Article: How can organizations be good consumers of data


How can organizations be good consumers of data

How can organizations be good consumers of data without falling prey to biases? Subeer Bakshi, Head HR Excellence, Bajaj Finserv, shared his perspective on understanding the impact of biases on decisions and how to avoid them at the Talent Analytics Conference 2018.
How can organizations be good consumers of data

“In God we trust, everyone else must bring Data,” is what can be said of today’s times. In today’s world, an increasing number of organizations are investing majorly in Data and analytics, be it workforce planning, Talent Management, or improving operational performance. Both companies and HR professionals nowadays make use of Analytics to improve the management of their human capital. 

“Intuition and experience combined with quantitative Data can lead to good decisions. Before following this approach however, one needs to understand why relying only on experience in not a wise call.”

However, Data comes with its own challenges of biases and assumptions. In a session titled ‘Understanding Cognitive Biases In Data Analysis’ at the Talent Analytics Conference 2018 hosted by People Matters in Bengaluru, Subeer Bakshi, Head, HR Excellence, Bajaj Finserv, spoke on how to be a good consumer of Data and debunking some of these challenges.

Subeer shed light on understanding the inherent cognitive biases in Data Analysis and its impact on decisions. Today, organizations have reached a point where due to the explosion of Big Data and statistical technologies, everything is backed by Data. If one doesn’t have a Data-backed position, then that organization doesn’t hold a prime position.

Avoiding cognitive biases in Data Analysis

The problem with relying on experience is that human beings, while working as analytical engines, often spot patterns where there aren’t any and miss patterns that are already there. One instance of this is when while hiring, HR people hire people who have supposedly rock star CVs but end up being average performers.

Avoiding cognitive biases in data analysis

While understanding patterns and having the confidence to act on them is a plus point for humans, there are also limitations to this in the form of cognitive biases. For instance, one pitfall is getting anchored to the available information which could be unrelated, being led by it and making a decision based on it.

In the field of HR, how it could manifest is when an organization gets industry benchmarks, such as attrition figures from consultants. What the organization ends up doing is anchoring itself to those figures, even though the circumstances of one’s organization could be very different. Without looking at the demographics or the basis on which the benchmark is based, an organization could subject itself to this anchoring trap.

Similarly, one could suffer from the over-confidence trap stemming from expertise, which could again impact decision making. Then there is the prudence trap wherein by being overcautious we tend to adjust estimates. This could lead to sandbagging of targets resulting in the misallocation of resources.

The solution which Subeer advocates are to combine experience with Analytics or Data-backed decisions.

Challenge key assumptions

Most of the time, Data Analytics is used to confirm one’s biases, organizations should do the opposite as well. One also needs to falsify the information to challenge the bias. This is what is meant by critical thinking or criticizing one’s own basic assumptions. 

For instance, when the HR function is assessing different psychometric tools, say for performance, it also needs to see what else is being eliminated. So, the need to falsify one’s assumptions is the central premise of combining intuition with data in order to arrive at a good decision.

In order to be a good consumer of Data, one needs to describe and challenge every key assumption and the reason for building it. Moreover, one needs to be aware of one’s sensitivities and the business bets. 

Thus, organizations need to find patterns, discern what’s important, judge what’s reasonable and, accept the fact that it is the source of their decision making. However, along with the realization, they also need to challenge some of their assumptions, find deeper patterns, counter overconfidence, falsify what they know and avoid decision traps.

Read full story

Topics: Technology, #TAC2018

Did you find this story helpful?



How do you envision AI transforming your work?