Consider this a Public Service Announcement.  Or just a rant.  And pity Trish:  she has to hear me “go off” about these things all the time.  I would like to think that writing about this topic is a way to get it out of my system, but who am I kidding?  I will most likely continue to correct TV announcers for the rest of my life.  Herewith, a short list of things about the non-scientific part of the world that drives the scientist in me nuts.  Daily.

The word “data” is a plural noun.  The singular is “datum” but you hardly ever hear that used.  Instead, the word “data” is used as a singular noun.  Of which it is not.  One might say, “The data shows that….” Or “The data is clear on that topic,” but one would be WRONG.  This item is one of those little seemingly insignificant things that, once pointed out to you when YOU use it wrong, sticks in your craw forever.  Like “I could care less” or “chomping at the bit” or “irregardless”.  For the record:  it’s “I couldn’t care less” and “champing at the bit” and “irregardless” isn’t even a word!  Use “regardless” or “irrespective”.  We are veering from science to grammar a bit but my point is about using words in a way that shows that you understand them.  There is an awful lot of that in the scientific community—using jargon in a way that indicates lack of understanding of real meaning.  I can forgive the non-science world for using “data” as a singular noun since it is probably the only way most people have heard it used.  However, any scientist (or journalist) should know better.  Now this will drive you bonkers just like it does me.  You’re welcome.

Significant Digits Matter.  This explanation must begin with a story.  When I was between my junior and senior years in college, I stayed up in Baltimore to continue my undergraduate research work.  I was young and just getting my first exposures to what real “research” meant and I had only a cursory understanding of what I was doing.  My advisor would send me up to Swarthmore College on a regular basis to use an instrument there.  While there, I worked side-by-side with two students from that school who were much more capable scientists than I was.  We were all doing similar work and would have a research review to discuss our results each Friday when I was there.  I was showing a table I had put together presenting the data I had collected that week.  The guys were snickering and whispering to each other.  Finally, they laughed out loud and said, “Look at all those significant digits!”  I wanted to drop through the floor.  The instrument gave me a string of numbers.  I put that string of numbers in my table.  If the instrument reported a value of, say, 1.942583, then that’s what I put in the table.  I did that with each entry, so the screen was a sea of long numbers.  The problem was that the “confidence” in the measurement was only to the second decimal place.  There was no real difference between 1.942 and 1.944.  I should have just reported 1.94.  Most of you have probably fallen asleep by now (Trish does) but I have a serious point.

Let’s take an example we hear about regularly:  We’re in an election year (Ugh).  What if I told you that Candidate A leads Candidate B in the polling by a measure of 52.4 to 47.6.  Sounds like a significant lead, no?  But if the margin of error is +/- 3 points (not uncommon) then this race is a dead heat.  Polling numbers should never be presented to three significant digits, but they often are.  And then they are over interpreted.  The pundits go breathless analyzing what Candidate B is doing wrong when there is really nothing definitive they can say about the race!  This happens all the time when data of any sort are presented in the news.  No one seems to know how to put numbers in context or certainly use the correct number of significant digits.  How do I know if any number is meaningful when I have no context?!  You  know what commentary is like on monthly jobs reports.  New jobs went DOWN by 20,000 from the month before!  We must be heading for a recession!  Millions will be out of work again!  Food lines!  Crime increases!  BE AFRAID!  Of course, new job numbers are regularly revised by well more than 20,000 each month since collecting accurate data is difficult.  THIS is why I am constantly yelling back at the news. 

When I taught Freshman Chemistry in grad school, I taught the lesson about significant digits in the first week of class, but deducted points all semester long if significant digits were wrong in calculations.  When someone complained that the topic was from the first week of the semester, I’d answer, “And clearly you didn’t learn the lesson.”  Significant digits matter, if for no other reason than to keep you from going crazy when you watch the news.

The Scientific Process is not Widely Understood.  This issue has been magnified as we’ve gone through the COVID pandemic.  Never have I seen the scientific community communicate so poorly nor the general populace demonstrate their lack of understanding of the Scientific Process so profoundly.  It was literally painful to watch the news each day.  The Scientific Process basically works like this:  First, you form a hypothesis.  Your hypothesis may be something like “wearing masks helps reduce infectious spread of viruses” or “pharmaceutical A can cure people of COVID-19”.  Next, you devise experiments to test that hypothesis.  Those experiments must be carefully designed so that you are sure you are testing only the question in your hypothesis.  You have to articulate all the little assumptions that you are making about what might or might not affect what you are measuring so you can account for all those other influences on the results.  Designing experiments this way is really difficult!  Then, there must be enough data such that your conclusions can be statistically valid.  You run your experiments.  You analyze your data and draw your conclusions.  If the results don’t fit your hypothesis, you can modify your hypothesis and/or design more experiments to gain additional insight by changing your assumptions.  Maybe, for instance, pharmaceutical A only works if administered within 3 days of infection.  Or maybe there were people who got relief from COVID-19 when they took pharmaceutical A but it was not BECAUSE they took pharmaceutical A.  We’ve already discussed the danger of anecdotal information.  Finally, you have your peers review your assumptions, experimental procedures, data, and conclusions and have them tell you everything you did wrong.  Science is so much fun!  But, in following this process (which takes a fair amount of time), you get reliable results, not anecdotal information.  It’s not fast.  It’s not fun.  It’s kind of boring.  But it WORKS.

There are other topics.  Like the fact that the general populace tends to use “chemical” as synonymous with “toxic chemical”.  Or the common assumption that if something occurs in nature it is, by definition, “natural” and therefore safe and healthy.  Last I checked, arsenic is naturally occurring.

And that brings me to my final thought.  When I asked my niece (a plant pathologist) for her thoughts, she reminded me that what often bugs her the most is how annoying the ego of scientists can be!  Hmmm.  I have often described understanding chemistry as like learning a foreign language.  If you don’t learn the early stuff well, anything more advanced is totally unintelligible and is either suspect or just misunderstood.  I was lucky.  I had the interest coupled with good early teaching.  Most people are not in that situation.  Maybe, then, I should get down off of my pedestal and not let these things bug me so much!

[Editor’s Note:  Can I hear an AMEN?]