Learning Outcomes – What have they ever done for us?

We have a monthly pedagogical research group meeting  and the topic for December was learning outcomes (LOs). We based our discussion on three documents and this was followed by a presentation by Dr Bob Norman on the research he’s doing into learning outcomes. The three documents were:

  1. A journal article by Trevor Hussey and Patrick Smith
  2. An article in the Times Higher Education Supplement by Professor Frank Furedi, and
  3. A guide to writing and using good learning outcomes by David Baume (Leeds Metropolitan University).

The Hussey and Smith article divides learning outcomes into three broad categories: session outcomes, module outcomes and outcomes for a degree programme as a whole, and argues that session outcomes are the most pedagogically useful and programme outcomes actually represent a misuse of the term learning outcome. A common criticism of learning outcomes is their misuse (potential or real) as auditing or management tools. Hussey and Smith state that learning outcomes have been hijacked by managers as an auditing tool. But for hijacking to occur requires that learning outcomes had a pedagogical purpose in the first place. So what purpose do learning objectives serve? Their purposes fall into two categories – learning outcomes as audit tools and learning outcomes as teaching tools. I’d argue that these are two fundamentally different things rather than two categories of a single concept.

So, what purpose do they serve? For example, do learning outcomes drive student behaviour? Surveys of student attitudes to learning outcomes suggest not, finding that most use of LOs was at the summative assessment stage (i.e. as a revision aid) rather than using them to guide their studying throughout the module, and certainly not in any sense as an aid to reflective practice and their development as independent learners.

Alternatively, are they simply statements of curriculum or what the tutor is going to cover? This is obviously useful to the student while studying the module and also, if they are made more publicly available, when choosing a degree, but that brings us back to the role of LOs as being something other than a pedagogic tool. And do a series of learning outcomes really describe the richness of the educational experience we’re trying to give our students?

Early on in the discussion someone asked a pertinent question: what evidence is there that learning outcomes have any impact on educational outcomes? In other words: do they actually work? The question was followed by silence, since none of us knew of any research supporting the use of learning outcomes. In one way, that’s not surprising since here in the UK, the push for learning outcomes didn’t arise from any pedagogical need but were imposed as part of a quality assurance agenda and consequently resistance to their use (as shown in the Furedi article) becomes part of the wider reluctance to accept the creeping managerialism of the higher education sector. And as Hussey points out in a comment replying to the Furedi article:

“First, few people dispute the need for a teacher to tell his/her students roughly what to expect from a teaching session, or what, broadly, will be the contents of a course of study. Secondly, doing this is not the same thing as stipulating supposedly precise learning outcomes, written to a strict formula.”

In some ways, this is similar to debates and policy around learning styles. For example, saying that people learn differently at different times or in different contexts isn’t all that contentious, but a simple, imprecise, rule-of-thumb like this then gets extended to the stage where secondary school teachers place cards on little Freddy’s desk labelling him a kinaesthetic learner and are expected to tailor their teaching to that particular style. My standard response to that is ‘Great, now let’s see you teach them to code kinaesthetically – Yay! Dance that subroutine!’ The original (and useful) general awareness of personal differences has been subverted into an inflexible process that is actually counter-productive. In a similar way, the general usefulness of letting students know where they’re going has been subverted into a series of audit checkpoints better suited to accountability to management than usefulness to the students.

An important point about learning outcomes written beforehand is that they are intended learning outcomes. Another category of outcomes, often neglected in discussions of learning outcomes from an audit perspective, is emergent outcomes. In other words, those outcomes that arise from the interaction of the student and tutor and that (by definition) can’t be planned in advance. MacAlpine et al (1999) talk about a ‘corridor of tolerance’ where the corridor represents the level of diversion and divergence from the planned outcomes that is acceptable in a given context. Hussey and Smith (2008) argue that these should be teacher decisions and I strongly agree. Isn’t that what teaching is all about? – the interaction that leads to the ‘aha’ moment or the diversion that leads to an intense interest in something only tangentially related to the teaching session (and something that’s probably not going to be ‘on the test’). My first degree was in Marine Biology and Zoology. I forget what the trigger was but I remember being fascinated by wolf ecology and I’d check the journals room for the latest issues that were likely to have relevant research in them (this was in the days before electronic journals and abstract databases were widely available). I was like a child waiting for next week’s comic, with the result that wolves are an animal that continue to fascinate me to this day.

Furedi talks about learning outcomes devaluing the art of teaching because in their most restrictive form they constrict what can be taught. Hussey and Smith in an earlier paper (2002) make the point that all learning outcomes are objective and trying to make them subjecting by adding more detail only makes them objective at a different level. The more we try to make them subjective the more restrictive they become, and the pedagogy undergoes a selective pressure towards didactic teaching rather than constructivist practices (but of course, the more subjective they are the better they suit their role as auditing and management tools). Refreshingly, the Baume document, although written as an institutional guide to writing learning outcomes, explicitly acknowledges (p7):

“… learning isn’t the tidy process that the use of learning outcomes may sometimes lead us to think. Unexpected, serendipitous, learning happens. Such learning is also worth making space for, recording and valuing.”

Since I was expecting this document to approach learning outcomes from the management perspective (because it’s an institutional document) this was a welcome nod to teaching as the messy and creative process it often is.

Learning outcomes don’t have to be prescriptive though. Gabriel Egan in the comments to the Furedi article responds to Furedi’s arguments against learning outcomes and shows how learning outcomes can be written in such a way that the encompasses the organic messiness of real teaching:

“1) Demonstrate that you have acquired knowledge of, and can articulate fluently (in forms as yet undecided), aspects of the topic that cannot be predicted beforehand and that are as yet just as unknown to us, the tutors, as they are to you the student.

2) Describe and explain your engagement with the uncertainty of academic exploration of this topic and show your skill in articulating the indeterminacy of the pedagogic outcomes that arise from the topic’s inherent complexity and the subtlety of the tutor-student relationship.”

Now these are my kind of learning outcomes, but whether these learning outcomes would make it past institutional validation processes is another matter 🙂

Another question is how do learning outcomes link to assessment? Constructive alignment (Biggs, 1999) matches the teaching and learning activities to the assessment to be used, but Hussey and Smith point out that at the session level the outcomes may be too small, too granular to be assessed directly, and that they may need to build upon each other and be subject to student practice before reaching a point where they can actually form something to be assessed. This shouldn’t be surprising. If we’re aiming to develop and then assess the higher order skills then it’s unlikely a tick-box approach to the outcomes of a session would be appropriate.

Constructive alignment brings us on to a function of learning outcomes that I haven’t really seen acknowledged much, that of outcomes as a design tool. The choice of outcome has a direct impact on the learning and teaching activities and subsequent assessment. For example ‘… will demonstrate the characteristics of …’ implies a very different course to ‘… can recall the principles of …’. That’s one way I’ve approached learning design in the past – matching learning outcomes to a grid of teaching activities/technologies and a grid of assessment techniques. I wouldn’t argue it’s the best approach to learning design, but it’s a quick and dirty way to get something up and running with a fair likelihood of being effective pedagogically, and it helps us to think a little more reflectively about our teaching practice. The downside to outcomes as a design tool is that we operate within institutional constraints that we may have little or no influence over e.g. we have X contact hours in room Y.

So where do I stand on this? Learning outcomes are useful as a pedagogical tool rather than an auditing tool. They work best at the session and module level, and as a rough and ready design tool, but only if they reflect the messy, creative and sometimes chaotic process that good teaching can be. The endpoint can be (flexibly) planned, but the journey should be a voyage of discovery in good company, not guided by a robotic sat nav that leads us over a cliff.


Biggs, J. (1999). Teaching for quality learning at University. Buckingham: Society for Research into Higher Education and Open University.

Hussey, T. and Smith, P. (2002). The trouble with learning outcomes. Active Learning in Higher Education, 3(3), 220–223.

Hussey, T. and Smith, P. (2008). Learning Outcomes: a Conceptual Analysis. Teaching in Higher Education, 13(1), 107–115.

MacAlpine, I., Weston, C., Beauchamp, C., Wiseman, J. and Beauchamp, J. (1999) Building a metacognitive model of reflection. Higher Education, 37, 105–131.