Re: FW: Grading Alternative

["John E. Purchase, Ed. D." <jopur@muskoka.com>]

Roger Key wrote:

"Gradings are easy and require little effort on behalf of the grader."

Except for that statement, he seems to be on the money.  Marking and grading will not change as long
as society believes that these practices serve a significant purpose.   However, assigning grades to
student work is not easy or requiring little effort (unless one uses the "stair method"  in which
one throws the bundle of tests, assignments, etc., from the top of the stairway and grades according
to which step each paper lands on".

Myron referred to his example on the den-web of a Competency Matrix as an alternative to
"Grading".   The rows state the specific skills which the student tries to master and which both the
student and instructor use to check off the skills mastered.   However,  the columns of the example
are graded.   Whenever any "level" of proficiency is used in the scheme,  grading is taking place
regardless of the title or code given each proficiency column.  The grades with which most of us are
familiar are such "levels" compressed over the  assignment,  project, test, or course.   Although
the reported level does not transmit all the specific details involved in evaluating the given
performance,  those details, in varying degrees,  are not lost but exist in student and instructor
records.  Codes such as "Not Complete-Pass-Fail",  "A-B-C-D-F",  "U-S-O",  "Poor-Fair-Good-Very
-Good-Excellent",   are all grades.   Even if a Competency Matrix were only a check list with one
column,  it would still imply "mastered" if checked and "not mastered" if not checked.  Further,  if
a stakeholder interested in making use of the skills of the individual thus profiled looked at all
the numerous objectives specified on the checklist, he or she would still be required to judge the
overall level (summative standing) of the candidate's competence unless every single objective had
been mastered.

All evaluation is relative,  i.e.,  related to norms or  related to criteria.  Although some
measurement and evaluation authors would like to have clear distinctions between norm-referenced and
criterion-referenced,  on close inspection of all relevant sources of information one uses to make
the evaluation,  the hoped for distinctions blur and meld.

John E. Purchase
======================================================================