The Small Importance of Big Data
Many small things are critical to big data. Yet, as history shows, revolutionary uses will surprise us. Our imagination lags far behind technology’s possibilities. We often see technology as replacers for existing tools and methods. Cells phones would replace landlines and payphones. They are but have also become mobile accessories of the internet and our electronic partners.
Continuing the theme, big data will replace traditional research and problem-solving by providing better and quicker ways to do the same thing in the same way. This won’t tap though much of its potential. We will force big data to work within two of our problem-solving biases rather than cure them:
These biases exist because they’re simpler, not necessarily better, for human problem solving. Combining them, our bias is for one big solution versus many small ones, meaning that we focus so much on locating the 20% increase or savings, that we’ll miss the twenty each producing 1%. “Little Things that Mean a Lot” (The Economist, July 19, 2014 edition) offers many examples of many small solutions adding up big.
Yet, even this oversimplifies because the many small solutions are integrated, either complementing or offsetting one another. “Which?” and “How?” will require us to avoid locking big data into another problem-solving bias, compartmentalizing, defining problems within our understanding rather than within the integrated reality they exist. This means learning to live with unknowns in solutions, another bias.
If big data must work within our problem-solving biases rather than around them, our expectations of it will likely go unfulfilled. This means of course damaged careers and enlightened competitors taking us by surprise. History already knows that’s in some of our futures.
About Post Author
Yes, there are those hindered greatly by the biases listed … Maybe an even greater issue is seeking to solve problems with current capabilities (“let’s see, we have the big data – that’s new; so we can just integrate that data into our current capabilities.”). And, of course, even worse, are those who don’t want anything new – who believe they can deal effectively with problem solving using the textbooks with their included example problems FOREVER!!!
I like very much your attention to the biases. Educators must facilitate the development of effective learning and effective problem solving skills with the same passion and motivation they bring to the learning and application of core knowledge!!! Anything less is a total disservice to the it students!!!
Yes, John, you’re right. Those are certainly additional impediments. I hadn’t considered that people would continue to use sample problems as templates for current problems. I guess that’s also support for people trying to frame new problems into those of past problems without much consideration for their unique aspects.
Thank you for the compliment. The challenge is trying to instruct the existence of these biases. I’ve found that the tendency is very much the same with customer service training. It’s not about me. I’m doing it right. It’s these other guys that need it.
As always, thank you, John.
Excellent observations. I suspect that the biases you mention, Mike, will work in favour of Action Research and against the classic research paradigm. Those countries and institutions that are already investing in participatory processes will benefit. Well, I hope so!
Marilyn, I’m not well versed on Action Research but from what I briefly researched I would say it’s a matter of how the various outcomes are interpreted. For instance, these biases aren’t so much about collecting better or more data but rather the interpretation of those data in drawing conclusions about plans and action steps in those plans. For instance, this post suggests that discoveries in big data might be viewed as too small to make a difference and so overlooked.
A participatory process will help as long as its participants are diverse. It’s less about more participation than it is about diverse participation. A participatory process isn’t going to do much good in regard to these biases if the participants are very homogenous in functions, attributes, experience, skills, connections and personalities.
Thank you again for visiting and commenting. I appreciate the compliment too.