[quoting Phillips' paper] Then, I would like you, the reader, to contribute your reflections on the problems you have experienced in your own projects.
I suggest that another source of inspiration about difficulties is looking at novice design efforts. For instance, in teaching my students "user-centered design" (which shares considerable overlap with "learner-centered design"), I reliably see several problems such as "functional fixedness" and over-confidence. The functional fixedness manifests itself in products that look like what the implementation tool does well, not in a good tool-independent solution to the problem. The over-confidence appears as projects informed by intuition, not exploration.
However, I'll talk about a particular project I was involved in that was less than ideal (out of my control, of course).
I was contacted by a guy had some funds to develop a CAL project in a technical area (he's apparently had success before in this as well). He'd hired a person from my alma mater to serve as the actual programmer. The pedagogical problem was identified as transfer from theory to practical labs: students had trouble using equations to model the observed performance which was less than the ideal predicted by theory. This led to a proposed design (not detailed). The pedagogical goal shifted some time after that (I ended up being called in only when the programmer was having difficulties with design) to providing a model and some exercises to help students acquire a difficult concept (that was one of the elements that was difficult to use in the equation modeling). The main investigator stayed away as much as possible, and had the programmer spending time writing up conference papers about the project, as well as focusing on bells and whistles (where's the sound? the video?). Fortunately, the programmer had a cognitive background and was eager to learn about good learning design.
It was insufficiently planned and needed more specification before implementation. It did finish. It was going to be evaluated by a Ph.D. student, last I heard. The programmer left to do a Masters. I'm mostly off campus now, so I've lost touch.
It looks pretty good. They got some graphic design help. I think it's not horrible, but only by means of some fairly serious revisions at the periodic reviews. The whole thing needed better management.
Reflections on another project indicate problems from handing off to some others to change graphics who were not involved in the design, and they made other changes inconsistent with the overall design goals.