When you collect any data you need to know what exactly the numbers mean. You need “metadata” (data about data).
For example, you want GDP. Is that real or nominal GDP? At market prices or factor cost? Annual or quarterly – and if quarterly do you want it seasonally adjusted? Is the data for England, England and Wales, Great Britain or the United Kingdom (all four are different)? Note that the official titles of data series are often different from the economic titles: real GDP is usually called “GDP in constant prices” and nominal GDP is usually called “GDP in current prices”.
If data is for a country then how do you deal with boundary changes? Nearly every European country has had a boundary change during the 20th Century (including the UK – the southern part of Ireland left the UK to become the Republic of Ireland in 1921) .
It is very important to make notes of exactly what the data are so that you do not misuse the data in your analysis. Does the definition of the data match the definition of the quantity you are studying?
Another issue in downloading data is to check the date at which the data is downloaded since data (especially recent data) is often revised substantially, especially for series such as GDP.
Finally, it is very important to note any requirements on the use of the data and to acknowledge the source correctly. Remember that some unofficial data sources may be a lifetime’s work for a researcher and to use their data without acknowledgement is plagiarism.