This is not simple, mainly on the analyzing of the data. Ok, what I want to do is something like this: Collect data, analyzing data, and conclude. I will do the last 2 parts, and you guys will just help me to collect data, I need as much as possible.
1.) What data to collect?(tentatively)
Year of admission, department admitted, UEC full results, SPM full results(if applicable), first choice at the time of application.
2.) How to collect?
I need accurate data. If the person cannot remember exactly his results, don't take his data. And, I need digital data=). I dun care how you get the data, just key it into database, which i think it will be convenient if the database is in our server...But I have to get opinion from Kian Zhong la.
3.) From who these data from?
For anyone who is admitted into NUS using UEC results. Thats mean majority of them would be AMCISA members. But if you know someone who is not our member, but he/she use UEC result to apply, you can ask him/her to help.
4.) Summarizing data
My job too. Basically just find a best way to summarize all the data collected. Maybe one or few graphs, or any other means necessary.
It may or may not be enough until steps 4. I'd prefer to continue until steps 6(below), but it also seems to me that, it's enough to just present the summary of the data, and let those who read the data interpret the data as they wish.
Below is what I will be doing if further analysis is needed.
5.) How to do the analyze?
To be frank I dun have a clear way to do it. But it's more or less playing with the data, and try to figure out the relationship. I will not go into details of the analysing, but if you want to know how will i do the analysis, feel free to ask me, here or elsewhere.
6.) Conclusion of the results
Related to the analyzing. Please be noted that, this might be a failed survey. It may be cause by too little data collected, or there is actually no relationship between the data.
So, 4/6 steps end our survey, depends on what you guy think. I'd prefer 6 steps one, as summary of data always misleading. But another problem comes in if you do the analysis: you might give wrong interpretation of the data...So, ALOT EXTRA WORKS will be needed if you wan to do the analyzing.
Think, and question me about this survey. You can question anything about this survey, including the way I wan this survey to be. Really, anything, just question. I will answer.
P/S: Statistics is fun=) taking experimental physics now, and im going to use alot of stats!=)))
Just to remind myself: Night cycling march 6, next meeting 11th feb
okay a =)
ReplyDelete