Published : 24 Dec 2011, 07:18 PM
Earlier in the week, scientists at the Large Hadron Collider in Switzerland announced that they are closer to actually discovering a new particle: the Higgs boson. Dhaka University (DU), which shared the glory of having introduced the concept of bosons in early 1920s, seems to be set on an autopilot; it has been causing the institution to sink slowly for over last three decades. With years of comparative data widely available, it is not a guessing game anymore as to where DU and many other Bangladeshi universities are heading. Many of those who are driving these different institutional engines appear to be either unwilling or unequipped to interpret data that cannot be just brushed aside anymore.
A University of Grants Commission leader was quoted recently as describing DU's ranking decline as evidence of "other universities of the world progressing whilst Dhaka University is failing to keep the pace with them." When pressed by the media, the person came back swinging, suggesting that many of the ranking criteria are based on issues that are irrelevant to the universities in Bangladesh. The current DU leadership seems to have withdrawn already from the global knowledge race. More folks are happy simply harping on DU's political and cultural role and the glorious political past but hardly ever of academics and research.
Not surprisingly, when it was time to celebrate the 90th founding anniversary of Dhaka University on July 1, 2011, it was clear there wasn't much to celebrate after all other than a 90-pound-cake, colourful parade, a drama, sports competition, a discussion at TSC, and release of nine pigeons. Dhaka University, my alma mater, was placed between 550th and 600th in the Quacquarelli Symonds (QS) rankings in 2010. Worse still, the university's status seems to be on a sustained decline; it was 365th in the QS rankings in 2005. According to 2011 Webometrics ranking of universities, the Oxford of the East was ranked 3627th amongst universities with BUET ranked at 2108th.
While there are genuine reasons such as lack of sufficient investment, research equipment, and library resources that hold us back, our steep decline can be traced back to a single event –"Dhaka University Order 1973" that ended up in effect replacing "meritocracy" with "democracy," once and for all. This was brought upon us by an emotion-driven political process and was participated to by the then academic leaders without much thoughts or regards for the future. A current Dhaka University website entry describes this disastrous milestone as when "democratic norms and autonomy became integral features of the institution." From then on, the so-called democratic practices would be used in identifying the vice-chancellor, deans, and so forth compromising fully the quality assurance so vital to higher education.
Latest websites entries of Dhaka University Faculty of Science departments show the share of their full professors to be now at 43.1%. In case of Faculty of Biological Sciences, in particular, that number stands at 53.4%. These percentages would approach 51.5% and 66.5% respectively, if we exclude the lecturers. None of this is surprising given that promotions are based now solely on political affiliation and whether one has spent an arbitrary length of minimal time in the previous rank irrespective of scholarly or teaching record.
A quick study will reveal that each succeeding VCs since 1972 typically had relatively worse research record than those being replaced. Recent Google scholar search for citation of the published works of all VCs at Dhaka University since 1990, for example, yielded sobering numbers. The average of their cited works is 4.8 and that of their lifetime citations is 40.4. The one with the best record had 14 cited works and a total of 153 citations.
The one exhaustive study that ranks institutions in a non-arbitrary fashion is the SCImago Institutions Rankings (SIR) Report, currently in its third round. In 2011, it ranked 3042 entities based on citation and publications. These 3042 organisations (including 2010 universities) together generated more than 80% of world's scientific output during 2005-2009. While Bangladesh seems to have 75 universities, only four of its entities had enough output to be included in the SIR Report. In comparison, 111 of 310 Indian and 13 of 126 Pakistani entities were included in the SIR.
The SIR Report measures institutions in terms of 5 indicators: (a) output in journals; (b) International Collaboration (IC) which shows its output ratio produced in collaboration with foreign institutions; (c) Normalised Impact (NI) in percentages, shows the relationship of its average scientific impact and world average; (d) Quality of Publications (QP) measured as ratio of publications that gets into top 25% journals; and (e) Excellence Rate (ER) which indicates % of its output within the top 10% of the most cited papers.
During 2005-2009, 58 of 111 entities in India and 8 of 13 in Pakistan were ahead of Dhaka University. In 5 years, Dhaka University researchers authored 845 articles in journals. The university website claims to have 1805 faculty members. That would imply a 5 year production rate of 0.468 papers per faculty member. In other words, it took 11 faculty members a year to produce a single article. 46.8% of its articles had one or more international collaborators. 26% of the articles appeared in journals ranked in the top 25% of their respective categories. The NI value is 0.8 which implies that its published works on average had an impact that is 20% below the world average.
Excellence Rate is 8.1 indicative of the fact that 8.1% of these articles made within the top 10% of most cited papers. Table 1 and 2 list scores respectively for Bangladesh and Pakistan. Table 3 shows the most reputed as well as a few with the lowest output of the India entities.
By comparing Tables 1 and 3, we see that Bangladesh is dependent on foreign collaboration. The case of Center for Health & Population Research, in particular, shows over-dependence while also yielding positive quality, impact, and excellence. In case of others, their publications continue to involve collaborators but low-impact journals. Table 2 shows that in case of Pakistan, government and private entities as well as colleges contribute to research while equivalent Bangladeshi organisations such as Atomic Energy Commission, BCSIR, Sparso, and others have all become victim of the same malaise that is destroying the universities.
The Indian examples of Table 3 show result of a drive for excellence not only in academic and government laboratories but also in relatively small colleges. In fact, successful graduates of better institutions are now influencing smaller, less-known institutions. The output of larger entities is a function of both investment and quality of people they have engaged to drive their knowledge engines. Their IC values show they are increasingly reaching word-class status and not depended any more on collaborations. In 2011 list of U.S. News World's 400 Best Universities, there were 66 entries from Asia that includes 6 each from India and Taiwan along with 12 from Japan, 9 from South Korea, 8 from China, 5 from Hong Kong, 4 each from Israel and Malaysia, 3 from Saudi Arabia, and 2 each from Indonesia, Singapore and Thailand.
In recent years, purposefully, each of the Bangladeshi institutions have been publishing their own sets of low-quality journals that give the false impression that the faculty members are engaged in research yet most of these work will never get noticed by peers anywhere else in the world. The six faculties of Dhaka University now produce six half-yearly English journals and one ten-monthly Bengali journal. This is in addition to ten department-produced journals. The six university journals are offshoot of "Dhaka University Studies" that begun in 1950s while "Journal of Statistical Research," for example, started in the 1970s. None of these journals is indexed in any database. They do not have enough issues a year nor have focus, depth, and quality in any one field of interest and as such, nothing therein stands a chance of being cited by anyone except through self-citation. It seems that all this has taken an epidemic proportion – each Bangladeshi university, public and private, has one or more similar journals.
The symptom of this nationwide epidemic is apparent from the latest listing of the Bangladesh Journals Online (Bangla JOL). As of December 14, 2011, a total of 73 journals are included in Bangla JOL with five (including a "Stamford Journal of Microbiology") joining since November 7, 2011. This alphabetized Bangla JOL listing begins with "Anwer Khan Modern Medical College Journal" and ends with "University Journal of Zoology, Rajshahi University." Of the 73 that are currently included in Bangla JOL, as well as those countless others that haven't yet joined Bangla JOL, only 8 journals, i.e., less than 11%, are included within the top 18,854 journals in the world.
Table 4 lists the Bangladesh-based journals, their relative measures as identified recently by SCImago and the years since their inceptions. The value of a journal and manuscripts included therein is gauged by the number of other researchers who would make use of it in their follow-up research. Each journal is thus measured and ranked by SCImagoin terms of SJR, indicative of the scientific influence of its average paper. SJR ranges from a maximum score of 63,545 and a minimum score of 0. SCImago also calculates two of the most important and widely-accepted measures: H-index and impact factor (IF). Both measures have a high degree of correlation between them. The H-index, in particular, expresses the journal's scientific productivity by number of articles (H) that have received at least H citations. IF, on the other hand, indicates average citations per manuscript over two prior years.
The highest ranked journal (3581st amongst 18,854 in the world) from Bangladesh happens to be "Journal of Health, Population and Nutrition," published by the International Center for Diarrheal Research Bangladesh, now in its 29th year of publication. Its H-index of 23 indicates that 23 of its papers have been cited by at least 23 researchers. On average, 58% of papers published in this journal during the previous two years were cited by one or more researchers. In comparison, papers published over the previous two years, for example, in "DU Journal of Pharmaceutical Sciences" (ranked 14207th) and "Bangladesh J. of Obstetrics and Gynecology" (ranked 17139th), respectively in their 9th and 24th years, wasn't cited at all.
It is illustrative to note that the other 65 Bangla JOL journals as well as all others from Bangladesh do not contribute too much of anything as far as scientific impact is concerned. That long list of non-impacting journals happens to also include the "Journal of Bangladesh Academy of Sciences," now in its 35th year of publication.
Will people continue to go for meaningless low-impacting journals? Yes, they will as long as the system that controls their appointment, promotion and retention decisions will not be bold enough to say "No." That requires courage and can be pushed by only those who have broken free of these questionable outlets and practices and have set examples already by subjecting their own works to real scrutiny of the global research community.
While there are many naysayers who may argue that it's impossible or too late for Bangladesh to turn around the boat, I am reluctant to accept it. For the last 14 years or so, I have been fortunate to be a witness to the possibility of a knowledge creation sector of Bangladesh. This happens to be in computing and information technology and is being driven by talents from within Bangladesh with assistance from their colleagues overseas.
The International Conference on Computer and Information Technology (ICCIT) has transitioned itself to now become the largest scientific conference held in Bangladesh. It attracts typically 400 submissions a year and has an acceptance rate approaching 35%, indicative of high quality. Each of its articles is indexed by the IEEE Xplore, the largest research database containing over two million records. Since 2008, over 30 of these accepted manuscripts now get selected each year to be further enhanced which upon additional rounds of reviews appears in international journals.
I cite ICCIT as a game-changer since it gets to engage the most number of universities in Bangladesh. Secondly, ICCIT doesn't end with just a conference but is indexed worldwide and transitions some of its better works into journal articles. How significant is this? If it can be sustained, this could in 5 years contribute to an additional 150 cited works for Bangladesh – that's roughly 20% of what got reported for Dhaka University or BUET.
As a proud alumnus of Dhaka University, I do have a simple dream. This is an extension of a poster I saw recently in Dhaka: We would rather not be ostriches and bury our heads in the sand. The mediocrity that has plagued both the University Grants Commission and DU now for some 38 years and in turn has compromised the universities at large will be all swept away by a tidal wave.
The dream gets a bit specific. By 2021, before Dhaka University reaches its 100th founding anniversary, it will free itself from (a) the clutches of the Dhaka University Ordinance of 1973; (b) vagaries of national politics that need to stay outside its doors; (c) the shame of its internal journals; and (d) the practice of appointment, promotion, and retention except on the basis of a quantifiable record of scholarly research and teaching.
Is it too much to dream?
—————————-
Mohammad Ataul Karim is Vice President for Research of Old Dominion University, Virginia, USA.