Beyond its academic dimensions, heterogeneity poses serious problems to systems attempting to classify the universities for research, ranking or public policy purposes. Chile is a good example. The first attempt to classify national universities – based on selectivity, size, prestige and nature (public or private) – resulted in eight categories. Despite some of its merits, this classification was criticised on conceptual and practical grounds, including the fact that the categories were not exclusive ones.
Other observers have tried to classify Chilean universities, using selectivity and annual publications as primary criteria, and the number of students and the years of accreditation granted to the institution as secondary criteria. They described seven categories of institutions – some improvements over the previous ones. However, this classification was also flawed on several accounts, including the use of selectivity as a main criterion. For example, one category listed selective research universities, while another group was described as non-selective, teaching, large-size and low-accreditation institutions.
A recent approach faced the challenge of classifying Chilean universities – using as main criteria the existence and number of accredited doctoral programmes and the annual number of internationally indexed publications. Applying the first criterion, the universities were divided into two groups: (1) without accredited doctoral programmes; and (2) with doctoral programmes. Then, the former were further divided, according to the number of publications, into two categories: (1a) with fewer than 20 annual publications; and (1b) with 20 or more annual publications.
The first category (1a) was named ‘teaching university’ and comprised 23 institutions. The second, called ‘teaching university with limited research’ (1b), included 11 universities. In turn, the universities with accredited PhD programmes were divided into two categories: (2a) those with up to five programmes, and (2b) those with more than five doctoral programmes.
The first category (2a) was called ‘university with research and doctoral programmes in selected areas’, and 11 institutions met this criterion. The second (2b) was named ‘research and doctoral programmes university’ and comprised six universities. As expected, the four categories had marked differences in the mean values of the variables used as ‘primary classification criteria’.
Thus, the teaching university group (1a) averaged four publications per year; the teaching university with research projection group (1b) averaged 41 publications per year; the ‘university with research and doctoral programmes in selected areas’ group (2a) averaged 94 annual publications; and the ‘research and doctoral programmes university’ group (2b) averaged 636 publications per year.
In turn, while the average number of doctoral programmes was 2.2 in the group of ‘university with research and doctoral programmes in selected areas’ (2a), it averaged 18.5 in the group of ‘research and doctoral programmes university’ (2b). Consequently, the primary classification criteria had successfully grouped Chilean universities in markedly different categories.
Particularly striking was the tenfold difference in the number of publications observed, between the two ‘teaching universities’ categories – indicating that in this aspect the category ‘teaching university’ (1a) is indeed quite different from its ‘teaching university with research projection’ (1b) partner. On the other hand, this difference implies that in approximately 30% of the Chilean universities practically no research is conducted.
The four categories were also compared in terms of the values of institutional size and academic performance (accreditation) – unrelated to the publications and doctoral programmes indicators used to define the four categories. The statistical significance of variations in mean values between categories was tested using a one-way analysis of variance. This test provides a method to establish whether or not the means of several groups are statistically different.
The analysis of variance test was complemented with post hoc tests, which do establish more specifically means that were significantly different from each other. Results indicated a major diversity in mean values in most of the indicators explored, including: number of students, number of faculties, percentage of faculties with advanced degrees, number of faculties per study programme, percentage of accredited study programmes and years of institutional accreditation. The main differentiations were found between the ‘teaching university’ (1a) and the ‘research and doctoral programmes university’ (2b) categories, with mean values of the other two categories falling in between.
The categories defined by the new classification are associated with basic institutional characteristics and academic performances. Thus, for comparison purposes the institutions included within a given category could be considered to be ‘academic peers’. The latter seems a relevant point, since most of the available comparative studies – including national and international university rankings – generally overlook this aspect. From this perspective, it is unfortunate that the research universities, especially those considered to be ‘world class’, have become the paradigm of academic quality. While recognising the need for any country to have a ‘critical mass’ of those institutions, from the standpoint of diversity and their intrinsic value, the only paradigm that a university should have is the best institution within its own category. The new classification used for Chilean universities can be applied in other countries, with some adaptation to local realities. For example, other cut-off points for annual publications or number of doctoral programmes accredited by a national agency could be used.
The new classification might also provide an overall diagnosis of a system, in terms of the percentage of teaching and research institutions present. In university systems diversity represents a value in itself, since it implies, both for the students and the faculty, more options to decide where to study or work. When classifying and comparing universities, particularly in developing systems, all classifications do freeze in time what are essentially dynamic situations. In the future, many institutions will reform their category, as research activities expand and new postgraduate programmes are created. By the same token, faithful to their missions, many other universities will remain in the same category, while improving their academic performance.
Ultimately, in the academic world what really counts is coherence between mission, human and financial resources and the will to achieve the highest possible quality standards. Thus, it is crucial to classify universities properly.
* Claudia Reyes is executive director of Red Universitaria Cruz del Sur in Santiago, Chile; Pedro Rosso is rector emeritus and professor of paediatrics at Pontificia Universidad Católica de Chile, in Santiago, Chile. email: Claudia Reyes and Pedro Rosso: email@example.com. This is an edited version of the article “A New Approach for Classifying Chilean Universities” published in International Higher Education, No 67, Spring, 2012. http://www.bc.edu/research/cihe/.