Follow us on Twitter      |   Join us on    

Who’s Interpreting the Data?

A common refrain from almost every government CIO I’ve interviewed is how excited they are about Big Data and analytics. Sieving through large, messy data stores to find little nuggets of useful information to help in public policy is what is driving the passion around the technology. But when you ask for a progress report, quiet reflection replaces bright enthusiastic smiles as CIOs ponder daunting human resource challenges.

You see, Big Data as a technology phenomenon is really maturing fast. There are many tools, both open source and proprietary, to address almost any problem government can throw at it. And yes, there are amazing analytic tools to generate impressive charts and dashboards. So what’s the problem? Finding people who can glean the insights from the analysis.

To be sure, this is not a problem unique to the public sector. Private enterprise has grappled with this since business intelligence and analytics gained prominence the past decade. But where the private sector has established entire departments devoted to analysis, the public sector has had no such luxury, resources, or some would say, will to resolve this.

Anecdotally, many in public sector view big data analytics as content, not insight. This is a mistake. Great charts and dashboards are a great source of fresh content in and of themselves, but they don’t help in policy making unless someone knows how to interpret the analysis. Changing the mindset and recognising the primacy of analysis as a tool for policy making will require personnel who can ‘read between the data lines’. The public sector needs employees with additional skills to thrive in a data-driven governance model to benefit from Big Data analysis. So what skills are critical?

Going beyond the numbers
Being numerate and comfortable with reams of data is increasingly a critical skill in the public sector. Just look at how many statistical reports governments are routinely generating and you understand that extent of the problem. You don’t have to hire statisticians in your department but public sector managers who have to deal with the reports must be adept at interpreting data, metrics and results.

The private sector’s response to this problem has been to create jobs for data scientists. These are individuals who ask the right questions because they understand the unique requirements of their companies. I see a future for such jobs in the public sector very soon.

Experimentation
Increasingly, I see the public sector adopting and embedding a spirit of experimentation in their job scope for their employees. If Big Data analytics is in your department’s future, you need to train your staff in the principles of data analysis and then let them experiment. This will facilitate a real-world understanding of testing and design, sampling selection, etc., and how to construct intelligent hypotheses. If this is done right, staff will feel empowered to test their hypothesis by setting up random testing models so the agency can evaluate the validity of the data analysis. The more experimentation, the better the veracity of policy decisions because you know the data has been validated.

Deploying Big Data analytics is about fostering a culture of analysis and focus on data as it is about adopting new technology. The future of governance demands that we build a public sector workforce that has data analysis and interpretation in their DNA. An experiment-focused, numerate, data-literate workforce serving the public is the way forward. What do you think?

Add your comment


Most highly rated

Better learning with web 2.0 and virtual worlds

In a visit to Ngee Ann Secondary School yesterday (22 July), FutureGov found students deeply ...

Artificial Intelligence Tools to be Used in Singapore Schools

The Infocomm Development Authority and Ministry of Education of Singapore have initiated plans to introduce ...

Students take a green stance with social media

Ngee Ann Secondary School’s students are on a bid to “change the world” with ...