I must admit that I do not understand the internal statistical tests they describe to show the validity of their instrument. Not my area, sadly. But I did have a few thoughts about what the papers showed (and hence what their measuring instrument would show).
First, there’s a bit of ambiguity in the background. They seem to express sympathy with the Benner/Dreyfus claim that expert mastery involves progression from the procedural to the intuitive end of a spectrum and yet they also characterise the basic procedural approach as ‘logically defensible’ implying something illogical about the intuitive. I assume that this is just an infelicity in their summary. But there is a difficulty here insofar as the logical defence of a piece of intuitive judgement – for example pattern recognition of symptoms – is that the pattern is there to be seen for those with the right expertise. If one acted on intuition but were not an expert – if one guessed, in other words – that would not be the expression of greater mastery. In other words, there needs to be some quality control via the question: by whatever process the judgements was made, was it (generally) right? And sadly it seems that their instrument does not measure this because it is a self-marking system of how the participant thinks about how they make a decision not whether they are actually making the right decision, reliably.
Second despite the apparent tilt in favour of an assimilation of mastery/expertise and intuition, the questionnaire presupposes that the ‘nursing task’ falls into four stages.
The structure of the instrument was designed to reflect four main stages: (a) data collection, (b) data processing and identification of problems, (c) plans of action, and (d) implementation of plan and evaluation. This decision was based on abundant research evidence indicating that decision making in nursing comprises different stages…
This four stage articulation, however, looks to presuppose a procedural approach. On an intuitive approach, by contrast, one may see at a glance – from the symptoms to – the underlying medical emergency and hence the intervention needed. The gestalt character of intuitive knowledge does not partition into temporal stages. So there looks to be something very odd, at least, about building a test of the procedural versus intuitive nature of judgement but assuming the prior correctness of the procedural.
Ah, you might reply, even an intuitive judge can reconstruct what would have been the logic of the judgement, had they needed to, slowly after the fact. (Perhaps an expert at speed chess can reconstruct their strategy afterwards even if there would have been no time while actually playing. Perhaps this is what makes them an expert.) If so, the four stages do not prejudice the instrument. The problem with that, though, is that the two examples of questions from the instrument that they give suggest that a ‘phenomenological’ distinction can be offered by the participant of how they think of the process of judgement and this then used to place them on the spectrum.
My hunch is that the instrument measures how participants theorise about how they make judgements rather than how they make judgements. Hence the question: is the former likely to correlate with the latter? Perhaps one needs to have the capacity to think about ‘data’ and ‘hypotheses’ to be able to adopt a procedural approach at all whether or not one did adopt a procedural approach.
But then what is it that you want to measure? If Benner and Dreyfus are right, we want nurses to make skilled intuitive gestalt judgements, to see what’s going on and what is needed. But if the choice is between someone who makes generally incorrect intuitive judgements and someone who modestly follows a procedural approach then one will want the latter. So the context matters.
When I taught medical students at Warwick, I took mere awareness that values matter to decision making, that values can be hidden and that rational disagreement about values is possible to be a successful teaching/learning outcome. It was at least half the battle for students to grasp this (whether they actually went on to make sensitive collaborative value-based decisions afterwards or not). So I can imagine an equivalent.
Although expert intuitive judgement is the gold standard, in cases where that is not possible then one will want nurses who can explicitly frame hypotheses, derive contrastive predictions and use these to test the hypotheses. One test of this is: can they at least frame the idea of hypothesis testing? Do they know what ‘hypothetico-deductive’ means? Do they understand that confirmatory evidence is still compatible with an infinite number of theories? Etc etc.
But I would not test this capacity by asking them whether they confront nursing in a four stage model which starts with data collection. Frankly, that seems a weird, alienated, inhuman approach to the business of nursing.