In his blog, Jay Cross points to relative old, but interesting article on informal learning: Informal Learning and the Transfer of Learning: How Managers Develop Proficiency and notes, "If you're still relying on formal training to develop managers, you might want to give this one a read."
The paper is more interesting for what it omits, rather than what it purpose seems to be. That is, it seems to tout the importance of informal learning, rather than from what I see is the real key finding — its the mix that matters. For example, it makes several comments along these lines:
"Because skills learned informally are likely to share similar features with transfer tasks in terms of context and content, the potential exists for skills learned informally to be more readily transferred than skills learned in formal training contexts."
"Our study suggested that managers learn mostly from informal learning, that proficiency is the product of informal learning, and that metacognitive knowledge and self-regulation skills moderate informal learning and the transfer process.
In the paper they show the following chart (p377):
When referring to the chart they note, "The distribution indicates that managers reported learning all twenty skills predominantly from informal learning activities." Yes, while the managers believed they learnt more from informal learning, the chart actually seems to be showing that they learn a core base from formal instruction, and then they build from their proficiency from there. In addition, some core skills only need a drop of formal learning to get the process going, while others require a heavy dose.
This goes back to the previous post in which I noted that some learning episodes that are strictly informal may be too narrowly based in that the learner only learns part of a task or superficial skills that may not be transferable to the job (Bell and Dale 1999).
Thus, just as we have a "blend" of learning media and processes, we also need the proper mix of formal and informal learning. This means you not only have to select the proper blend of formal learning, but also select the proper mix of formal and informal learning. In turn, you then have to select the best blend of informal learning that will help the learners transfer their skills to the job.
In the paper they mention a study in which pilots with more flight experience perform better on a simulated flight test (that is, a transfer task) than did their novice counterparts. Now I don't believe that anyone is going to argue this point, but the other part to it is that those better performing pilots would have never been able to perform in the first place if it was not for their core skills gained with formal learning.
In the comment section on my previous post Michael Hanley notes, "...the reality is that all of these exist on a 'learning continuum.'" This learning continuum is also the subject of a post by Clark Quinn: The Formal/Informal Continuum.
Thus, its not a matter of designing learning from one side of the continuum or the other because you need the core skills from one side and the proficiency of actually being able to put those skills into practice from the other side. In addition you need that mix from the middle that is not readily identifiable as either formal or informal.