Tuesday, February 28, 2017

Fake News: Spinning and Winning



Fake News: Spinning and Winning
“Truth crushed to earth, shall rise again” (William Cullen Bryant)

In Nelson Goodman’s (1954), Fact Fiction and Forecast, the notion of projectible predication arises to differentiate hypotheses based on regularities well grounded in experience and those which are not. There is a parallel treatment of counterfactuals (If X, then Y, and not-x, as in “If this thing in my hand was made of copper, it would conduct electricity but it is actually a wooden popsicle stick”) involving relationships well-grounded in experience and those which are not. What might we say about facts, fictions, and forecasts, in educational research or in the current political climate?

What are some facts in the area of educational research? A study by the 2014 Program for the International Assessment of Adult Competencies (PIAAC) involving 33 countries shows that 7 of those countries scored significantly higher on a literacy scale and six scored significant lower (measured understanding, evaluating, using and engaging with written text) than the USA and the USA was slightly below the average (see https://nces.ed.gov/fastfacts/display.asp?id=69).
Here is an associated counterfactual claim: If an adult person M (say that is me – an American) is a lifelong resident of Japan, then M is more likely to be literate than N (where N is an adult American picked at random). Is that counterfactual claim reasonable?

Here is another claim supported by extensive educational research: Directive feedback (providing corrective information) tends to work well with learners new to a topic or domain whereas facilitative feedback (providing guidance and cues) tends to work well with more advanced learners (Shute, 2007; see https://www.ets.org/Media/Research/pdf/RR-07-11.pdf). Now, suppose that P is a learner new to the domain of logic and epistemology (me, for instance – my dissertation was in that area) and someone claims that P is more likely to benefit from directive feedback in the area of logic than Q (a middle school student in rural Alabama). Is that a reasonable claim?

One way to treat counterfactuals (IF-Then claims with the IF-clause clearly false) is to dismiss them as trivial or even meaningless. Yet some seem to make sense to some people. Other counterfactuals can be used to make jokes as in: “If I knew everything, then I would know _______________ .” I forgot to mention that this was a pop quiz. How did you fill in the blank? I used this phrase: “… then I would know where parallel lines meet.” Math humor is not so humorous to very many people.

On to fictions. I just love fictions. Sometimes I think about my training in philosophy … one of my professors said that the never-ending business of philosophy was to help us understand the boundaries between sense and nonsense. My own take on philosophy is that it is a kind of thought in slow motion. Fictions – claims that do not hold up under scrutiny. Scrutiny is when you close one eye and take a closer look for those of you taking notes. There are some blatant fictions as this one I discovered in a book on medieval logic: “I just ate the last cannibal” spoken in a group of monks who had taken vows of silence. Bouwsma’s (one of my professors) example was this: “I just suffered a fatal heart attack.”

There are many less blatant fictions. Here is one: “Humans only came to the Grand Canyon area about 4,000 years ago.” Here is another one: “All of the fossils found in the Burgess Shale in the Canadian Rockies were fossils of creatures still living somewhere.” Consider this one: “There is no evidence that human activities contribute to climate change.” Some people believe what they want to believe and are reluctant to take a closer look at evidence or consider alternative perspectives or beliefs. There is a difference between advocacy for something and evidence supporting something. A critical issue concerns the nature of good and compelling evidence. Just as there is a fuzzy boundary between sense and nonsense, the boundary between advocacy and research is somewhat fuzzy. 

Just as counterfactuals turn out to be somewhat problematic, there is another kind of IF-THEN claim that is also problematic. I call it an unconditional conditional and it has the general form of If X, then Y, where no matter what is put in for X the person making the unconditional conditional claim will maintain the truth of Y. No refutation of the unconditional conditional is considered possible. In such a case, one cannot conclude that Y is a fiction … one can only walk away from the unreasonable challenge of the advocate of the unconditional conditional in trying to offer evidence that Y or the unconditional conditional with Y as the then-clause may not be true. 

Does this ever happen in educational research? In medical research? In political discourse? The challenge of finding examples in each of those categories is left to the reader – this is the mid-term exam. Hint – the answer to the first set of three questions  is ‘yes’ – this does not constitute timely nor informative feedback. It is merely encouragement to keep on keeping’ on.
When you have completed the mid-term exam, you may want to continue on to forecasts. My forecast is that some of you will pass the mid-term. After all, it was a take-home exam … or take-to-the-bathroom exam.

Having said a few things about IF-THEN claims, it seems natural to apply some of that discussion to forecasts, as these often come in the form of complex IF-THEN statements, such as:” “If W, and X and Y, then Z” – W might refer to the learning or instructional context and X might refer to the students or teachers and Y might refer to the intervention or treatment. Obviously, each of the parts of the complex IF clause could be compound, which means that the forecast result Z depends  on a conjunction of a set of contributing factors. If Z does not occur, the advocate of Z is likely to look for one or more deficiencies in the set of contributing factors. Another approach is to construct a replication study or a revised version to see if Z might occur. Yet another approach is to revise Z and conduct a replication study. Forecasting or predicting and then confirming or refuting or refining is not easy … it is what scientists and meteorologists and other investigators are trained to do. 

I have a vague memory of reading Fact, Fiction and Forecast about 45 years ago. I was fascinated by the concept of the hypothetical predicate ‘grue’ for things that are green before some future date and blue after that date. At this time, ‘all emeralds are green’ and 'all emeralds are grue’ are both true and confirmed by the same evidence. However, few believe that after that future date that emeralds will all be blue. I also realized that I did not understand what a meteorologist meant by a forecast of 40% chance of rain. Was it that 40% of the area covered by the forecast would surely receive rain, or that any random spot in the forecast area would have a 40% chance of rain or that it will rain 40% of the day or ??? Forecasting still bewilders me. I recall a sports enthusiast being asked to predict the outcome of an event about to begin. The sports enthusiast replied “Let’s just watch and see what happens.” My respect for sports enthusiasts rose significantly that day.

I suppose we need a final exam since we have had a pop quiz and a mid-term exam. The final exam is a single multiple choice question:

Which of the following statements is true?
  1. There is someone in this room who loves all and only those persons in this room who do not love themselves.
  2. Never in the course of human history have events so resembled the present as they now do.
  3. It is a fact that X leaked Y but that fact is fake news
  4. There are an even number of planets in the Milky Way galaxy
  5. If X is a human being, then X knows less than X is typically inclined to believe that X knows.
  6.  More than one of the above is true
  7. More than one of the above is false.
 Truth? What is truth? I will go where you go, answered Ruth. My trumpet is louder than yours so follow me said someone else. The truth they are telling might only be the truth that is selling. And the slow one now will later be fast said the Nobel laureate.