Lies, Damned Lies and Statistics

Image by Tumisu from Pixabay

There's something to be said for what you think is a lost cause.  If you can't do anything about it, at least raise a laugh - a joyful belly laugh or a nervous whimper or a knowing snigger - there are choices aplenty for those who will squeeze fun out of anything and lemonade out of lemons.

The Lying Kind

"There are three kinds of lies: lies, damned lies and statistics!" these lines attributed to Mark Twain, among many other contenders since the 1850s, got me thinking about the connection with what we are doing with numbers, kids and even schools these days. Actually we have probably been doing it since Mark Twain's time too which might have prompted him to opine, "Never let school interfere with your education."  We have been reducing the kids to performance statistics, perhaps preparing them for performance appraisals in corporate life?

98% of our students have scored 90% or more says one institution

100% success rate!  says another

These stats eventually guide the admission process for those institutions.  "We only take the best since we have a reputation to maintain".  Which begs the question, "If we only take the best, what do we do with them the rest of those 12 years?"  And the other question, "Why do a few students repeat a class every year?"  But, don't panic!  Statistics are on our side.  We have "less than 0.01% failure" -- perhaps adopted from the need for stringent quality assurance norms in later life - the ones named with Japanese Terms or Math Formulae.  And of course, with the rules of decimal rounding, these are negligible numbers in reality.  Except for that child who didn't make it.  He suddenly knows he isn't the best.

Image by Colin Behrens from Pixabay

We Excel in Everything

With the forced use of computers (especially now in the online learning phase) we now have potent tools to prove points or lack of them. We can show graphs and negate the niggling doubts in a parent's mind about the capability of their kids. A well made Bar Chart can drive a harried parent to drink (no connection with the bar chart, the one where the prices keep going up!). If you look at the full page advertisements for coaching in JEE and the like, you will see several toppers and accompanying statistics to indicate they all came from that advertiser's stable. Depending on the paper you read, the same multitasking topper features in other stables too -- didn't they shut the real stable door in time?  To be fair, the Topper probably did enlist in several coaching centres to avoid putting all the eggs in one basket.  But which one finally got him to the top?  Statistics in every centre will prove the point -- it was them!

Data Validation and Data Torture

Ronald H. Coase, a renowned British Economist is the author of the quote- “if you torture the data long enough, it will confess to anything”.  And that is precisely what happens when various magazines come up with "surveys" to pronounce a school the Best in Class in some particular area ranging from the highest number of students to pass a particular benchmark to possibly the best teacher-to-pupil ratio (private tutors are not counted). Helped by the computer it could further granulate this knowledge into single-gender schools, co-educational schools, rural, suburban and city schools -- all of which eventually creates some level of bias in our minds, at least in the education world.  Because it's proven with statistics!  I love the asterisked line in one such survey that claims that "we interviewed 2659 parents in 6 cities, from the upper affluent class!"  So much for broad-based data.

Best in Class?

Measurement and Statistics can also be very helpful.  I recall a student of mine was most happy as he had "come first" in an examination.  Big smiles turned into question marks, then frowns as I pointed out that he had apparently achieved only 80% of what he was expected to master.  The fact that his arch rival had achieved 78% made him gloat till he realized that he was still a whole 20% short of mastery. So, ranking has this effect.  You only need to be a little better than the next guy.   Once this student was able to understand the "so what"  he realized that his greatest competition was himself and he then went on to achieve more on his own. In the school where I taught there was a special prize for "Effort" -- awarded to students who had improved their scores though they were not top of the class.  This encouraged bottom-scrapers like me to actually attempt to make a little more than the pass mark. We didn't go on stage to get prizes -- who would be left to sit in the rank and file and applaud?

Vital Statistics?

We spend a lot of time teaching kids the What, the Where, the When, the Who and occasionally the Why and sometimes the How.  And then grading and scoring them on what they remember of what they learnt.  Rarely do we approach questions like So What?  How does it matter?  How will it change your life to know that half the class scored better than you did? Will you get the chance to change that?  No.  It's now indelibly inked on the Report Card and is now your badge of under-performance.  Average, above average, below average .. my knowledge, my understanding, my so-called performance is only judged in relation to that of the others.  

The question that institutions need to ask themselves is, "How important are these statistics?" And the obvious addenda, "How well understood are the results?"

For example, teachers could be spending hours on analyzing marks and then producing statistical data -- the number of students in a particular achievement band, standard deviation scores (now available with in-built functions), pretty pie-charts, line graphs and bar charts, some of them stacked.  Who is interpreting this data?  Who is drawing out inferences and basing their action plans on this?  How many parents (or teachers) are able to absorb and then act on the basis of this input? Does it give any information to the administration on whether the teaching has been effective, or more effective in relation to previous terms?  Can the statistics be used to identify cause and effect?  Can an individual teacher be helped to change his or her style and approach based on the statistics?  I sincerely hope these questions are answered in the affirmative.  We would see quantum leaps in quality.

But the answers to all these questions, as Bob Dylan would sing, "are blowing in the wind". Time to stop flying kites and bring them down to earth, reel in the stats instead of reeling them off.  

My Learnings from all this

• Data is only as good as its use.  Storage is only useful when retrieved.  Colourful graphics are only useful when there are action points emanating from them.

• Everyone can learn to input and analyze, but to Interpret requires a different skill. Technical use of  software does not ensure proper and useful interpretation.

• Producing information from raw data assumes the raw data is correctly input.  Wrong data, wrongly interpreted can destroy dreams.

[If you found this interesting, please leave your own views in the Comments section on the blog.  Share and invite others to the blog.  Also, let me know if there is something you would like me to engage with. - Leslie]