We’re smarter together. Learn from this collection of community knowledge and add your expertise.

Integrity - An essential skill in a SAS User's Toolbox

by Regular Contributor on ‎10-08-2015 05:26 PM (769 Views)

To thine own self be true – Shakespeare

 

Data analysis is one of the greatest passions in my life; being able to take complex data and convert it into a story is absolutely amazing.  Being able to support users, and show them tips and tricks to make them more efficient, is definitely an added bonus.  One aspect of data analysis and data management that’s often neglected in blogs and data science books is integrity.  I realise that this may sound like common sense, but I’ve heard enough stories and been in some situations that could have gone very badly, and without my strong sense of honesty (and the support of management), I could have found myself in a really bad situation.

 Integrity.jpg

 

I’m writing this to give you the best advice I can give – be true to yourself.  This isn’t to say be arrogant, condescending, or ignore the direction given by experts.  Rather, constantly ask questions, learn, read – and in time, that information will amalgamate to allow you to become skilled.  But in that time, you may be tempted to take a shortcut, manipulate data to fit your expectations, or provide an interpretation of the data that may be not quite accurate.  Doing something like this is easy, especially if you’re under pressure to get the results you (or your boss) want.  But take a step back and think what the data represents – someone’s lab results that may be a red flag for a disease, account information that may be the difference between someone losing or keeping their house, credit card data that may mean someone’s financial status is damaged. 

 

Hopefully you have someone in your life that you can go to in times that you have questions.  If not, find one – whether it’s a professor, someone in your workplace, even someone online – as long as it’s someone that you are comfortable with going to with a problem, that can understand the data requirements, and who you trust and will listen to.

 

Good luck, stay true, and don’t use exploding pie charts!

Comments
by Super User
on ‎10-08-2015 06:14 PM

Love this!

If people don't trust you, they won't trust your work! Statistics is difficult for many, so there is a strong requirement on trusting the person preparing the analysis. Integrity in both your work and in your life is important for a successful working relationship :)

by Respected Advisor
on ‎10-08-2015 11:20 PM

Thanks. Food for thought.

 

What I find most difficult is dealing with (statistical) insignificance. When, for example, somebody who trusted me with very precious data has strong beliefs in some hypotheses that, it turns out, can't be verified. This is especially frustrating in exploratory studies where insignificance can be the result of statistical corrections for multiple testing. Maintaining trust, integrity and even friendship in those circumstances can require a lot of work.

Your turn
Sign In!

Want to write an article? Sign in with your profile.