Shouldn’t Excel come with a warning?

How did they manage to embrace big data sets before spreadsheets were invented? I have asked myself that question many times when working on 5MB Excel files. The younger generation does not pose this sort of interrogatives simply because they lack experience of working with voluminous calculations without the use of spreadsheets. It is us, somewhat older lot, that can appreciate the benefits of automated tools against the backdrop of distant memories of working with a pen, a piece of paper and a calculator.

Spreadsheets have evolved significantly since mid-80s when I first came into contact with a 1–2–3 Lotus. I was then staring at a black screen of an IBM Personal Computer (with CGA monitor) populated with amber-coloured numbers. To start with, I struggled to embrace the base concept but once understood, the fundamental premise of a spreadsheet has been logged in my mind as a structural, material and consistent source of beneficial support.

The appreciation of the functionality that a spreadsheet offers has never waned. On the contrary, it grew along with the volume and complexity of data handled and with the subsequent software enhancements. Pull-down menus, functions, graphing capabilities, to name just a few, contributed to this continuing appreciation of spreadsheet as a work tool.

After twenty years of intensive use I started to notice some disturbing aspects of this seemingly faultless tool.

My concerns did not regard the functionality of the tool per se but rather behaviours and attitudes that it promoted amongst its users, not least my own (although I must say that in my experience a critical introspection is far more “costly” and enduring activity than fault spotting when observing your fellow humans).

Roughly about the same time I came across “The Black Swan” a world-wide bestseller by Nassim Nicholas Taleb. The book provided me with a unique experience in that, despite my struggle to understand some of its content, it proved to be so liberating.

One of the concepts that grabbed my imagination in particular manner was that of epistemic arrogance, i.e., the difference between what someone presumes they know and what they know in fact.

This concept resonated with me to me as I could see its applicability to the way we use spreadsheets. I noticed repeatedly how the use of spreadsheets can inflate our believes about what we know, underplay our ignorance and obscure the areas of uncertainty. I shared this thought with the author of “The Black Swan” in brief email proposing that use of a spreadsheet can be, in fact, a particular case of epistemic arrogance as it has a potential to delude its users in that:

• five decimal points precision creates an illusion of correctness
• 12 scenarios considered create an illusion of an in-depth analysis
• choosing the “most likely” scenario creates an illusion of its infallibility

Albeit brief, I got a satisfying reply from Nassim Nicholas Taleb: “Bingo!”.

The accumulating evidence of the health risks of smoking lead to placement of progressively more visible health warnings on cigarette packages. Using spreadsheets without being aware of the potential negative side effects can well be detrimental to our cognitive faculties. Shouldn’t we therefore start placing relevant warnings on spreadsheet packages? Below a couple of proposals.

P.S. One of the features that I appreciate and use very often is Filter. Why have they not introduced filter for columns yet? You must have surely seen those eye-crossing files with data populating columns from A to CD (and beyond) resulting in a scrolling dizziness syndrome :)