Week 8

By wiesel |

Saturday, September 12, 2009

oh yeah, week 8. Should probably blog then. I must say i am a little bit hung over so apologies for any increase in ridiculous comments.

More aroma. I'm getting used to Visual Studio BI stuff more now. I got the hierarchy the wrong way around in the tute. I'm not sure if i was dyslexia or me thinking the other way around. It did have me thinking about whether i was understand or just following directions though. I got that working after some help from Steve. I hadn't saved the database file along with my other aroma stuff so i just downloaded the aroma file from moodle to save me time, lol. You gotta love how you don't have to do the tut. before to do this tut. it's very convenient.

Lecture was about Business Performance Measurement and it was titled Business Performance Measurement (I) which means next week at least will be about this too. I noticed you broke the diagram (pictured), all the words are outside their shapes. You don't win the noble prize for BI with that quality of diagram. I put that in here to remind you to change it. Also remember to change the transition on slide 15 too. Unbelievably this week, all my lectures had probability and/or stats in them. I am so bored of normal distributions right now. If week 8 wasn't hard enough, without having the same thing told to you in varying degrees of confusing ways that you have understood since primary school or at least 11th grade. I mean (no pun intended) seriously, so many people don't know the difference between mean, median and mode; it's shocking. And standard deviation too, it's so easy to understand. What's worse than that is somebody teaching it badly, not you POD, but other lectures of the methods of research for IT (like what i did there?), go in to the deep dark depths of Bayesian theory to describe relatively simple things. end rant.

most thought provoking idea came at about 9:55 in the lecture "There is no conclusive experimental evidence to verify that the use of graphics improves a decision maker's understanding of data" which i can agree with. But understanding is different to speed or other characteristics which may be beneficial to the user. Especially because every user learns differently, i'm a bit ahead of myself, cos i know a bit about VAKT from when POD talked about it in IE. I can see arguments for graphs even if they don't aid understanding as much as others. Prettiness is probably underrated academically, but some people will reject a system if it doesn't look the way they expect it to look, not to mention all the ideas of HCI to do with signals that appear to be similar will likely be confused; tables look quite similar to each other, but by making different graphs we can differentiate the information domains. There are a thousand other ideas that I have like that we a conditioned to like graphs from school. But i think the biggest idea that is for graphs is the speed of use. If i am used to a graph i can tell what is happening in a split second. I don't think i could do it as quickly with a table. I think the amount of information that you can take in visually with a graph in an amount of time is much greater than in a table. I think a person would remember better too. I also want to ask questions about how the studies were done. If the question that needs understand is something simple like 'what were sales yesterday?' then a simple number is obviously the best graphic. but for questions such as 'how did sales vary over the three weeks?' i think graphs could be better than plain numbers.

I have been spending way too much time thinking about this clearly, i mean it's as good a thing to think about as anything, just i didn't need to. In conclusion i'm looking forward to the explanation of this in week 9, see you then.