Data Science and Data Scientists are hot topics. The Harvard Business Review famously called the role of Data Scientist, the ‘sexist job of the 21st century’. Many organizations in fields as diverse as transportation, financial services and health care, are on the hunt for these elusive experts to drive their business forward in the digital and online age.
Unfortunately, many data science projects fail. This is often due to the data scientist and the business having divergent expectations.
So how does one get the best out of a data scientist? Drawing from Tom Davenport’s work, Megan Yates highlighted ten questions one should ask a data scientist. An executive need not have a degree in statistics to pose these questions but they will help to align expectations and outcomes.
- Do you have any data to support that hypothesis?
- Can you tell me something about the source of data you used in your analysis?
- Are you sure that the sample data are representative of the population?
- Are there any outliers in your data distribution? How did they affect the results?
- What assumptions are behind your analysis? Are there conditions that would make your assumptions and your model invalid?
- Can you tell me why you decided on that particular approach to analysis?
- What transformations did you have to do to the data to get your model to fit well?
- Did you consider any other approaches to analyzing the data, and if so why did you reject them?
- How likely do you think it is that the independent variables are actually causing the changes in the dependent variable?
- Are there more analyses that could be done to get a better idea of causality?”
Megan Yates concluded the presentation with a few examples of how data science has been used in Human Resources. Hewlett Packard for example, used machine learning algorithms to predict which of their more than 330,000 employees worldwide were likely to leave the organisation. This insight was incredibly useful because it allowed the organisation to take preventive action – and sometimes before an employee might even have thought of leaving.
LinkedIn saved 15% of their recruiting budget by applying data science to recruitment and successfully predicting hires to within 5% of actuals. This allowed the company to resource the talent acquisition team appropriately and significantly reduced fire fighting.