The data detective
Want to change the face of medical research? Want to extract nuggets of vital information from, well, information? Love the idea of sifting through data to unlock its mysteries and potential? Then the role of data detective may well be for you. While it sounds like it was invented by writers in search of a gritty, modern protagonist, it's a very real career. So real, in fact, that according to an article in Nature, a data detective, John Carlisle, recently discovered suspicious medical trials and exposed medical misconduct and mistakes. And he isn't even a trained data analyst. It's a career for anyone with a passion for patterns, and with tenacity and curiosity.
This role is ideally suited to people who love the numbers, find the patterns and seek out solutions within the puddles of data that gather below dripping taps of organisational big data.
"The data detective covers the investigative phase before any data cleansing or data preparation takes place," says Martin Rennhackkamp, chief data scientist, PBT Group. "It covers the areas of data accessing and data assessing, and includes data exploration and data analysis."
Data exploration is undertaken to determine what data there is and at what level of detail, as well as what relationships exist between the data. Data analysis investigates the characteristics of the data such as quality, comprehensiveness and completeness. The data detective handles the investigation of the data and the metadata that describes it, sifting through exploration and analysis to get to the nuggets hidden within.
“Using both intuition and a collection of data-sleuthing techniques, they can interpret the data for deeper analysis and predictive modelling,” says Clinton Deavall, founder and chief data scientist, Dinkum Analytics. “To succeed in this role, you need to be an analytical thinker, curious, methodical, and be patient. You also need the technical skills required to dissect, manipulate and transform the data.”
The data detective offers the organisation vital skills for ongoing data management. Data science projects are costly and if the company fails to find hidden biases in the data and spends significant sums on building flawed models, then time and money have been wasted. The data detective is the one who finds the clues that lead to the patterns that help the business bypass this problem. It’s a challenging career, but one that adds immense value to the organisation, especially when it comes to innovation and development.
“The data detective tells the business what they don’t know and challenges their thinking in terms of old assumptions on how the business works and performs,” says Beth McGuinness, data and analytics lead, IQ Business. “As organisations automate more and replace existing business channels with digital channels, they need people who can speak the language of data and make it useful,” she adds.
This role finds what the business doesn’t know, closes the loopholes, unlocks the detail and potentially resolves problems before they start. It’s as demanding as careers go, but it’s also one that will net you a spot in any company in any country thanks to its ubiquity and its relevance.
The mindset of the data detective
What defines the role and the person who fills it?
David Loxton, CEO, Africa Forensics and Cyber: They do more than just process data, they explore and dig into the data. They’re analytical, critical and good at problem-solving.
Martin Rennhackkamp, chief data scientist, PBT Group: A good data detective can make discoveries in existing data that others have failed to find. They can give voice to the data being investigated.
Beth McGuinness, data and analytics lead, IQ Business: They can identify patterns and then convince others that these patterns exist. They are also tenacious, capable of looking until they find the patterns.
Clinton Deavall, founder and chief data scientist, Dinkum Analytics: They approach every problem with no preconceptions about the data or the system that’s generating it. This enables them to uncover clues by asking the right questions.
The clue is in the course
Becoming a data detective is a fairly convoluted process that sees people come from diverse backgrounds with varied skillsets. Some come from business resources and enhance their technical skills, while others focus on the technology and develop their business skills. What they all have in common is maths, science, coding, analysis and data modelling skills. These you can get at:
- Data Science Professional Certificate – Pearson
- Statistical Sciences – University of Cape Town (UCT)
- Data Science, Statistics, Mathematics and Applied Mathematics, Operations Research - UNISA
- SAS Academy for Data Science
- Data Analytics – Damelin
- Business Systems Analysis – UCT
- Statistics and Actuarial Science – University of Stellenbosch
- Big Data Analytics – Wits
This article was originally published in the March 2020 issue of ITWeb Brainstorm magazine.