I just came across a fascinating ph'd desertation rejoicing in the title: Framing in Educational Practices Learning Activity, Digital Technology and the Logic of Situated Action". It's by Annika Lantz-Andersson and provides, among other things, some insight into how children respond to the problems they encounter while using educational software.
An illustrative bit:
The results from this study show that in break-down situations, when the students hesitate in the interaction and when they do not get the result they expect, they end up being uncertain about how to continue and how to act in the situation. The students, thus, showed that they had difficulties in how to frame in the situation; had they themselves performed an incorrect modelling or did the difficulties lie in the digital design? Using Goffman's (1974/1986) terminology, the students experienced a conflict in how to frame the events in the situation in a relevant way. In this study, the ambiguity of how to interpret the nature of the incorrect answer runs through the entire problem-solving activity, and the students were uncertain about the framing that was relevant in order to understand the feedback they got.
When students received feedback that their answer was not correct, they were forced to reconsider. The manner in which they did this was a) to check their interpretation of the problem, that is, the modelling issue, b) to check whether the digits and calculations were in order, or c) to enter into the framing of the software and begin to consider various features of the syntax or whether the answer function could be incorrect.
In most of the excerpts in this study, the latter strategy dominated. The main result was that for long periods of time the students operated within the framing of the functionalities of the software, and, while doing this, understanding mathematics seems to play little or no role in their deliberations. This implies that many of their actions and interactions were devoted to speculation about the syntax features of the digital tool and to testing whether there was something wrong with the answer function. The students, hence, engaged in extensive meta-level talk that did not primarily take place within a mathematical framing but was, rather, geared towards considering various features of the design of the software.
One of the frustrations in doing large ERP installs is that users follow a well trodden path on these: you get some early adopters who learn a lot about the system and how to use it - but then turn-over happens and the people in place focus on what they use regularly to tell the new comers what works and what to stay away from in the ERP software. The effect, over several employee generations, is that less and less of the software gets used while resistence to change grows as some combination of manual and magic work arounds for real or perceived limitations in the software takes hold.
Where PCs are used in conjunction with a help desk this process gets dramatically accellerated as those early adopters with some emotional stake in the system's success run into whatever real problems may apply and place the blame for those first on the PC and then on the help desk - and, in that process, first exhibit the classic stockholm syndrome hate/love conflicts that go with being held hostage by both, and then rapidly lose interest in pushing the software forward.
It's not at all unusual, and largely, I think, as a consequence, to see companies using decade old ERP releases, excusing this at the IT level on the grounds that it would be too expensive to replicate the customization done during implementation, but using, in reality, only the most basic system functions while supporting complex, and often largely undocumented and unauditable, workarounds for all, or nearly all, of the processes they think differentiate them from the ERP supplier's other customers.
Read that dessertation carefully and I think you'll see that what's driving the responses she documents is a kind of supra-framing: the kids know the technology isn't reliable and prefer betting on its being wrong to reviewing their own thinking and actions. Translated to the commercial IT world of big name ERP that supports, I think, my belief that a lot of these guys are spending tens of millions reducing employee productivity to the point that the shareholders and non IT employees would be better off with a suitably scaled up QuickBooks "solution."