Among bloggers, academics and practitioners of quantitative methods it is standard gossiping procedure to make fun of allegedly dreadful spreadsheet programs like Microsoft excel. Indeed such programs are error-ridden, (too) low-tech, and – a cardinal sin for many quant people – do not require programming skills! This cannot be any good. Instead, quant people preach an intellectual hierarchy going from the apex, programming using something like Python, to advanced statistical packages like R, to other specialized software, with programs like stata or spss residing at the bottom. Spreadsheet programs like excel don’t make the cut; they are the pariahs of the software landscape for quantitative methodologies.
Indeed there are good reasons to mock excel, as, for instance, this Dilbert shows
Yet, I think that mocking excel also is a sign of methods specialists to risk being a sect, a cult looking for its own hermetic truths rather than catering to a wider audience and disseminating its skills to a wide array of people. The downside of such an elitist strategy is that many people are illiterate when it comes to basic spreadsheet techniques like computing an average, handling simple data transformations etc. The abyss between the ‘chosen’ high-tech people, and the ignorant masses becomes even larger. For instance, just to launch a more or less user-friendly ‘surface’ of R such as Deducer one often needs to be able to work around all the problems of Java scripts etc, something that turns most people with limited computer skills even farther away from getting literacy in this area.
I think it is time for the quant people to lose their attitude and teach programs not only in accordance to what they think is best, but also those that are most in use in the private/ non-profit/ or public sector. If they really think that excel is bad on practical or ethical grounds (after all excel costs money), why don’t they create decent and easy to understand online alternatives or freeware programs? The zero- or low-cost alternatives to STATA or SPSS/ Excel are still not convincing, but would be dearly needed. Google, if you have time, put some effort into this! I would happily embrace such a software package when teaching basic stats classes.
Polidigwerkschdaeddle 