# University project help (correlation)

#### kristian9210

##### New Member
Hello dear stat-folk!

I'm currently in the process of re-writing my project due to a failed exam in it.
In this regard i'm having trouble figuring out whether i can even use statistics (and how) to describe a theory i have towards a possible correlation.

In this part of the project i am investigating whether i can state that there is a correlation and to what degree between the quantity of product A,B,C,E.. etc produced, and the amount of errors that the products contain. (the generel theory is that, the more often operators perform manual laborer on the same type of product, the better they become at producing them with no errors) the industry in which i'm researching is EMS (Electronic manufacturing services), my case company produces PCB's (plastic circuit boards).

My initial strategy towards this, have been listing every type of product and the quantity in which they have been produced throughout a period, thereafter listing the amount of errors found for each type of product. Hereafter i have converted the amount of errors into a percentage out of the total produced for each type of product. And then taken the correlation with the CORREL command in excel.

This may have been formulated a bit rough, also english is not my native language, as people may have recognized by now. But to give a better idea of what i have done ill post a link from google sheets: https://docs.google.com/spreadsheets/d/1miaMC539lsFm0s3aP-MXAU8_NGwbBorOFXmlXoh9kLo/edit#gid=0

My question is, is this a valid way of investigating the correlation im trying to prove, and how to i go about estimating the P-value of this statistic.. etc.

My concern is that the statistic is invalid due to the difference in sub group size?
And what estimator for P i should use, significance levels etc..

Im confused.. i hope someone could shed light on my problem.

Best Regards