No Zero Policies and the failure of Educational Research

A personal pet peeve is this "no zero" policy some experts are pushing.  In particular, let's look at what one person pushing this, Dr. Douglas Reeves, has to say:

"First is the use of zeroes for missing work. Despite evidence that grading as punishment does not work (Guskey, 2000) and the mathematical flaw in the use of the zero on a 100-point scale (Reeves, 2004)..."

From the same article as above, "get the facts; gather evidence that will create a rationale for decision making."

Sounds like a plan; let's look at his first citation, Guskey, 2000:

"Instead of prompting greater effort, low grades more often cause students to withdraw from learning.  ... Other students may blame themselves for the low grade, but they may feel helpless to make any improvement (Selby and Murphy 1992)."

And here the problem becomes obvious when you look at their citations:
"Selby, D., and S. Murphy. 1992. Graded or degraded: Perceptions of letter gradeing [sic] for mainstreamed learning-disabled students. British Columbia Journal of Special Education 16 (1): 92-104."

Basing an argument on "mainstreamed learning-disabled" students only?  Even better, you're basing it on SIX of them:

"This study of six mainstreamed students (in grades six and eight) with learning disabilities, their parents, and their teachers..."

So there's one of his two reasons trashed.  Let's look at the other, the "mathematical flaw".

He makes two points.  The first is an assumption:  "To insist on the use of a zero on a 100-point scale is to assert that work that is not turned in deserves a penalty that is many times more severe than that assessed for work that is done wretchedly and is worth a D."  There's not much to say here; he points out that punishing students doesn't work, and I agree.  But giving a zero for zero effort is not "punishing"--that's just GIGO.  The only other time I--and, I'd like to believe every other teacher--would give a zero is in cases of cheating, where it is entirely appropriate to punish someone.  Certainly giving out zero's left-and-right is unfair (and probably unethical), but this is just a simple matter of fairness, in my opinion.

The second argument is also an assumption:  that you must be consistent in that every grade must be 10 points apart or you are being "unfair".    Ironically, he makes the statement that "that many people with advanced degrees, including those with more background in mathematics than the typical teacher, have not applied the ratio standard to their own professional practices."  He's assuming we say that every ten points must mean something; there's nothing that dictates that and again, this "justification" is just an assumption.

This argument is dismissed by assuming that a "failing grade" is anything that doesn't meet a certain threshold.  That is, you have to do so well before we'll consider you competent and after that point, you go up one letter grade for every 10 percent.    His argument relies on the "fairness" that every grade must be 10 points apart.  There is nothing in a piece-wise function that makes this mandatory--it's just something he has assumed must be true.  He also seems to fail to take into consideration his 0-4 point scale radically changes how we assign grades.  Due to the cardinal nature ("ratio standard") of grading it seems to me the scale must remain linear, so 0-4 can just as easily be represented as:

  0-19%  F  ("0")
20-39%  D ("1")
40-59%  C  ("2"  Note: You are now "average" even if you know less than 50% of the material)
60-79%  B  ("3" Slightly above half--60% and up--is now categorized as "above average")
80-100% A  ("4"  And here you have an even larger percent of kids who are "excellent".)

If a student only knew 20% of the material, would you consider that "competent"?  Note that you've doubled the ranges for each of the grades A-D.  If you think we had a grade inflation problem before, wait until this becomes acceptable.  Good luck determining who is truly excellent when "A" (a "4") means the top 20% of kids.  The grades become too ambiguous to be useful under his "four point" (apparently the zero doesn't count?) scale.  Note that if the zero is reserved for ONLY assignments not turned in, the scale becomes even more inflated/ambiguous; 1-25% is a "1", 25-50% is a "2", 51-75% a "3", and the top score will now include everything from 76% up!

But perhaps the greatest bit of wisdom is from a statement he made in the national press, where he (I assume) tried to briefly summarize his "ratio standard":

"It's a classic mathematical dilemma: that the students have a six times greater chance of getting an F," says Douglas Reeves, founder of The Leadership and Learning Center, a Colorado-based educational think tank who has written on the topic."

The "chance" of getting an F?  While chance/luck will certainly play a role in everything we do, short of guessing on every answer, I'm pretty sure ability, practice, and preparation are going to have a far greater influence on your "chance" of getting an F.  In other words, the esteemed doctor's "classic mathematical dilemma" rests on the assumption that grades are random variables.  (Maybe he should consult some of those highfalutin folks with "more background in mathematics than the typical teacher" before making any more statistical arguments...)

So we have some research that was, to be polite, very poorly done (the egregious misinterpretation of the original source would warrant an "F" in my class...I'd be hard pressed to not give him a ZERO.) and a "mathematical dilemma" that is based on several flawed assumptions and would result in some horrible unintended consequences.  And yet our educational leaders are buying this hook, line, and sinker.