发信人: johnuseast (john), 信区: Statistics
标 题: 也谈生统师面试
发信站: BBS 未名空间站 (Sat Nov 2 06:12:42 2013, 美东)
Many people like to talk about interview questions.
Nothing is wrong about it.
However, there is something lacking: the interest in the framework behind
I therefore encourage job hunters to ask the following questions:
(1) why and how the interview questions were formed?
(2) how to make the interviewers happy with your questions?
(3) how to improve your answers in later interviews.
To answer your first question, you need to learn the environment
biostatisticians work in. You need to know what they are talking in daily
basis - you need to know their mindset. I saw some one asked in this forum
the question "what do biostatisticans do in the pharmaceutical company?"
and got the answer "go to meetings". That is the first good answer, yet,
not enough: what do they talk in the meetings? How did they prepare the
meetings? What did they write before they go to meetings?
Then you should try to find one thing: the statistics they use compared to
the statistics you learnt at school.
"Statistics is statistics … " you might say.
But different areas have their own focus. For example, biostatisticians
working on oncology studies will certainly be thinking of all available
methods for survival analysis. Yesterday, I came from a meeting for cancer
study when I started interviewing a candidate for a middle level biostat
position. I was surprised that a person with 3 years of pharmaceutical
experiences could not explain what Cox model is.
The reason is simple - she had been working on CNS for 3 years.
To understand the situation is one thing. to be pleased is another - a
biostatistician without knowing the basic methodologies in survival analysis
will not please any colleague.
Here is the first step: if you are serious about finding a job in the
pharmaceutical industry, get your class notes on survival, experiment
designs, nonparametrics, read through them, tie them together, test yourself
by providing solutions to the problems in each area.
Bottom line: make your self at home with what you've learnt from school
Then the next thing is how to tie your knowledge to the problems your
interviewers are facing on daily basis. for example, if your dissertation
topic is related to missing data handling, you might want to google what are
the current opinions on LOCF.
Then the next step is practice.
I find some students can talk very nicely when 1 to 1 in a casual fashion,
but performs very differently when facing a few people. Some student even
told me that she was scared of any onsite interview!
No easy solution but practice. When you have a good summary of the
knowledge you've learnt at school or during work, describe the concepts to
your friends, your husbands, your wife …. or your mirrors ….or recently, I
found two methods very helpful:
(1) making videos. I once asked bancova students make presentations using
videos. it really helps to improve your presentation skills.
(2) be an interviewer. Try to "interview" your friends by putting yourself
"biostatististiain". Some of students recently helped us to
interview some applicants and they said it really helped them to rethink all
the questions during the process.
(3) record what you talked, review them, and then improve.
Here is my general advice - There are no majcs here. there are no shut-cuts
. what you need is hard work. I'm for discussing interview questions.
However, I suggest making the dissuasion as a way of enhancing your
knowledge, not as finding ready-made answers. In other words, do not expect
you can pass the interview as if you are taking SAS base certificate by
remembering the answers to the questions.
Bancova.com - bridging academics to industry
※ 修改:·johnuseast 於 Nov 2 06:18:34 2013 修改本文·[FROM: 98.]
※ 来源:·WWW 未名空间站 海外: mitbbs.com 中国: mitbbs.cn·[FROM: 98.]