a given age, and members fallen is the amount of falls within the total member. 

age status membersFallen totalMember percentage
37      1      208       14824        1.4
38      1      443       14736        3.0
39      1      610       13190        4.6
40      1      563       12209        4.6
41      1      539       11186        4.8
42      1      528       11141        4.7
43      1      546       10118        5.4
44      1      527        9677        5.4
45      1      560        9399        6.0

This data generates a crescent graph, and what I'm willing to do is to calculate the uncertainty of this percentages and plot an error bar in scatter plot.

Thanks

Pedro Martins
  • 110
  • 2
  • 8
  • 1
    What plotting code did you use so far? You can look at `prop.test` for one way to generate confidence intervals for proportions. You should be explicit about what modeling assumption you want to use to generate confidence intervals. – MrFlick Jun 03 '15 at 20:19
  • Welcome to SO! You may get an answer faster if you share with us what you already tried with a [reproducible](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example?s=1|7.0421) example. – akhmed Jun 03 '15 at 20:41

1 Answers1

0

Are you thinking of the confidence interval for binomial proportion? see http://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval E.g. binom.test(208,14824) the 95% CI is [1.22, 1.61] for your sample estimate 1.40 (percentage)

fishtank
  • 3,718
  • 1
  • 14
  • 16
  • Thank you! It helped me a lot. My Solution in R was use **prop.test(k, n)** (http://www.r-tutor.com/elementary-statistics/interval-estimation/interval-estimate-population-proportion) command to calculate the ci. – Pedro Martins Jun 04 '15 at 17:40