Monday 7 December 2015

Visualising Constraint in Educational Institutions with Parallel Coordinates

I'm preparing a presentation for the SRHE conference later this week about status functions and constraints in institutions. Status functions are Searle's idea for how social reality is manifested through particular kinds of declarative speech act - i.e. "This is a University/certificate/banknote...etc". The argument of the paper is that they can be analysed by looking at institutional strategies (these are, in the end, collections of status functions), technologies, league tables and so on, and each of these status functions has a constraining effect on institutional life. Moreover, status function declarations exist in webs where there are many inconsistencies, contradictions and often 'knots' where competing status functions constrain each other to create a kind of stability. I think Searle is not quite right in saying that social reality results from status function declarations; but it seems reasonable to argue that social reality is certainly constrained by status functions. Maybe things like Universities, monarchies, nation states and e-portfolio(!) only exist because of the knotted constraints they tie in each of us...

Analysing institutional strategies is one way of indicating status functions, but Searle also says that status functions have to be upheld by the 'collective intentionality' of the community for which they are intended. Deep down, we agree that David Cameron is Prime Minister, and a £10 note is worth £10. We're probably less clear about education, but the knotted constraints of education including the value of certification and social and cultural capital for employability leads people to go to extraordinary lengths to ensure that their children get "the best" education, and leads young people to ask few questions about heavily indebting themselves for a degree. And that's before we consider the status functions of education itself: learning outcomes, assessments, curricula, timetables, lectures, VLEs, e-portfolios, academic papers, textbooks, and so on. We all buy into the whole thing and tend to ask few questions about something which would seem quite perverse to a Martian visitor!

Whilst the status functions of strategies and league tables are there to be seen, seeing the 'collective intentionality' requires that we ask people about it. The paper reports on some work I did with colleagues at the Far Eastern Federal University (FEFU) in Vladivostok. For Russian universities, the status functions of league tables and prestigious journals are particularly constraining because they are all in English, and dominated by a Western European/Anglo-Saxon academic culture. Russian academic traditions are different, yet the status functions are made by Western publishers like the Times Higher Educational Supplement. The Russians have to buy into it, but it's hardly a level playing field (the same argument applies to non-research universities in the UK, of course).

The problem is that things like the QS rankings constrain the strategic decision-making within institutions, where managers feel that their institution should be doing all it can to raise its league table position. This can put teachers in an impossible position.

So at FEFU we asked teachers to rank the constraints they felt prevented them from enhancing research and teaching. We then asked them to consider those things which constrained them least, and asked them what they might do to overcome the least constraining things. The data we got back was quite messy, but a visual analytic approach helped to identify some patterns.

Parallel coordinates, pioneered by Alfred Inselberg (who has a cybernetics background), is a powerful technique for doing multivariate analysis in a visual way. I've used the javascript library D3 which does a nice job of making interactive parallel coordinate graphs. There are no surprises about the consensus as to what is most constraining with regard to research: the second bar along in the graph above represents "too much teaching" and the 9th represents "bureaucracy". But the strong constraints are possibly less interesting than the weak ones. The only problem with "weak constraints" is that people don't really see them as constraints; rather they might perceive a 'weak constraint' as 'irrelevant'. Quite a few teachers felt that the abilities of their students wasn't a great constraint on their research activities (indeed, I suspect they may have seen it to be an irrelevant factor). In answer to the question "What can you do to address the least significant constraint?" some responded by emphasising the possibilities of doing more interesting things in class and enthusing their students: this at least was within reach for teachers - although it doesn't seem to fit with the constraint of feeling that there was too much teaching!
What I find interesting at a broader level about this is that the graphs provide a kind of model of the collective intentionality of the staff, and that with this knowledge, institutional policies which work with, rather than work against, the prevailing collective intentionality are more likely to be successful. More importantly, modest strategies for more inventive teaching (say) or freeing up the curriculum could change the institutional constraints to the point that ways of raising the international reputation of the university can reveal themselves without directly tackling things like the QS rankings. 

Oblique strategies and effective realistic analysis of where the land lies may be far more successful in the long term. What is important is to have analytical techniques which can pinpoint areas where strategic status functions in the form of new initiatives might be created which are likely to be upheld in a constructive way by the majority of teachers.

1 comment:

Unknown said...

In the world of human endeavour. constraints need to be over-come - in the world of software development, they are seen as absolutes. My semi-automated approach to time-tabling used a mixture of (absolute) constraints and desires. Absolute constraints cannot be broken, but the desires were used to assist in point-scoring solutions.

IDEF modellling was based on 'control' signals coming from 'managing' processes. The use of the visualising software may help structure constraints into hierarchies or identify loops which are responsible for instability.