My research was about finding missing values titled "Bootstrapping with SAS and Exploring Dirty Data". At 2:30pm, we had free cookies. At 3pm, I started my saliva-splashing work. Blah...blah...blah... At 3:40pm, I was bombarded with big Q and small q. One of the questions was, "If I bump into somebody on the street and that person asks me to define bootstrap, what should I say?"
I would say, "It's a method where you perform sampling with replacement over and over again to find the relevant mean estimators such as the intercept and the slope, plug them into the equation and get your predicted y values." Just like the usual 串 Lulu, I added, "I just spent half an hour explaining about the methodology, that's the best answer I can give if you want me to wrap up the whole thing in one sentence". I looked straight into the eyes of Prof Paul Eloe who raised this Q. He smiled.
Reminded me of OYL's 你想怎样? during her presentation. Also reminded me of my own 串 answer to Nukman's Q of "You are blaming on the machine for your inaccurate result". Though it was 7 years ago, I still remembered vividly my answer: "Now we only have 1 accelerometer (to do measurement), they(the department/research team) are ordering 1 more accelerometer. Later, when they can take measurement at 2 points simultaneously, probably they can prove me wrong!" The very next thing I remembered was, Prof Zahari was trying hard to stiffle his laughter. And the guys sitting at the back row gave me a round applause.
This comment has been removed by the author.
ReplyDeleteZoom in on the desk to the left in the picture of Lulu with Li Qian for a nice surprise.
ReplyDelete