Beyond learning outcomes

This is a slide deck for a talk I’m giving today, at a faculty workshop, on the subject of learning outcomes.

I think that well-considered learning outcomes can be really helpful when planning and designing learning activities, especially where there is a need to assess learning. They can help keep a learning designer focused, and to remember to ensure that assessment activities actually make a positive contribution to learning. They can also be helpful to teachers while teaching, as a framework to keep them on track (if they wish to remain on track).  However, that’s about it. Learning outcomes are not useful when applied to bureaucratic ends, they are very poor descriptors of what learning actually happens, as a rule, and they are of very little (if any) use to students under most circumstances (there are exceptions – it’s a design issue, not a logical flaw).

The big point of my talk, though, is that we should be measuring what students have actually learned, not whether they have learned what we think we have taught, and that the purpose of everything we do should be to support learning, not to support bureaucracy.

I frame this in terms of the relationships between:

• what we teach (what we actually teach, not just what we think we are teaching, including stuff like attitudes, beliefs, methods of teaching, etc),
• what a student learns in the process (an individual student, not students as a whole), and
• what we assess (formally and summatively, not necessarily as part of the learning process).

There are many things that we teach that any given student will not learn, albeit that (arguably) we wouldn’t be teaching at all if learning were not happening for someone. Most students get a small subset of that. There are also many things that we teach without intentionally teaching, not all of them good or useful.

There are also very many things that students learn that we do not teach, intentionally or otherwise. In fact, it is normal for us to mandate this as part of a learning design: any mildly creative or problem-solving/inquiry-oriented activity will lead to different learning outcomes for every learner. Even in the most horribly regimented teaching contexts, students are the ones that connect everything together, and that’s always going to include a lot more than what their teachers teach.

Similarly, there are lots of things that we assess that we do not teach, even with great constructive alignment. For example, the students’ ability to string a sentence together tends to be not just a prerequisite but something that is actively graded in typical assessments.

My main points are that, though it is good to have a teaching plan (albeit that it should be flexible,  reponsive to student needs, and should accommodate serendipity)learning :

• students should be participants in planning outcomes and
• we should assess what students actually learn, not what we think we are teaching.

From a learning perspective, there’s less than no point in summatively judging what learners have not learned. However, that’s exactly what most institutions actually do. Assessment should be about how learners have positively changed, not whether they have met our demands.

This also implies that students should be participants in the planning and use of learning outcomes: they should be able to personalize their learning, and we should recognize their needs and interests. I use andragogy to frame this, because it is relatively uncontroversial, is easily understood, and doesn’t require people to change everything in their world view to become better teachers, but I could have equally used quite a large number of other models. Connectivism, Communities of Practice, and most constructivist theories, for instance, force us to similar conclusions.

I suggest that appreciative inquiry may be useful as an approach to assessment, inasmuch as the research methodology is purpose-built to bring about positive change, and its focus on success rather than failure makes sense in a learning context.

I also suggest the use of outcome mapping (and its close cousin, outcome harvesting) as a means of capturing unplanned as well as planned outcomes. I like these methods because they only look at changes, and then try to find out what led to those changes. Again, it’s about evaluation rather than judgment.