Subscribe to DSC Newsletter

What most people call “analysis,” I refer to this as “guidance.”  It is not guidance in terms of guiding the company; but rather, I provide a narrative to help guide people through the data - of which there is a great deal.  I play the role of a tour guide.  I remember when I was a teaching assistant for a social science class - and there was a contentious area that would likely be the focal point for essays - I said that it didn’t matter to me what “opinions” people expressed.  Nobody had to worry about my personal sentiments affecting their grades.  However, absolutely each student had to support their assertions using proper references.  Writing the paper was an exercise in making a point and substantiating it - hopefully in a manner that I could understand and grade.  I operate a bit in reverse with my guidance or analysis of data.  I examine what the body of data seems to say, and then I try to make a point.

 

It is surprising at times how different one’s personal conclusions might be from what the data has to say.  I would even suggest that it is dangerous not to use data - i.e. to use sentiments or instincts.  Even if a person uses data, it is necessary to ensure that the data supports the points being made.  Quite frankly, untrained individuals will tend to draw faulty conclusions from data.  In Ontario, there is an expectation that employers will not simply terminate an employee on performance issues: there should be some evidence that the employer spoke to the employee not once but several times in an attempt to resolve problems.  It is for this reason that I believe workflows tend to be restructured - creating surplus positions - opening the path for terminations perhaps due to absence of work.  I believe that many companies keep poor records relating to the day-to-day operational performance of employees.

 

However, an organization that fails to keep operational performance data at an employee level likely lacks the intellectual capital or capacity to improve the performance of departments and teams.  The company will know when “things” are going well or poorly but not how to get all of that complicated “stuff” to improve.  I think specifically of the human resources department: it probably doesn’t have much working knowledge of operations.  It accepts the narrative as delivered probably without much evidence.  There might certainly be strong opinions - but not compelling data.  Imagine a critic making a statement like, “That worker is lazy - just sitting around - not doing the job.”  However, the data might say something entirely different; this leads one to conclude that the critic has no idea.  I am not saying that this person doesn't “appear” to have any idea.  There is quantitative evidence that he or she has no idea.  I call this phenomenon the “Little Hitler Syndrome.”

 

I had no difficulty giving essays lousy marks if students failed to provide proper references.  However, that was in an academic setting.  It is important to recognize that the mindset persists or thrives among the vernacular in everyday life.  I have discovered that the average person even if they have some level of post-secondary education will sometimes descend to their tribal instincts.  I get it, really.  A big reason I write blogs is for the freedom to exclude references - if it seems problematic to include them.  A hurdle that I face is that I cannot include data from my workplace.  It is not mine to share.  It therefore probably seems “normal” - since I do it routinely - for a blogger to carry on a narrative without sharing evidence.  However, the fact that I do not post the evidence simply means that I haven’t posted the evidence - not that there is none.  The question really is not whether I am a quasi-intellectual (making assertions that sound intelligent but which lack support) but rather whether I am lying (making assertions although I have no supporting evidence).

 

I would take greater offence if readers called me a quasi-intellectual than a liar.  I understand people saying that I lie - since I might post no evidence.  To suggest on the other hand that I have no evidence, that I am a quasi-intellectual, this is disturbing and unsettling to say the least.  Fortunately, when I am providing “guidance,” I am up to my eyeballs in evidence.  (But now readers might say that I am lying, which is fine.)  In all truthfulness, I am not literally up to my eyeballs in data - not in a spatial sense - as if it were pushing up my nostrils into my nasal cavity.  That is merely an expression, which admittedly seems inapplicable to electronic files and databases.  I have an intimate relationship with the data resources.  I know where everything is - unless I forget - in which case I usually have an approximate idea that leads me to the right location.  The entire narrative begins with the data.  I have no preconceptions of what to expect.

 

After all of that ghastly preamble, my main point actually is that evidenced-based management makes it possible for individuals like me to remain employed.  It is true.  If nobody believed in making decisions using data, I wouldn’t be needed.  Note the two components:  “making decisions” and “using data.”  Decisions don’t have to be made using data.  Decisions can be made on the absence of data.  There is also the idea of making decisions “with” rather than “using” data.  What is the difference?  Sometimes people will have a preconceived notion of what has to happen or should be done; they just need the smallest bit of evidence as back up.  An employee for example might have stats pointing to fairly good performance overall - although there might be some areas of weakness that could be improved.  Conceivably, a manager could ignore all of the positive data and focus only on the deficiencies.  Technically speaking, this is not evidenced-base decision-making but rather the rendering of evidence-embellished decisions.

 

The emergence of evidence-embellished decisions in an organization is job-threatening to the data scientist.  The process substantially ignores the truth.  Since anyone can tell a lie at any time without any deliberation, there would be less need to invest to research and understand the truth.  Little Hitler might eventually gain full reign over an organization.  From that point onward, the organization’s connection to reality starts to wane as it literally makes less use of “evidence” to substantiate its behaviours.  The narrative become entirely fictionalized.  Products might be produced that people don’t care to buy.  A retailer could set staffing levels that are inconsistent with the volumes of shoppers.  Employees might be retained not because of performance but because they fit certain arbitrary profiles.  "Jude is a real team player.  He plays golf and hockey.  Good firm jaw.  People person all the way."  In short, many wrong things can happen when Little Hitlers are approached for their uninformed opinions.

 

In order to prevent the tiny annoying fascist inside all of us from taking hold of corporate resources, it is necessary to encourage the build-up of worthwhile data and provide a means to ensure that it is used for guidance.  Moreover, the guidance that is provided should not be from the heart or gut - but only from the data.  Believe me, the heart and gut give few if any constructive business insights - more likely to lead to racist or sexist marketing materials and human rights violations.  When evidenced-based management succeeds, I firmly believe that Hitlers shriek in horror.  Legitimizing the use of data requires concerted effort and I suggest also an understanding of its role in the power structure of an organization.  Space for the lowly data scientist begins to shrink when people in decision-making stop caring about the truth.  Data is the weapon of choice for certain managers - but not all of them.  Data scientists are responsible for providing the best tools possible for evidence-based managers - essentially to thwart the rise of fascist dictatorships.

Views: 164

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Shay Pal on March 6, 2018 at 4:19pm

Very aptly said, and thanks for posting!

Videos

  • Add Videos
  • View All

Follow Us

© 2018   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service