Make your own free website on Tripod.com


« July 2017 »
S M T W T F S
1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31


Some URLs
Main Home Page
My Research Journal
Thursday, 8 November 2007
Interesting update to the LP software :)
Mood:  a-ok
Now Playing: Heart and Soul (The Monkees)
Topic: LP Software

So, I bumped into Shannon and Joanna from Geography during lunch today and I was telling them about my research and the problem I was having with the students who are not recognising that some constraints are binding ... i.e. for example when the labour hours has finished but lumber feet has increased that students still think they can build chairs because they're not looking at the whole problem and connecting it but rather just part and then relating it to real life. Joanna then said this reminded her of the machine in the museum which was used to teach capital flows etc. in economics by using pumps and water tubes, and checking to see what happens if they put a pressure (such as increase taxes) what would happen to the capital flow.

This got me thinking that perhaps students might benefit (but obviously not something for me to do in this study but perhaps as a recommendation) - an animation or graph that shows how much resources they have remaining - i think an interactive graph/animation might be interesting - one where the resources are being increased and what happens to production etc. Just a thought :D.


Posted by prejudice at 1:34 PM GMT
Thursday, 11 October 2007
Trying to figure out my design!
Mood:  irritated
Now Playing: the noises of the house and the housemates
Topic: LP Software

I'm trying to figure out why I made the glass-box the way it is ... that is, why I don't show all the steps rather than ask them to click iteration for each ... and for the life of me I can't remember why. I've checked my meeting notes but not able to see anything there or either in this blog. I think the reason it's because I inherited the style from the expected-values pilot study, since I did it that way for each section in the expected values.

Well, so in fact the glass-box may not be completely glass-box -well its in a sense, but it doesn't show all the steps with one click, you've got to do several clicks to get all the steps. I felt and I vaguely remember I did it this way so, it would be somewhat mid-way between the OB and the BB. I think I had two options, I could have made it more BB looking (i.e. one click and just all the steps shown) or more OB looking (click iteration for each of the steps to appear). I guess at the time I was thinking about the amount of interaction the student had withe the software and with GB they had more interaction. So, is it really GB? I think it is, it's GB with interaction. Perhaps in future studies, one can look at GB with and without interaction.

Now, why did I choose GB with interaction rather than GB without interaction ... this is what I'm uncertain of. I know I wanted it to be closer to OB but why? Well, I had two choices really, and I wanted to end up with a Latin square design ... so only could do three anyhow. I think I wanted to look at whether the level of interaction would affect how the participant uses the software and as I had already drop one level of interaction in the OB, this was a compromise situation. In the GB with interactions, calculations were shown at each step and still maintains the phenomena of a GB. I don't know if that is a strong case - but there it is. So, we really have three levels: no interaction (BB), some interaction (GB) and lots of interaction (OB). So, it is somewhat on a scale. We can also have it has no steps (BB); steps shown in a modular fashion (GB); and steps determined by user and shown in a modular fashion (OB).


Posted by prejudice at 9:05 PM BST
Friday, 16 March 2007
Excel VBA Programming
Mood:  d'oh
Now Playing: Jab Tu Muskarati Hai (Koi Mere Dil Se Pooche)
Topic: LP Software

Well, I'm still working on the Excel spreadsheet programming - taking a bit longer than I expected. Technically speaking it is finished - however, need to make it attractive looking and easy for students to enter the data and prepare the sheets for the different problems (still go to do that) and then test these out - and I need to do this ASAP!!! My timeline was to finish by next Friday (but don't think I'll be able to) .. probably the two days I have here when I come back from Dublin I can do something - hope so. That means setting up the appointments and making sure they work. I think I will need at least need three persons, one for each software.

Well, will see how it goes.


Posted by prejudice at 5:07 PM BST
Thursday, 10 February 2005
Lindo and its tableau
Mood:  a-ok
Now Playing: Kabhi Kabhi (one of my favorite songs!)
Topic: LP Software
I'm thinking that probably for the tableau when it comes to problems with excess and artificial variables, that the way to treat this is to first convert the problem into a two-phase problem, with the new objective function ...

...Just did that ... found out it doesn't work, since there is a number in the objective function, and all numbers in the objective function must be associated with a variable. Oh well, not sure what method it is using to solve these problems, but I did the tableau and pivot actions several time, and although when I am doing it, it is claiming the solution is infeasible, it eventually reaches an optimal solution. I wonder if this is the interior point algorithm ... I don't think I have ever done that algorithm - so, probably why I don't understand what Lindo is doing.

Posted by prejudice at 11:07 AM GMT
Wednesday, 9 February 2005
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?



Posted by prejudice at 4:08 PM GMT

Newer | Latest | Older