Student assessment system remains controversial

By Dennis Nilsson and Noelle Loh



There is one overarching problem in NUS when it comes to the grading of students: a dire lack of transparency.

In an article in the February issue of the NUS Students’ Union publication, The Ridge, students are reported to be “uncertain” when asked about the grading system used in the university. The report, which raised the questions of whether a bell curve grading system is used in NUS and, if so, why and how, also suggested that information regarding the NUS grading model is obscure.

The Observer has made its own investigation into the academic grading system used in NUS, and responses from students and university personnel alike reveal uncertainty and inconsistencies.

The people who know, or who should know, about the ways of NUS grading are either not willing to discuss the matter or do so in rather vague terms.

An air of confidentiality, secrecy and uneasiness surrounds this issue in NUS.

Of the 19 university administration personnel and faculty members The Observer contacted via email with regards to the academic grading system applied in NUS, only three agreed to an interview.

The others either did not reply or declined to be interviewed, citing busy schedules or the inappropriateness of discussing the grading system in NUS. In his reply, vice dean of the School of Computing, Khoo Siau Cheng, referred The Observer to the provost office, as “the principle (of grading) is set by the university.”

However, the vice-provost of education, Lily Kong, who leads the University Committee on Education Policy, was one of those who did not reply.

Why is it that NUS staff members are so silent about this matter? What is all the secrecy and “inappropriateness” about?

Why all the secrecy?

Dean of the Faculty of Science at NUS, Professor Tan Eng Chye, said the university tells faculty members to keep information concerning academic grading confidential and would not like students to know about such matters.

“We don’t want to create the wrong image,” he said in a face-to-face interview with The Observer. “We are afraid that if we give you the guidelines, then everyone will think that that is the rule.”

“We leave it to the professors to do their grading,” Tan explained. “We have guidelines, but if there are deviations in guidelines, we actually allow the professors to explain to us.”

He added that the university would let these discrepancies pass should the lecturer be able to account for them.

Tan elaborated on the nature of the “guidelines” and the “checks and balances” done at the department level.

“The guidelines are only applied to modules that are large. Maybe more than 30 people,” he said.

“We look at the proportion of people who gets D’s and E’s. We also look at the proportion of students getting A’s,” the FoS dean explained.

Co-ordinators for each level and deputy heads will look for discrepancies in the marks given in each module, he added.

The Observer also asked specifically whether NUS students are marked according to a bell curve system.

“We have guidelines, certain guidelines, but we are flexible with those guidelines,” Tan replied.

The controversy of the bell curve

To grade according to a bell curve, also called the normal curve, means that the percentage of students who are to be awarded each grade is predetermined according to an estimation of what the norm is.

The grade frequency should form a bell shape graphically, with most students receiving a mediocre grade in the C range, fewer students receiving the grades above and below and, finally, only a few people assigned the top and bottom marks.

As an example, a bell curve grade distribution in a class of 100 students could look like this: A - 10, B - 20, C - 40, D - 20, E and F - 10 (disregarding + and – grades for the sake of clarity).



An important point about this system, which is also a major point of controversy, is that a student’s performance is thus assessed relative to the performances of fellow students in that module and not solely according to the student’s own performance.

In the above example it has been predetermined that only the top 30 students will get A’s and B’s. This means that a student would have to settle for a C if his or her score is ranked “only” 31st or 32nd in that module, regardless of how that performance would have been assessed in its own right.

In this way, this “normal” distribution cannot be trusted entirely to reflect how well each student performed according to the expected criteria.

“Hypothetically, if a bell curve – which is only a statistical observation, not a rule to be enforced – is artificially imposed on a class of fewer than 400, we can have very little confidence statistically that this is ‘normal distribution’” said Dr. Linda Perry, a professor from the Communications and New Media programme.

A C student could in theory have scored 92 out of 100, or an A student could have scored just 67 out of 100.

There are other highly problematic issues related to the bell curve system.

How do the teachers make the actual scores fit the bell curve if they are not “normal”?

What if two or more students get an identical score and are placed on the border between two grades? Who gets lucky? Who does not?

Does the teacher then throw the dice?

If where a student falls within the curve includes an element of luck, then how does that affect his professional prospects?

Tim Gorman, an American graduate student in Southeast Asian studies at NUS, said he believes a bell curve grading scheme can have a detrimental effect for NUS students who will later compete with Western students in the job market.

The bell curve system is not common in the United States, he said, and if NUS students have “a bunch of B’s and C’s” it would make them “look like they didn’t do very well in their undergraduate years.”

If the bell curve system is followed, a lot of students will get a lot of B’s and C’s, without necessarily deserving it.

“I don’t think it is necessarily fair to be graded against your peers,” Gorman said. “Certain classes will have a concentration of stellar students and some classes won’t.”

The secrecy of final examination marks

While students continue to guess whether a bell curve grading system is de facto used in NUS, another question looms — why are the grades of students’ final examinations not disclosed?

It is standard practice across faculties to show students the marks they attain for the various components of continual assessment for each module, but not that of final assessment, making how teachers arrive at a student’s final, overall grade unknown.

And while students “may request for a review of their examination results,” as stated in a circular sent out by the NUS Registrar’s Office prior to the release of examination results every semester, it does not mean that they get a chance to look at their final examination papers.

According to Professor John Richardson, vice dean of the Faculty of Arts and Social Sciences, only students who receive a failing grade can check the papers for themselves.

“The policy is you can pay $10, I think, to have the arithmetic checked. We’ll check that it’s been properly marked, that there have been no mistakes in the sum,” he said. “Students who have not failed cannot check their papers.”

Tan acknowledged this university procedure of reviewing examination results but expressed doubt that students ever get to see their examination scripts.

“Even if you fail, I don’t think you’d get to see your paper,” he said.

For Tan, such a review practice has been implemented because there would be a group of students who would “continually tell the professor that (the) grading is unfair” and the university wants to avoid this.

He said he sees such scenarios occur during the semester when teachers return students’ term paper assignments.

Assessment policy questioned by teachers

Teachers themselves are not blind to implications of the NUS academic grading system.

According to the Centre of Development of Teaching and Learning’s brief of Aug. 10, 1999, a seminar on student assessment was organized earlier that year by the CDTL and led by the then-deputy vice-chancellor, Professor Hang Chang Chieh.

Hang is currently director at the Centre for Management of Science and Technology.

At the seminar, NUS staff participants voiced concern over the use of a bell curve grading system and the inherent implications.

The norm-referenced (bell curve) system “does not tell us what the students can do,” said one unnamed participant, as recorded in the CDTL brief.

“I think top international companies would be more interested in what the student can do rather than what percentage of NUS he has achieved,” the seminar participant added.

As the surveys conducted by both The Ridge and The Observer have shown, most students seem to assume or believe that teachers are applying the bell curve grading system when assigning NUS undergraduates their grades.

“I’ve heard it from professors. They talk about their under-grad classes. They have to grade on the bell curve,” Gorman said. “From what I know of it, it is something they are forced to do, like a school policy.”

Hang attested to students’ assumptions about the use of the bell curve grading system.

“We always have a bell curve. We cannot accept the fact that only 100 (students) are the best in the country and all of them score only A’s and B’s and no C’s. We cannot accept that,” he said to The Observer.

An unnamed participant at the CDTL seminar agreed with this assessment. The participant is recorded to have said, “We are told to satisfy the bell curve.”

The participant added that “in this case, the already deserving A and B students are not properly assessed.”

Competence of teachers questioned
According to Richardson, the grading system in NUS is one that involves a “moderation process” that tries to ensure that grading is fair across modules and departments.

“What we are trying to do is to avoid grade inflation and grade deflation,” he said.

Richardson, who is British, shared that one would be shocked to know of the large number of students who are awarded with first- and second-upper class degrees in England.

“If all students are getting second upper-class honours, how do you tell a good student from a poor student?” he said, adding that grade inflation is “pointless.”

During The Observer’s investigation, Richardson has not been the only one of the opinion that the grading system, bell curve or otherwise, is offering solutions, not problems.

One of the problems that the system is claimed to help solve is, most curiously, that of teachers’ competence in grading.

In the 1999 CDTL seminar, in response to suggestions that the use of a bell curve be replaced by a system that gives teachers the autonomy to assign final grades, Hang said there would be “misuses and problems.”

“The problem lies with the fact that this is not Cambridge University, U.C. Berkeley or Stanford, where every single teacher can be safely empowered and the student abilities are more homogenous,” he said.

“While staff who are fair will spend 50 percent more of their time grading, others will spend more time on consulting,” he explained.

Even though Tan said he felt Hang’s seminar comments were too strong, the FoS dean does not deny them, pointing out that teachers have different backgrounds and may not be familiar with NUS standards.

“It’s not so much to say that we are not sure they can do it. Of course they can do it,” he said. “(We apply our grading system) just to make sure that they understand our context. We want to make sure that we set high standards.”

The implications of Hang and Tan’s words are serious. How can it be that the teachers in a university that is ranked by The Times of London as 18th in the world — based on factors that include the institution’s strength of internationally renowned academics —cannot be fully trusted for their abilities in assessing students?

The incongruity of this situation does not evade Gorman.

While the graduate student said standardizing grades might help because the standard of grading can be “wildly off,” he pointed out that one should respect the autonomy of instructors.

“If you can’t trust the judgement of a professor, then what’s the point?” he asked.

More importantly, that Hang’s 1999 comments on teacher empowerment and academic grading in NUS contrast starkly with the views he shared with The Observer does not simply highlight a flaw in the grading system in the university. It could also make students and academics alike doubt the credibility of NUS.

During his interview with The Observer, Hang said teachers today have improved and are “fully qualified” to assess students’ academic performance.

“We cannot control our teachers on all levels, but you easily walk into any faculty or any department, the top one-third (of teachers) is as good as anywhere in the world,” he said. “When I say top one-third, I’m being conservative. I would say at least (one-third).”

On the same note, Hang said he felt teacher empowerment in grading was still lacking in NUS.

“Now we treat a young assistant professor with no experience the same way we treat a full professor. All of them are not empowered,” he said.

“If they have a bell-shape curve, no questions asked. If they have a very skewed curve, they are in for tough questioning. So most people will stick to the bell curve,” Hang elaborated.

At the CDTL seminar, an unnamed participant said he also desired more flexibility and teacher autonomy in grading. Satisfying the expectation of a curve meant teachers had to “do all kinds of tricks to adjust the marks to reflect the bell curve,” he said.

Questions remain unanswered

It would be in the interest of everyone, students and staff alike, to learn more about these “tricks” and to have an open debate about the NUS academic grading system.

As it stands, with no transparency in grading, students are left to wonder if the part of their numerical grade that is kept secret from them is a part of the “tricks” used to force the curve.

Similarly, it would be in the interest of everyone for students once and for all to learn from the NUS administration whether a bell curve grading policy is applied in this university and, if so, how.

One of the points of learning, basically, is to replace assumptions with knowledge. The reasons given for the lack of transparency and for keeping this issue in obscurity are less than convincing.