3.10 Frequently Asked Questions

My data is already in an Excel list from an online capture system, do I still need to enter the specifications?

Yes, but this should be a quick process. The reason that we need the specifications is to know how the data should look so that we can check for inconsistencies before analysing the data. Also, to know the appropriate ordering for the data. When the data is imported, a variable with levels such as 0-2 years, 3-5 years, 5-10 years and 10+ years can’t be ordered correctly without a specification. You can use the UNIQUE function in Excel to extract all the unique values from a column (ie =UNIQUE(A2:A50) to get the unique values in column A up to row 50).

I have a lot of variables and I don’t want to enter specifications for all of them!

It is essential to provide specification for all the variables you send to Biostatistics, but here are some tips to help: 1. Include only the variables you would like analysed in your DataDictionary. 1. If you still have a lot of variables, it is likely that a number of them will have the same type and range, so you can copy and paste those to save time. 1. If you have a lot of variables with spaces or special characters, place the original variable name in the description field, and give your variables simple names - these can be as simple as q1, q2, etc (you don’t need to type those, Excel will allow you to highlight a few and then drag down automatically renaming to q3, q4 etc).