Why non-tech people make data better

On the home page of DataCharrette.org, I mention that our event is unusual in that we specifically designed it to welcome non-technical people in addition to data specialists. There are several reasons for this, but it relates to my favorite explanation of what data scientists do.

In response to the Quora question “What does a data scientist do? Can you become one without being hired as one?” Steve Andrews give this excellent description:

  1. Clean data.
  2. Clean data.
  3. Clean data.
  4. Clean data.
  5. Clean data.
  6. Clean data.
  7. Clean data.
  8. Clean data.
  9. Clean data.
  10. Clean data.
  11. Clean data.
  12. Clean data.
  13. Do some math.
  14. Try to get everyone to understand my findings.
  15. Try to get everyone to understand my findings.
  16. Try to get everyone to understand my findings.
  17. Try to get everyone to understand my findings.
  18. Try to get everyone to understand my findings.
  19. Try to get everyone to understand my findings.
  20. Try to get everyone to understand my findings.
  21. Try to get everyone to understand my findings.
  22. Try to get everyone to understand my findings.
  23. Try to get everyone to understand my findings.
  24. Try to get everyone to understand my findings.
  25. Try to get everyone to understand my findings.
  26. Repeat.

The point of this is that the technical part of data science, in many cases, constitutes only a small fraction of the work involved. Steps 1-12 (“clean data”) can be done by people with even limited technical training (and, at the charrette, it’s usually done in Google Sheets). Step 13, “do some math,” requires technical training. Steps 14-25 (“Try to get everyone to understand my findings”) can often be done marvelously by non-technical people, once they have consulted with the person who did step 13. In fact, it’s often done better by the non-tech people. But, regardless, it makes it clear that there’s room for everyone in the data world, and certainly at the Data Charrette.