Share
Tweet
Share
Share
Email
Comments
Peter Flom, Independent statistical consultant for researchers in behavioural, social and medical sciences
Excel isn’t undervalued as a tool for statistical analysis. If anything, it’s overvalued as such a tool. I would not trust analysis done in Excel unless I was very sure of the analyst (and the analysts I am sure of don’t use Excel). Why?
- At least the way most people use it, it is point and click. That’s a bad way to do statistical analysis because it makes replication hard and makes error spotting almost impossible.
- It has many fewer options than dedicated programs such as R, SAS, etc.
- It makes it harder than other programs to check assumptions
- While it is possible to make good graphs in Excel, the default graphs are awful and making them good is tricky.
- It is too easy to use. Yeah, too easy. Because a tool that’s easy to use is easy to misuse. Excel for statistics gives people the idea that they can do analysis. Often, they are wrong about that. Doing statistics well requires training.
- As others point out, it can’t handle big data. (This actually doesn’t bother me, I don’t deal with big data).
- It’s too easy, even for an experienced person, to go completely wrong. Cell referencing and row/column referencing is great for spreadsheets, but not for statistics. And you can go completely wrong with no errors, warnings or whatever.
Read Peter Flom's answer to Why is Excel such an undervalued tool for data analysis? on Quora
Related Items:Articles, Big Data, big data analytics
Recommended for you
Click to comment