A, B, C or 80, 90, 100,
that is the question!
百分制或等第制

Da Hsuan Feng 馮達旋
Senior Executive Vice President
National Cheng Kung University

“In a nutshell, the NDUH(國立東華大學) grading model is to strictly
match a letter grade with a numerical range.”
Preamble
Recently, universities in Taiwan are engaging in vigorous and healthy
discussion, and debate, about whether students grades, especially the
undergraduates, should be the traditional and time honored numerical value
from 0-100, known as (百分制 or hundred points system) or letter-grades of
A, B, C (等第制.) See for example the following news reports:
http://news.chinatimes.com/focus/0,5243,50105619x112010061100108,00.h
tml
http://news.chinatimes.com/focus/0,5243,50105619x112010061100092,00.h
tml
http://news.chinatimes.com/forum/0,5252,110514x112010061300347,00.ht
ml
No doubt, a student’s grade is a characterization of his/her learning quality
and knowledge level, at least it should be. With the aim of becoming
“global,” the incentive for this transformation comes from the desire for
Taiwan universities to dovetail with western universities, especially those in
United States. As is well known, “Grade Point Average” or GPA is part of
their educational DNA for students in United States!
With nearly 22 years as a professor of physics in North America, I would
like to take this opportunity to discuss my personal understanding and
practice of the grading system in one university, Drexel University in
Philadelphia. I should mention that my information is some fifteen years
out-of-date, although I believe that the situation may not have altered
significantly. Without having done a careful universal analysis of all US
universities, I cannot say that what we have done in the physics department
at Drexel University fifteen years ago represents the general procedure of
grading in the country. However, based on many discussions with
colleagues in many universities, I am fairly confident that our method does
not significantly deviate from the norm.
I would also take this opportunity to give a comparison of my experience
and the Taiwan model. I hope this could serve what the Chinese would call
“to throw a stone to draw a jade” (拋磚引玉.)
The National Dong Hwa University 國立東華大學System: The Taiwan
Model
Perhaps what struck me most about the Taiwan “new” system is the way A,
B, C’s are defined. In fact, I learned that such a system has been in existence
in one Taiwan university, National Dong Hwa University (NDHU 國立東華
大學) in Hualian. For more than 15 years, this university has been practicing
such a grading system. In a nutshell, the NDUH grading model is to strictly
match a letter grade with a numerical range. I should point out that the
system is more microscopic then that, since it would have A+, A and A- and
so on. Two other universities, National Taiwan University (NTU 國立台灣
大學) in Taipei and National Tsinghua University (NTHU 國立清華大學)
in Hsinchu will follow NDHU system soon. Many other universities,
including mine (NCKU 國立成功大學) are taking a “wait-and-see” posture.
For discussion sake, let me stick to the simple A, B, C (without A+, A and
A-) since the principle is the same. In the NDHU system, which will also be
practiced by NTU and NTHU, an A grade, for example, would correspond
strictly to a particular range, say, 90-100, and B 80-89 and so on. Hence, if
one defines A to be a GPA of 4.0, B to be 3.0 and C to be 2.0 and so on,
then the average GPA would correspond to a well known numerical range.
Take, for example, an extreme case, an average 4.0 GPA after four years of
undergraduate studies means that all exams, in any subject, taken by this
student, must be in the 90-100 range.
I should mention also that in the US, there is no universal equivalence of
GPAs. For example, a 4.0 GPA from a less known university may not
have the same perception of quality for potential graduate schools or
employers of a 4.0 GPA from a well known university. This, as we know, is
a fact of life since “perception is reality!”
My experience
Let me discuss briefly my experience at Drexel University in Philadelphia.
As a physics faculty member, I taught an entire spectrum of courses, from
freshman physics (for scientists and engineers and for humanities majors,)
different advanced undergraduate courses (such as undergraduate quantum
mechanics and so on,) to an assortment of graduate courses (including
seminar courses on selected topics.)
For introductory courses, the number of students could be of the order of
many hundreds. Thus, instructors would involve a number of faculty
members and many teaching assistants (usually graduate students.) Grades
for each examination were decided by a meeting involving all instructors
(faculty and TAs,) where the the entire distribution of numerical grades of
students (each is between 1 – 100) are analyzed before deciding what would
be the cut of As, Bs, Cs, Ds and Fs. Although the fluctuation from
examination to examination is not significant, it is worth underscoring that
throughout my many years of teaching, I do not recall ever seeing 90’s and
above would be the definition of A. Usually it is lower. This way of grading
is possible because the statistical distribution is meaningful for such courses.
There is always a normal distribution or in some cases, bimodal or trimodal
distribution. In this scenario, the issue of 89 being a B and 90 being an A
almost never arise. Whether a student can get an A or B, or any other grade,
would depend on how he/she compared with his/her peers.
I should mention that I do not know whether my colleagues in freshman
chemistry or freshman calculus carry out identical procedures of grading,
but I cannot imagine that they would significantly deviate from this scheme.
Also, I suspect that all universities in North American have more or less the
same scheme, at least for the introductory science and mathematics courses.
I should also mention that such grading systems are not full proof. In my
many years as professor, there was numerous debates as to how one could
improve grading system. In the end, none were found to be acceptable or
workable.
For advanced undergraduate physics courses, grading is trickier. For
example, in my junior/senior quantum mechanics course, the number of
students was usually small, say less than 15. In this situation, my personal
interactions with the students throughout the course, how diligent they
worked on their homework problems (I tend to give reasonably hard
problems,) how penetrating were their questions, and so on were all taken
into consideration. In fact, the final examination was normally only some
30% of the entire grade. At the end, it is entirely possible that the whole
class be awarded A’s even though on the final examination (in my case, they
tend to be very difficult because I want to test how far I can stretch their
intellectual minds under intense examination condition) their grades could
be low, as low as less than 50 (on a perfect score of 100.) Of course, by
definition, this way of grading has a high degree of subjectiveness. I could
say, however, that when a student gets an A from me, he/she usually
deserves it, from an intellectual and working habit point of view. In this
manner, when the student asked me for a letter of recommendation for
graduate school or employment opportunities, I could give an in depth
analysis of his/her ability as well as learning and working attitude! The point
I want to make here is that a grade of A is not immediately equivalent to a
well defined numerical value.
Although details could be different, I am fairly certain that all my colleagues
in the physics department practiced similar method of grading.
Discussion
There is an old American saying: “if the toilet is working, don’t fix it!”
To this end, in the NDHU model, which will soon be practiced in NTU and
NTHU, it is incumbent to ask the critical question: “why do it?” There are
several reasons lurking behind this question which to me requires some in
depth exploration, if it had not been done already.
- One is that in the NDHU model, A, B, and C’s are strictly
defined by a numerical range. One may question what new
information is contained in such GPAs when one compares to the
numerical averages. Unless I am missing some important points
here, it appears to be almost the same since there is a strict
equivalence. Since the main reason for this change is to dovetail
with the US system, it seems to me that it has only gone halfway.
- The incentive for this transformation is to dovetail with western
(US) universities grading system. In the past half a century,
thousands upon thousands of Taiwan students went to the West,
especially United States, for graduate studies. In their
applications submitted to the Admission Office of a US
university, the traditional numerical grades were provided along
with other supporting documents (such as letters of
recommendation.) Surely by now, US graduate schools have
thoroughly been educated about the Taiwan method of grading.
- Perhaps an interesting question to raise here is how to increase
the information contained in the new system, if Taiwan
universities were to universally make this transformation of
grading system. I am not saying that the way I outlined is better
(or worse,) but I do feel that some of the concerns of the students
raised (B+ and A- different by only one point) could be
minimized.
- Finally, in the 21st century, only a relatively small percentage of
students graduating from Taiwan universities would pursue
advanced degrees in the West. Most will enter graduate schools
in Taiwan or enter Asia’s workforce. In Taiwan, employees have
gotten used to, almost to the subconscious level, the traditional
ways of identifying the quality of students. Thus, any change
would require a new understanding, followed by “new”
calibration. To this end, it is not inconceivable to require that any
new method should contain significantly large (or larger) amount
of new information to characterize a student to make recalibration
a worthwhile exercise. Otherwise, it may be unfair to require the
employment communities to learn a “new way.”
With so much higher education competition globally, it is very heartening to
see that there are many new and creative efforts to impact higher
educational paradigm in Taiwan. This discussion/debate about changing the
grading system shows that what is happening in Taiwan education is robust.
Since grading system, at least for higher education, is a “label” denoting
quality, the practice must be carefully monitored and improved, increase its
strengths and minimize or remove its weaknesses. Moving towards a GPA
system may be one such step.
I am sure in the next few years, with the NDHU grading model being
practiced in NTU and NTHU, the pros and cons, which any new effort
would have, will be able to be “shaken out”. I look forward to the outcome
of this effort.