This is the crucial and all important sentence that we learned in the Methods lecture today. Clumsily translated, it means that, erm..."'Should' does not follow Being", or something like the existence of something does not mean that it should exist. Hmm...let's see if I can put it in more concrete terms. For example, if you do a survey of the pop. and the results happen to correspond to a nice theory that you have formulated, it does not mean that your theory is proven...on the contrary, it proves nothing. This is because no matter how much you try to make your selection of surveyees a fair one, it cannot cover the entire pop, and even if it does, the survey has only taken place in a certain period of time. There are just too many limitations of your expt/survey, and too many factors to consider.
The prof gave a v good example: A man went to his doctor and complained of pains. He proded himself in the shoulder, stomach and hips, "It hurts when I prod here, here, and here!" The doctor said in reply,"You have a broken finger."
And this prof actually licked the replacement batt for his mike today, and used a bamboo pole twice his height for support during the lecture (!!). Theatrics! And he loves to 'insult' the female gender, like saying things like, "Do you know that men in general have bigger brain volumes than females?" (Yeah, gorillas have larger brain mass than men too.)
Anyway, enuff of him and school.
I went to Köln the previous weekend, and it was truly fun to be with Singaporeans and see Xizhen after dunno how many months. (ever since the end of PDC!) That goes to show how busy we all are. =p I can't load the pics though, cuz clever me left my digi cam and earrings at Xizhen's place. =p
I tried to donate blood to the Deutsches Rotes Kreuz but the doc rejected me. She asked if I have slept enough, eaten enough and drunk enough. Of course I haven't, but that is normal for me on a school day anyway, sandwich for lunch, less than 1 L of water, etc. And I had a lecture afterwards. She told me that she couldn't take my blood then, cuz I would 'kip over'. (Which was true, I never knew how my neglectful habits have been weakening my system so much that I slept at lecture today) Afterwards, as I was having dinner, she came over to my table to ask if my lecture was fine, and if I felt ok, and told me that I could always donate at another time. That was so unexpected and pleasant of her to do so! I really have a v great impression of German docs, esp having been examined by 2 of them.
Right, I must read up on the swiss govt system now. No more procrastination!
=)Leeting
P.S Who left a comment on my last entry? Jo?
Subscribe to:
Post Comments (Atom)
1 comment:
I think that the example that you have given is so similar to what i have learnt in stats. So i thought i should leave you a comment. If you can still remember what we learnt about confidence intervals from stats in JC, there is always a level (like 95% confidence level) that we can accept or reject the null hypothesis that we have in mind. But yet again all these data that we collected are (like what you mentioned) subject to so many confounding factors like response bias, non-response bias, wrong sampling frame, etc. So actually a lot of statistical findings in the newspapers, magazines and other places, dun mean much actually.
In fact, sometimes the confidence level can be arbitarily chosen by the statistician in order to explain certain results or findings.
In order to be 100% confident that a confidence interval contains the pop. value, we can either sample the entire pop. (which is too time consuming) or we provide an absurdly wide interval of estimates. For instance, if we said the percent of students who think smoking should be banned is somewhere between 0% and 100%, we would have 100% confidence that we're right. But the interval is so wide that it's not informative.
Sometimes i really wonder what surveys are for. Haha.
Btw, i wasn't the one who left you a comment for your previous post.
Post a Comment