Recently I read an article where one wise data scientist intoned, "At the end of the most complicated and exhaustive analysis of data, a human being still has to draw an inference and make a decision." And when we reach that point where we have to assess the meaning of the data analysis, our biases come into play.
Many of us tend to trust or rely on data that supports our positions and expectations and suppress data that does the opposite. We also trust data from sources we like or, we rely on data that is the most recent. All of these biases contribute to the challenges and potential for mistakes from our data analyses.
I do not know how to evaluate accurately the data I m receiving and my team is using. What is a good Data Governanace technique?
1. Recognize and mitigate the potential for biases. Seek out data that expands the picture or conflicts with the data in front of you. Encourage an external observer to evaluate your assumptions around data.
2. Strengthen your understanding of data management. There are ample free sources of insights on the web, and many organizations offer seminars or workshops on data analytics and business intelligence. Many universities have added courses for this booming field. Keep sharpening your skills.
3. Ask yourself or your team, "What data do we need to make this decision?" Too often, we rely on the data at hand and ignore the need to seek more data to complete the picture.
4. Be critically aware of the difference between correlation and causation. As described earlier, confusing these two is a potentially dangerous pitfall for decision-making.
5. Quality-check your data. If your firm does not have a data quality or master data management commitment, invest the time to evaluate your data for obvious errors, including duplicate, incomplete or erroneous records. There are many commercially available software applications or to support this activity and many firms draw upon the expertise of data experts to query and assess the data quality. Also, consider external service providers who can help cleanse the data for you. Importantly, focus on continuously improving the quality of your data.
6. Advocate for stronger data quality and management efforts across your firm. This work has often been the domain of IT or technical professionals, yet data has the potential to serve as a strategic asset. Every manager must care about their firm's ability to better leverage data for decision-making and strategy execution.
7. Add technical and data-savvy talent to your team. Sales and marketing departments understand the power of engaging individuals skilled in the latest technologies and competent at navigating many of the data challenges outlined in this article. Technology and data are no longer the domain or responsibility of a single function in an enterprise.