Supervisory meeting

Mood:  lazy
Now Playing: song from Mohabbatein
Topic: LP Software

I had my supervisory meeting on Monday, it seems as if my research might go somewhere yet … we are thinking about correlating lecturer’s view of software with that of the approaches to study view of students … seems like fun … but got to unpack my computational and learning tool meanings at the moment.

This morning I demonstrating three software packages to James (Doug called in sick) – Lindo, WinQBS and Solver. James pointed something out to me how specialized some of the terms were like such as iterations, tableau, simplex algorithm, reduced cost, dual price, sensitivity analysis, shadow price etc. I had never thought about that, and I can see where problems can arise. James pointed out that depending on your background these concepts might be easier, for example people with linear algebra, may find the concept of elementary row operations (eros) much easier to comprehend … and possibly people in economics etc. may find shadow prices something easier.

I discover this morning that Lindo has a tableau function in which you can view the iterations, and even do them yourself, but it doesn’t seem to work with equations requiring the dual simplex algorithm, I think in that case, you may have to put in the new objective function. This is better than the WinQBS I think, for the simple reason, that you make the decisions, while WinQBS helps you along with the answer, by highlighting the next pivot variable and row (although you can choose to change the pivot variable) … may be there is a way to switch that off. Plus WinQBS allows you to put the variable in as text (like Lindo) or as a matrix (Solver). However, solver is nicer since it allows exploration with variables, i.e. changing the values of the changing variables or using even goalseek to see there immediate impact.

James gave me an idea though of probably using other lecturers from outside of the UK, as that might increase response in the survey, and we can do that as a web-survey … only problem with that, is that it does not link with the students … probably, I can choose a select set of lecturers and use their students, so a smaller sub-set of the study, in other words. I’ll have to develop the criteria for why I chose those lecturers … I don’t think it is good enough to say because they were interesting.

That reminds me … I think my literature review may have to contain a section expounding the good and bad between LSI and ASI. Also, have to learn some lingo … do students learn?? Or is knowledge transferred?