How can you become better at your work? You know the old standbys like attending a class, taking on a challenging project, and networking with new people who you can learn from. Just don’t stop there.
Go a little further than your competition by trying diverse ways of learning and reinforcing what you know. Try these nine ways to help you keep your resolution to become a better data analyst:
Find the human experience in your data. Choose a sample dataset or report from your own work. Ask yourself how the data relates to actions and results for an individual person. For example, does the data represent purchases, store or website visits, medical examinations? What happened to the people represented by the data? Retell what you see in the data, from the viewpoint of one person it represents, in no more than 3 or 4 plain English sentences – no jargon, no statistics.
Do your friends, relatives and neighbors understand what you do? No? Focus on the benefits – the people who are better off because of your work and the reasons why. Refine your story until you can explain, in no more than 30 seconds, what’s valuable about your work, to anyone.
Put it on paper. Write a brief summary, using complete sentences and paragraphs, of your latest project. Review it, and ask yourself whether a person who is educated, but not familiar with analytics, would understand the results and their implications. Revise. Get an outsider to review and provide feedback. Repeat.
Return to the beginning. Read an introductory book on statistics. Choose one that you have not read before.
The best analysts know the basics inside out.
Calculate by hand. Using a small sample of data, fit a simple model (such as a linear regression) or do a hypothesis test – by hand.
Concentrate on assumptions. Know the assumptions behind each technique. Look them up, one by one. Even working professionals are often misinformed about what is and is not assumed in individual testing and modeling methods.
Try the unfamiliar. Playing with new toys makes you more adaptable (and adds keywords to your resume.) Stretch yourself to use something you don’t like. If you usually prefer a commercial tool, try something open source, or vice-versa. Spend some time working with bigger, or smaller, data than usual. Learn a new analytic method.
If you really know it you can explain it. Find out if you know your stuff well enough to teach someone else to do what you do.
Emphasize process. The right process ensures maximum value to your organization and minimizes rework. Consider using a detailed analytics process model like CRISP-DM.