A Tale of Two Donalds

It’s a week since I posted a comment on Donald Clark’s blog, but comments there require moderation, and so far, at least, Donald hasn’t published my comment. As I mentioned it in my last post here, this is what I said:

Insightful critique, Donald - I found myself nodding enthusiastically at many of your points. But I find your conclusions a tad harsh. The main problem with Kirkpatrick is too many people regard it as "the only game in town", when it is one (just one) of a range of models and tools that have their place in different situations.

Brinkerhoff's Success Case Method (for example) is brilliant, but I wouldn't recommend it for every situation. The same with Kirkpatrick. I don't think we have to dismiss it altogether, just put it in the toolbox alongside all the other options. And I'm not convinced we need a single new all-embracing model to replace it (not sure you are advocating that, but that's one reading of an "overhaul").

Further discussion of situational approaches here: http://www.airthrey.com/papers.htm

On checking back to Donald’s blog, I found he posted an even more devastating critique of Kirkpatrick, back in 2006 – Donald Talks Bollocks! Perhaps Donald Clark’s views have moderated a little since then? I think the problem is too many people are looking at evaluation through the prism of Kirkpatrick, and perhaps that means they need the jolt of this sort of polemic.

Further thoughts:

1. "Happy sheets" shouldn’t be primarily to test how happy learners were: they are a simple quality test, an opportunity to check that the basics of a learning initiative are working – e.g., the trainer is effective, the course materials are relevant and comprehensible, etc. I realise this example is specific to a training course – arguably the sort of application where Kirkpatrick works best.

2. I don’t agree that Kirkpatrick asks “all the wrong questions”, but I think there are plenty of other questions, and many situations where other questions are more pertinent.
3. I don’t want to seem like an apologist for Kirkpatrick, but I think his model still has its place, as long as it’s deliberately selected as the most relevant model, as an informed choice over other models. It would rarely be my first preference, but I can still see it can work for some of the people some of the time.