Thursday, November 30, 2006

A test of confidence...

Let me show you Biostatistics students in my cohort what I learnt from Biostatistics. It will come in handy after this paper. No more stats for the rest of my life, though Andy has made a big impact in my life. He really made the module very interesting and I love his lectures. He made lectures so much easier to learn and easier to capture. Unfortunately, his papers is one of the killer papers in NUS and I didn't get to do the last question. It was the easiest question that was available. Only managed to finish most of it. Damn... It's all about speed and precision. It's all over now... Thank GOD... But for those who think of recruiting me into statistics, haha, sorry, I'll pass...

Lets say I want to test how confident I am before the paper and after the paper. I ask myself this very same question throughout the period of 2hours, plus minus the waiting time, "I'm going to survive this paper..." The number of times I ask myself before and after the test can be illustrated below

So after coming out with this table, we must use McNemar's test because we want to see whether they differ. Since b+c >20, we must use Chi-square approx test:

H0 : P(Confidence before paper) equals to P(Confidence after paper)
H1 : P(Confidence before paper) not equals to P(Confidence after paper)

Test Statistics:


X2 = 985.0090817

At alpha = 0.05 level, we reject H0 if X2 >X21,0.95. Since X2 > X21,0.95 = 3.84, we reject H0 and conclude that there is a difference in the confidence level. Thus, it means that I was confident initially, now I'm going to die... Thank you everyone for appreciating my work. If you don't then too bad...

Traumatized by ST2238
The Mixed Boy

0 Comments:

Post a Comment

<< Home