Which right test to choose

I am new in statistic and analysis, I have a question

I wish to compare is the data from my Centre significantly difference with the data from the published paper. the number stated is the percentage of occurence of the disease in my and published center data. My data is a non-normality data. Is there any possible way for me to test whether is there any difference between my data and the published data?

WIth thanks.

• Do you have the actual frequencies which you could present in the text as your image is not very readable? – mdewey Jul 21 '16 at 12:45

Comparing percentages without knowing the size of the Groups, from which the percentages were calculated, usually does not work. However, if the number of People is given, from which the percentages were calculated, than you could compute abolute numbers. Absolute numbers of patients can be compared using chi-square tests. So if one publication hat 2 patients out of 1000 and you have 828 patients out of 2000, than you can build a contingency table and perform an Chi-square-test

                mine  their's
disease A        828      2
not-disease A   1172    998


In R this would look like this:

> chisq.test(matrix(c(828,1172,2,998), nrow=2))

Pearson's Chi-squared test with Yates' continuity correction

data:  matrix(c(828, 1172, 2, 998), nrow = 2)
X-squared = 563.41, df = 1, p-value < 2.2e-16


And, of course, with a p as small as that, it's highly significant.