A problem with assessment in super-selective institutions

“As argued in Chapter 1, good teaching narrows the initial gap between Robert and Susan therefore producing a smaller spread of final grades than that predicted by the initial spread of ability. The distribution of results after good teaching should not be bell shaped but skewed, with high scores more frequent than low scores. At university level there is therefore every reason not to expect a bell curve distribution of assessment results in our classes.” Biggs & Tang (2011) Teaching for Quality Learning at University (4th edition), p.200

In a super-selective university this is even more so. If we assume a high quality intake, with very narrow spread of capabilities, then the eventual attainment spread should be extremely narrow. When we look at a student who achieved 65% (student 1) and compare them to a student who achieved 80% (student 2) in reality that difference might mean very little. The difference might simply be the product of entirely extraneous variables, random events (student 1 having a cold during exam week).

Unless we can demonstrate a difference in kind between the high achiever and the slightly lower achiever, this is meaningless. It might be (and I think I see this happening) that academics invest much into the identification and application of those differences in kind – “student 2 really got it, they have become a proper philosopher/physicist/economist”.

“The categories of honours (first class, upper second, lower second) originally suggested qualities that students’ work should manifest: a first was qualitatively different from an upper second, it was not simply that the first got more sums right.” ibid. p.210

But that then is also open to subjective biases. Biggs and Tang don’t really seem to have an answer to this. But they are very much entrapped by their strict adherence to definitive “intended learning outcomes” within the system of constructive alignment. Hussey and Smith’s alternative combination of ILOs and “emergent learning outcomes” within an “articulated curriculum” leaves room for student creative input, risk taking, genuine innovation, individuation and other (possibly) less determinate characteristics of learning as research/innovation/creativity. As such, the curriculum offers opportunities for more significant and transformative student input, and consequently aspects of student transformation-through-learning that can be meaningfully assessed and reported upon. Having experienced such learning activities, and achieved unforeseeable outcomes, the student is more likely to value and build upon their success. Thus the learning itself, and the transformation being evaluated, is a more reliable indicator of the student’s future capabilities. And that IS what we are looking for when we assess students in the university.

“The extent to which emergent learning outcomes (ELOs) contribute to the achievement of intended learning outcomes (ILOs) varies. Some emergent outcomes are relatively close to the intended learning outcomes and can be perceived to contribute directly towards their achievement. The contribution of others is less direct, being capable of inclusion on the basis of their contribution to the student’s knowledge of the subject in general, whilst the contribution of other emergent learning outcomes is to the field of studies in general and might be included on those terms. Yet other ELOs contribute to the overall development of the students as autonomous, self-managing learners, far beyond the field of study.” Hussey & Smith (2003) “The Uses of Learning Outcomes”, Teaching in Higher Education, Vol. 8, No. 3, 2003, pp. 357–368.