All the while it’s been Excel can do this, and Excel can do that. But what are its pitfalls? Where does it fail? What should we beware?
Surprisingly, there’s a whole community that’s at this topic! And I only recently stumbled upon them. That while trying to learn the new software “R”.
Here’s a link to Excel’s shortfalls, something that only authentic Statistical softwares are capable of.
I’d like to take the time to highlight some of these below –
1. Excel’s greatest boon – the spreadsheet page – is it’s greatest bane! Formulae and Values in the same cell, with no line of demarcation between the functions and the end results. Some operations such as sorting result in erroneous output.
2. Data Validation is a big problem
3. Computational complexity and inconsistencies, not to forget the time.
4. Bad graphing processes (I always thought Excel graphs were good!)
5. Excel Statistics is anything but Statistics! The probabilities, accuracy, plots, regression etc.. nothing matches with Statistical Software packages. (RAND function is potentially a great bug)
6. Solver is not something that can be relied upon, convergence issues.
7. No “HISTORY” to track the list of data actions/manipulations for ease of debugging or other reasons.
And the list goes on.. I’ve only scratched the surface I believe, but these are pretty much the high level accusations.
And now, what I’d like to take away from all this –
I know and understand that every software has its set of pluses and limitations, and Excel is no exception. It’s Microsoft, it’s unpredictable, it’s got bugs, it’s not perfect.. you can go on..
But it’s still the most widely used and understood piece of software, easy to use since we already know it so well, and works wonders on quick visual inferences. And Excel is powerful enough usually to do stuff that is beyond the average tasks, and I’ve myself used it for purposes that are beyond what I thought possible.
IF anything, it’s a great software, and currently, the easiest at handling data, charting and computation. And as long as we don’t rely on using hardcore Statistics using Excel’s built-in tools, I don’t see any harm why we can’t boast of Excel’s functionality in other domains! Also, most stuff above about Excel’s limitations on data complexity or computational complexity is something we already know, and here’s the best part – once used to Excel, we end up working around it always. ALWAYS! We just know where it hits the limit. It’s like having this old piece of machine and knowing how to use it. Agreed there are better tools out there, but currently, this one rules the majority of the population and until a time when the others take over, we can still boast of Excel’s capabilities and keep learning newer stuff to boost our skills and make our everyday life easier…