Pearson's correlation coefficient

#1
Hello everyone!
I'm fairly new to statistics and struggle a lot with it, so I could really use your help please!
As part of a project I'm doing, I want to see if there is any relation between multifetal pregnancies (twin,triplet,etc) and caesarean sections. I need to establish this relationship with data from a hospital in a city and then with data from a whole country. So, my variables are "number of multifetal pregnancies in said city" - "number of caesarean sections in said city" and "number of multifetal pregnancies in said country" - "number of caesarean sections in said country". I have anual figures for these 4 variables, for a total of 21 years (for example, in the city: year 2010 - 254 multifetal pregnancies and 1619 C-sections; year 2011 - 191 MF pregnancies and 1513 C-sections; year 2012 - 173 MF pregnancies and 1493 C-sections;...). My question is, can I use Pearson's correlation coefficient (r) to determine the relationship between the variables "number of multifetal pregnancies" and "number of caesarean sections"? If so, what is my n?? Is it the total number of years for which I have data (21 years)?

I would appreciate any help with this so much!!
Thank you :)
 

staassis

Active Member
#2
Your n is 21... Plot the two variables and see if the relationship looks linear or curve-like (non-linear). If linear, calculate Pearson correlation and make sure to generate the corresponding p-value (based on the t-test). If the relationship is non-linear, calculate Spearman's rho and / or Kendall's tau, with appropriate p-values. Without a p-value, you do not know if the true value of the association measure in question is 0 or not. This is important. If the association measure is 0, there is no relationship.