they ask by mikeholy


									Questions to ask the interviewers
1. Why is this position open?
2 What do you look for in a future faculty member?
 (or, What would you like done differently by the next person who fills this position?)
3. How is one evaluated in this position?
4. What are your tenure and promotion criteria? What are the promotion and tenure
5. What type of support does this position receive in terms of people, finances. etc?
6. What are some of the long term objectives you would like to see completed?
7. What are some of the more difficult problems one would have to face in this position?
8. How do you think these could best be handled?
9. What significant changes do you foresee in the near future?
10. What is the health insurance and benefits package?
11. (TEACHING) What is the teaching load per semester? How many courses? What is
the credit per course?
12. (TEACHING) How is teaching evaluated?


1. What accounts for success within the company?
2. What freedom would I have in determining my own work objectives, deadlines, and
methods of measurement?
3. By practicing your responses to some of these questions, hopefully you will not be
taken off guard if asked one of them. Most importantly, relax, go with the flow, and
before you know it, you'll be in your next job.
4. What advancement opportunities are available for the person who is successful in this
position, and within what time frame?
5. In what ways has this organization been most successful in terms of products and
services over the years?
6. How often has it been filled in the past five years? What were the main reasons?
7. (some special) program in your company
8. What are the grant-writing expectations? Is formal support available?
11. What is most pressing? What would you like to have done in the next 3 months.
12. What are some of the objectives you would like to see accomplished in this job?
13. How many statisticians are there in your center?
14. In what way do the statistician work with the medical investigators?
15. What do you expect me to do in the center?
16. What system do you mainly use?
17. Tell me about the duties and responsibilities of this position?
18. Why is this position open?
19. What kind of a person are you looking for?
20. Who would I be reporting to?
21. What kinds of people would I be working with?
22. What will be the greatest challenge in this job?
23. May I see the area where I 'd be working?
24. How is my performance judged in this job?
25. What is it like to work here?
26. So you have two different track, one is RS, another is RA, right?
27. How long is the tenure-track? how much is the chance to get a tenure?
28. Do you have much time to do research?
29. Do you have much service to attend?
30. How much is the tuition for a student
31. (not a good question) As a teacher, do you have lots of benefit
32. Do you often have chances to attend meetings?

Questions they might ask

(1) Human resource questions
Tell me a bit about yourself.

Currently I am a non-tenure track Research Assistant Professor in the Department of
Biostatistics at the University of Pittsburgh. I started this appointment from August 2004.
My primary responsibility is to provide biostatistical support for biomedical research
projects in the University of Pittsburgh Cancer Institute. I got my PhD degree from the
Department of Biostatistics at the Medical College of Wisconsin in August 2004. My
thesis focused on frailty models in survival analysis. I got my MS degree in Probability
and Statistics from Peking University of China. Before I came to MCW I taught
statistical courses at both the graduate and undergraduate level for a total of 8 years.

What are your qualifications?

I have 3-year non-tenure track research assistant professor working experience, 4-year
graduate research assistant experience, PhD degree in Biostatistics, good communication
skills and the ability to work as part of a multi-disciplinary team.

Additional (optional)
(1) 4 years' undergraduate + 3 years' undergraduate
   study in Peking university helps me to build a solid statistics background
(2) 8 years' statistics teaching made me more familiar with statistical areas, also gave me
much teaching experience
(3) From the 5 years' study in MCW I found it interesting to apply statistics in medical
and biological study. so I hope to work on this area. Your college is a good fit for me.

What do you like most, independence research or team work?

I enjoy both. On many projects I enjoyed team work with medical investigators and other
statisticians. When we met with difficulties we discussed together. By discussing with
them I learn a lot about how to solve problems in research. I also had a chance to use
what I learned in class. I think the best thing in collaborative research is people can
stimulate each other.

But I also like to do independent research, like in my dissertation work. It is challenging
also interesting. You would feel very glad when you solve a big problem. I think in
independent research you can think continuously without being interrupted.

What part do you think to be creative work you have done?

When I was doing my thesis research I noticed that the estimation procedure essentially
only use the laplace transform of the frailty, so I wrote the procedure in the form of the
Laplace transform, previous paper never did this way. Also in programming I often use
SAS/MACRO and UNIX script to save much time.

What do you do if you have different opinion from anybody else in your team?

First I think it normal to have different opinion in research. I would explain to him/her
why I want to do this way, and let him or her explain his reason. If we still cannot reach
an agreement, then I would say lets go back to think it over again very seriously. If after
that we still cannot reach an agreement, I would like to make our proposals open to all the
team members to let them discuss.

How do you lead a project if you've got a chance?

1) Create a group email link and let members communicate with each other through the
2) I will send everybody an email telling them some introduction about the project. and
tell them that we will have a meeting for it and let them think about what they want to do.
3) In the meeting I will assign the tasks to everybody based on their interest and the
project needs.
4) Set several stages. At the end of each stage we will have a meeting. Each member can
ask for a meeting if he has something to announce.

What do you think you can do for our center?

I have built up a solid background in statistics by more than ten years' studying and 8
years' teaching, also I have some basical knowleddge about clinical trials. also i have
research experience in hospital setting, so i am confident I can be a qualified component
of your center.

Your strong point and weak point?

I am creative, optimistic, diligent, easy-going, self-motivated, detail-oriented
My weak point is i know very little about American culture which may affect me
sometimes but I am trying to learn more and more.

1. Tell me about yourself? (try to hold your response to 2 minutes)
2. What do you know about our company?
3. Why should we hire you?

4. What can you do for us that someone else can't?
5. What do you look for in a job?
6. What skills and qualifications are essential for success in the position of ______?
7. How long would it take for you to make a meaningful contribution?
8. How does this assignment fit into your overall career plan?
9. Describe your management style.
10. What do you believe is the most difficult part of being a supervisor of people?
11. Why are you looking for a new career?
12. How would your colleagues describe you?
13. How would your boss describe you?
14. How would you describe yourself?
15. What do you think of your present or past boss?
16. What were the five most significant accomplishments in your last assignment?
17. What were the five most significant accomplishments in your career so far?
18. Can you work well under deadlines or pressure?
19. How much do you expect if we offer you this position?
20. Why do you want to work for us?
21. What other positions are you considering?
22. Have you kept up in your field with additional training?
23. What are your career goals?
24. What are your strong points?
25. What are your weak points?
26. How did you do in school?
27. What position do you expect to have in 2 to 5 years?
28. If you took the job what would you accomplish in the first year?
29. What was wrong with your current or last position?
30. What kind of hours are you used to working or would like to work?
31. Do you have your reference list with you? (Remember don't give it out unless it is
asked for).
32. Can you explain your salary history?
33. What questions didn't I ask that you expected?
34.Do you have any question for me? (See Questions for the Interviewer that you might
want to ask below).
(2) Why do you apply?
(3) Please state your uniqueness
(4) The vision of your graduate program
(6) Information about your own institution/program
(7) Research direction & funding sources
(8) What are your start-up needs
(9) What is your strengths/weaknesses
(10) Leadership
(3) Teaching interest
(4) Teaching and research philosophy

(2) Research questions
Tell me about your thesis research.

In my dissertation, entitled "Inference for the Shared Power Variance Function Frailty
Model and the Correlated Inverse Gaussian Frailty Model", I study frailty models in
survival analysis. In survival analysis, frailty models are commonly used to model
dependent event times. A frailty model is a random effects model, where the random
effects, that is, the frailty, acts multiplicatively on the hazard rate of each subject.

In the first part, I assume that the frailty is shared inside each group and the frailty
distribution belongs to a very general family, and find a stepwise procedure to specify a
simpler model. Also I compare the performance of the dependence measures under these
frailty models. The EM algorithm is employed to estimate the risk coefficients and frailty
parameters. A simulation study is used to evaluate performance of these estimators and
the model selection procedure.

In the second part, a more general frailty model is studied. This model assumes that
inside each group, individual frailties are correlated. Estimation and testing procedures
are also performed under this model. For both models we consider covariates so these are
extensions of the Cox model.

For both these two models, examples of real-life data are analysed to illustrate how these
models can be applied to biomedical studies.

What is your research topic?

My thesis is on frailty models in survival analysis. A frailty model is a random effect
model, where the random effect(the frailty) acts multiplicatively on the hazard rate of
each individual. My thesis has two parts: shared PVF frailty models and correlated
inverse Gaussian frailty models.

What is your research interest?

Survival Analysis, Clinical Trials, Statistical Genetics, Multivariate Statistical Analysis,
Linear Models

What do you do at UPCI?

Currently I am a non-tenure track Research Assistant Professor in the Department of
Biostatistics at the University of Pittsburgh. I started this appointment from August 2004.
My primary responsibility is to provide biostatistical support for biomedical research
projects in the University of Pittsburgh Cancer Institute. I work collaboratively with
medical investigators on many biomedical research projects. This includes writing and
reviewing statistical sections for clinical trial protocols and NIH/NCI grant proposals,
performing laboratory and clinical data analysis, and providing statistical consulting

Please give an example about the clinical trial you worked on

I helped medical investigators design a couple of phase I clinical trials. I mostly used
standard 3+3 dose escalation and de-escalation scheme for finding the maximum
tolerated dose (MTD). Also I helped write statistical analysis plan on these trials.

For instance, I established a phase I protocol to studying the safety of NK-92 cell
treatment in patients with advanced chronic lymphocytic leukemia, acute leukemia and
lymphoma disease. This trial is to determine whether the infusion of NK-92 cells, which
are highly cytotoxic to cancerous cells and are obtained from a continuously growing cell
line, causes undue side effects upon infusion into cancer patients, and to determine if a
target cell dose of 1010 cells can be infused safely.

I also helped medical investigators design some phase II clinical trials to study the
efficacy of some treatment. For most of these trials I used the standard optimal two stage
phase II design. I used PASS for the sample size calculation.

For instance, I established a phase II protocol to assessing the response rate and toxicity
of RAD001, an anti-proliferative drug with applications as an immunosuppressant and
anticancer agent, in patients with relapsed or refractory multiple myeloma which is the
second most prevalent hematological malignancy with approximately 15,000 new cases
per year in the US and remains incurable with a median survival of 3-5 years. I helped the
PI designed this clinical trial with Simon’s optimal two-stage design.

Please give an example about the data analysis project you worked on

On a project which targeted at the association between SMAD3 protein and radiation
resistance capacity, I built sophisticated statistical models (multilevel mixed models) and
developed novel data reduction techniques for cell growth and migration data created by
the cell culture imaging system, which had millions of observations and was hence very
difficult to handle without efficient statistical methods. He eliminated large objects based
on the distribution of object area at early scans. He examined the cumulative distribution
function of the velocity and developed fixed effects ANOVA models and multiple
comparison procedures to compare cell velocity at different times. Several multi-level
mixed models were also considered. Three SMAD3 cell lines (+/+, -/-, and -/-3) were
studied and compared. With my models it was found that SMAD3+/+ cells show
significantly better immigration capacity than SMAD3-/- cells, which suggested that
SMAD3 protein as a prominent member of the TGFβ receptor correlates with increased
radiation resistance capacity. These results are a great progress toward finding ways to
help human body defend against radiation injury. One paper describing these research
findings has been published in Radiation Research and three related papers have been
submitted for publication

Please give an example about the study design you did for a grant

Generally speaking, I helped medical investigators with grant proposals. I wrote the
statistical sections for them. I helped them design the experiments, estimate sample size,
and write statistical analysis plan. Some of these studies use ANOVA to study the effect
of some factors; Some of these studies are about performing selection procedures to
determine optimal drug combination or optimal dose.

I will talk about the Center for Countermeasures Against Radiation (CMCR) P01 grant.

The ionizing irradiation, which may be caused by environmental contamination with
radioactive substances, a nuclear reactor accident, or deliberate terrorism using
radioactive sources, could subject huge numbers of individuals to sublethal or even lethal
levels of ionizing irradiation. This P01 grant aims at developing new therapeutic
strategies to protect against and/or mitigate the deleterious effects of ionizing irradiation.
It is composed of five interlinked projects and six supporting cores including the
biostatistics core. Of the five projects, four target new drugs to unique sites of
mitochondrial irradiation damage, and one project uses a unique strategy of maintaining
mitochondrial generation of ATP during the time required for effective intrinsic repair as
well as action of other radioprotector drugs.

The specific aims of the Biostatistics Core are

   1. Contribute statistical expertise to the evolving design of in vitro and in vivo
   2. Perform statistical analyses of data from key in vitro and in vivo experiments;
   3. Develop more efficient dose-finding designs, based on accumulating data from
      CMCR experiments, to assess toxicity and efficacy of novel small-molecule
      radiation protective and mitigator agents used individually and in combination;
   4. Contribute to the review of pilot-project proposals, and provide statistical support
      to those that are funded by this CMCR;
   5. Work with the Project and Core investigators, including Bioinformatics (Core E),
      to ensure that data collection and database development are appropriate for the
      requisite statistical analyses;
   6. Collaborate with investigators in writing and preparing progress reports, abstracts,
      manuscripts, and presentations.

List some hot books and scholars

a) Analysis of multivariate data, by phillip Hougaard
b) wang(1991) and Dr. Klein : Positive stable
c) Li(1991?) inverse Gaussian
d) Klein(1992) Gamma
e) Dr. Petersen, dissertation
f) Yashin and Iachine, 1995

Describe your research. What are you currently working on? What is your five year

Currently I am working on inference for the shared PVF frailty model and correlated
inverse Gaussian model.
My five year plan:
1. continue with my thesis work. Discuss more about the large sample properties of my
estimates, also work on application of my methods to other areas like genetics
2. Find new partner to work with and new area to work on.
3. publish my results to survival journals.

4. Write proposals to apply for funds.

What are your plans for publishing?
In the first year I would publish papers based on my thesis work. For the years after, I
would continue with that and find more areas to work in.

How do you plan to support your research?
I would apply for some fund to support my research.

How will you seek funding to support your research?
Write proposal

In what journals do you plan to submit your research?
Life date analysis or Statistician

Do you happen to know of any businesses or governmental agencies in the San
Francisco Bay Area that use Biostatistical methods? If so, what Statistical methods do
they use? If not, in your opinion what are the main analytical techniques used in
Biostatistics applications in industry or government? Please explain.

 I know very little about the bay area, but I know in industries they use a lot of sequential
analysis, because the test for a new drug comprises of many
phases. Also experimental design, longitudinal data analysis, clinical trials and
categorical data analysis are used a lot in industries. In social sciences
they use a lot of sampling, questionnaire design and time series analysis.

Please discuss how you have incorporated computer software, such as SAS or S-PLUS,
into the Statistics classes you have taught or your interest in doing so in the future.

 For many statistics courses I would like to use computer software to illustrate how the
methods work. I would use some data sets throughout the whole
course for illustration and explained the statistics output to the students.

What did you do for your RA

2001 Work in our consulting group on several projects. In one of these projects, we
built a logistic regression model to predict probability of the child abuse. The factors we
considered are age, gender and many kind of bleed such as CSDH, PISH and
HEMCONT. In another project we built repeated effect models for the difference of CO2
reactivity at different MAC level. Here we are mainly concerned about the treatment
effect, that is, we are comparing the new treatment ISO and the old treatment SEVO. The
other covariates we consider are the blood flow velocity at both sides, respiration type
(hypo, norm, hyper), age, gender, weight, height and tumor size etc.

2001 Work on the propensity score for clinical trials. I did some simulation for
detecting selection bias.

I did a simulation study with Dr. Klein and use different method for the propensity score
estimation. The survival curve can be predicted well. We end up with a technical report.
(Background: In observational studies, investigators have no control over the treatment
assignment. The treated and non-treated (that is, control) groups may have large
differences on their observed covariates, and these differences can lead to biased
estimates of treatment effects. Even traditional covariance analysis adjustments may be
inadequate to eliminate this bias. The propensity score, defined as the conditional
probability of being treated given the covariates, can be used to balance the covariates in
the two groups, and therefore reduce this bias. In order to estimate the propensity score,
one must model the distribution of the treatment indicator variable given the observed
covariates. Once estimated the propensity score can be used to reduce bias through
matching, stratification (subclassification), regression adjustment, or some combination
of all three. In this tutorial we discuss the uses of propensity score methods for bias
reduction, give references to the literature and illustrate the uses through applied
examples. )

2001-2002. Biostatistics Core of the Specialized Center of Research Grant.
Study genetics of coronary collateralization by appling transmission-diseqilibrium test for
testing linkage between marker loci and associated disease traits. manipulating pedigree
data. implementing MCMC technique for a Bayesian TDT test for binary traits

2002-2003 Children's Hospital of Wisconsin. Dr.Nunchuck
 worked for Dr. Nunchuck for the overall performance of every medical resident based on
a survey of the doctors and nurses. Used factor analysis and cluster analysis for
questionnaire design. Analysis new born children clinical data to see how the childen's
height and weight change with the time .

2003- Dept. of Biostatistics, Cancer Center, Medical College of Wisconsin.
Worked with two faculty members from our department on multiple comparison
adjustment procedures in survival analysis on a grant from cancer center of MCW

Now On Dr.Klein's R01 grant doing some consulting job

Please describe your interest in and experience with computational statistics. What
statistical programming packages have you used most in your work?

 I have a strong interest in computational statistics. In our department I worked with
faculty on several projects, many of which require much computation. For instance, in
one of these projects I used simulation method to compare the performances of several
multiple comparison procedures. I used FORTRAN and IMSL to generate data and
compute the test statistics and then used SAS to summarize the size and power of these
test techniques. In another project I used SAS/IML to perform Bayesian analysis of the
family data. I employed MCMC technique to estimate the parameters and used SAS
macro to handle a large amount of data. The statistical packages I have used most are
SAS and Splus.

(3) Teaching questions

What's your teaching experience:

(1) 1988-1991 Wuhu teacher's college
              teaching assistant
              courses: Advanced Algebra, Advanced Calculus

(2)1994-1999 Beijing Institute of Technology
              courses I taught:
                          undergraduate level: Advanced Algebra, Advanced
Calculus, An Introduction to Probability and Statistics,
                          graduate level:          Advanced Statistics, Applied
Regression analysis, Multivariate analysis

What's your teaching philosophy

My teaching philosophy is based on some very basic principles derived from my own
experiences and expectations as a student and as a teacher. The goal of my teaching is to
ensure that students understand the concepts, course materials and they can solve
problems using what they learned in class. I try to make the courses I teach the 'state of
the art' so that I can also learn a lot myself! I usually can find many topics of great
interest while teaching. The course syllabus to me is a flexible guide and not a rigid
schedule. Although I must always know how the course should proceed, I must also be
prepared to adapt the course to the students' interests and concerns. I especially enjoy
teaching classes in Regression Analysis, Applied Statistics, and Survival Analysis
because the possibilities to be flexible. Most of my classes have computing and writing
components, because there is not much point to statistical exercise unless you can
calculate what you need to know on the computer and then communicate what you have

Specifically I would, first, let the students learn statistics through real life situations, so I
will provide a variety of examples to allow students to see the application of statistics in
their own discipline. Second, combine class teaching with statistical software such as
SAS, Splus and so on. I would like to teach my students on how to use them. I ask my
students to work on their final projects by team, which each group has to get a real data
set from an internet database and then analyze it by using all statistical techniques they
learned, and that goes for undergraduate students as well. Finally, each member of team
has to present it as his/her course presentation. I consider each class as an opportunity to
evaluate and to improve my own performance as an instructor. I encourage evaluative
feedback from the students. I particularly like working in an academic environment
because I get to work with students on various research projects and paper writing. And
finally, I want to combine class teaching with the internet. I will put my course syllabus,
references, assignments and solutions on the web so that the students can easily get
access to it.

How do you structure your courses?

At the beginning of the semester, I would send each student the syllabus of the course.
The syllabus would tell them the objective of the course and also the plan, the scoring
policy and the references. At the beginning of each class I would give a brief review
about what I taught in the last class, then I start to teach new staff, finally I would give a
summary of this class and also some assignment.

Describe how you'd teach an introductory statistics course?

Before the first class, I would send each student the syllabus of the course, the sylabus
would tell them the objective of the course and also the plan, the scoring and the
references. As to the content of the course, I think first I would give an introduction to the
basics of probability theory. Then I would start to teach statistics concepts, point and
interval estimation, hypothesis testing, simple linear regression and introductory
ANOVA. At the beginning of each class I would review last class, then I start to teach
new staff, finally I would give a summary of this class and also some assignment.

What courses would you like to teach? What textbook do you want to use?

    Statistical Inference -- by George Cassela, Berger
    Applied Statistics
    Regression Analysis
    Survival Analysis-- the book by Dr.Klein
    Linear Models

What course, not currently in our catalogue, would you like to develop?

Multivariate survival analysis, Categorical Data Analysis

How does your research inform your teaching?

First, from my research I get more and more familiar with statistical methods, which
make my teaching easier.

Second, I can use many of the problems that I met with in my research as examples in my

Third from my research I have ideas about what have been done in this area and what
have not, so I can also let students know better about this area.

What is the SMC teaching fellowship

SMC teaching fellowship is an award for excellent teachers.

Have you ever used a web-based course management system for presenting course
material on the Internet? Also, please discuss your
experience with or plans for using the Internet to present course material.

 I did not have any experience with using Internet for teaching, but I would like to do so.
I plan to use the Internet to show students the course
discription, syllabus, work assignments and solutions. From the web page the students
will have a clear idea about what I expect them to learn.

>A. Almost all of our classrooms have a computer and projector system for
>use during classroom lectures.
>B. Use of the computer lab is expected in our basic statistics classes
>for non-majors (about half of the sections we offer each quarter) as
>well as in major courses. There are 20 stations in the lab. If two
>students work together on a computer, then a class size of 40 or fewer
>students can be adequately accommodated. However, the actual number can
>range from 30-50. (We offer no classes in large lecture/discussion
>section format.)
>C. Our students are quite diverse with respect to ethnic background,
>native language, age, and educational background.
>D. Because many of our B.S. and M.S. majors are employed, we have
>traditionally scheduled most major courses in the late afternoon and
>evening. Recently, we have been increasing the number of later evening
>courses. (Most general and some major courses are also offered earlier
>in the day.)
>E. We have approximately 100 M.S. majors and approximately 20 B.S.
>majors and minors. Our B.S. majors are mostly double majors with other
>fields such as mathematics, computer science, psychology, economics, and
>biology. We graduate approximately 15 M.S. students each year with
>approximately 40 students taking courses at any one time. Students
>generally take between 2 and 4 years to complete their degrees. Most of
>our graduates readily find employment in Bay Area companies and
>governmental agencies. A steady stream of our graduates has continued on
>to work on Ph.D.s at other excellent universities.
>F. We have recently added options to our M.S. degree in Biostatistics,
>Actuarial Science, Applied Statistics, and Mathematical Statistics. This
>job opening is for a Biostatistician, and a major portion of the job
>will be curriculum development and student advising for this Option.
>For still more information about the curriculum in Statistics programs

>and at CSU Hayward in general, please visit our Departmental website:
>and the Universitys online catalog (see the Degrees & Programs link)
>for more details about the curriculum.

How much do you know of us?

ICON Clinical Research is a full service clinical research organization providing a
comprehensive range of clinical services in Phase I-IV clinical trials to the
pharmaceutical, biotechnology and device industries. ICON Clinical Research was the
first international CRO awarded ISO quality certification across all divisions and

Multiple comparison project.

Tell me about your research in the cancer center

In the year 2003, I worked as a Research Assistant in the Cancer Center and Division of
Biostatistics at the Medical College of Wisconsin. I worked with two faculty members on
multiple comparison adjustment procedures in survival analysis on a grant from Cancer
Center of Medical College of Wisconsin. Studied eight testing procedures and
implemented them in Fortran and SAS. Performed both simulation and real data studies.
Compared the performances of these procedures under different situations. Derived the
limiting distribution of the weighted log rank test statistics for multiple comparison.

What are the eight testing procedures?

1)The multiple comparison without adjustment.
2)Multiple comparison with Bonferroni adjustment.
3)Multiple comparison with Holm adjustment.
4)Multiple simulation
5)Multiple simulation with Holm adjustment
6)Multivariate simulation
7)Multivariate simulation with Holm adjustment
8)Closed test procedure.

Children's Hospital

Tell me about your work in the CHW

In the year 2002-2003, I worked as a Research Assistant in the pediatric surgery clinic of
Children's Hospital of Wisconsin. I documented and analyzed daily data (i.e. vital signs
like TMAX (body temperature), HR,RR,MAP, Cap Refill (blood pressure)) for the

premature neonates with necrotizing enterocolitis. Appraised the overall performance of
every medical resident based on a survey of the doctors and nurses. Used factor analysis
and cluster analysis in SAS and SPSS for questionnaire design. Helped a doctor finish her
paper abstract.

What analysis did you do on the clinical data?

For each neonate, we have observations documented by nurses. These are observations
on vital signs, like body temperature and blood pressure. What I did is input these data
into a database system and made summary tables about the sample mean, variances and
correlation coefficient between them. We include them in a technical report.

Why did you use factor analysis?

The doctors already designed a questionnaire which had about 70 questionnaire in it.
They used for the past years and it works well. But they still hope to reduce the number
of questions, so what I did is to use the data file, which includes the responses of the
doctors and nurses on the 70 questions and try to find the correlation between the 70
questions. We cannot directly use the correlation matrix itself, because the coefficients
are pretty close to each other, they are all moderate, so I choose to use the factor analysis.
I used the principal component analysis option in SAS Proc Factor. What this method did
is try to extract the principal components and find the loadings of each question on the
components. We collapsed the questions based on the loadings, I mean, those questions
having the main loading on the same component can be regarded as essentially the same.

Why did you use cluster analysis?

We use the Proc Cluster in SAS and basically it performed hierarchical cluster analysis.
At first, each question is a cluster, then we collapse these clusters step by step based on
the distance between clusters. We can end up with any number of clusters as we want.

The Genetics Project

Tell me about your research in genetics

In the year 2001-2002, I worked as a Research Assistant in the Division of Biostatistics at
Medical College of Wisconsin. I worked with faculty in the Biostatistics Core of the
Specialized Center of Research (SCOR) grant to study genetics of coronary
collateralization. Applied the transmission-disequilibrium test (TDT) for detecting
linkage between marker loci and associated dichotomous disease traits. Processed
pedigree data. Developed and implemented Markov Chain Monte Carlo technique for a
Bayesian TDT test for binary traits.

What is a TDT test?

TDT is short for Transmission/Disequilibrium Test. It is a viable alternative to other
existing sampling designs for testing for linkage between marker loci and associated
disease traits when marker genotypes are known for both the parents and affected

offspring from independent nuclear families. As a test for linkage, transmissions from
heterozygous parents to their affected children are used in the analysis. The test compares
the frequency of marker alleles transmitted to affected children versus the frequency of
marker alleles not transmitted, using a chi-square statistic. Thus, for the purpose of
testing for linkage, the TDT has the advantage of not requiring data on either multiple
affected family members or unaffected sibs.

How many alleles do you use?


How Gibbs sampling works?

Gibbs sampling starts from a certain initial point. At each step, by the use of the prior and
likelihood information, it returns a new point. We continue the process further again.
Finally, after many burn-in runs, when the process becomes steady, we get an array of
numbers which can be regarded as a random sample from the posterior distribution.

How does MCMC technique work in this case?

We generate random samples from posterior distribution by running a Markov Chain
whose steady state distribution equals the desired posterior distribution. The generated
samples must then be subject to convergence diagnostics to ensure that the Markov Chain
has attained its steady state. The prior and posterior distributions of the parameters are
graphical. The likelihood of the data provides substantial information about the disease
allele frequency, recombination fraction and disequilibrium coefficient. The credible
intervals of the parameters can also be calculated.

Tell me about your consulting experience

In the year 2000-2001 I worked as a Research Assistant in the Biostatistics Consulting
Center of our department. I helped finish several consulting projects. Built a logistic
regression model to predict probability of child abuse. Fitted repeated measures models
to compare two treatments. Calculated sample sizes for several types of experimental
step1. Receive the application letters where we ask for research hypothesis, the data,
relevant papers from the literature, research proposal or draft of their manuscript if
Step2. Ask about their main concern.
Step3. Write a report. Tell the results.

Do you know anything about clinical trials? Do you have any related experiences?

Yes. I know a little about clinical trials, mainly from the course I took in our department.
I have some experience in the hospital setting because I worked as an RA in the CHW.
Also I I analysed some clinical data in consulting projects.

Then, tell me what you know about clinical trials.
A clinical trial (also clinical research) is a research study in human volunteers to answer
specific health questions. Carefully conducted clinical trials are the fastest and safest way
to find treatments that work in people and ways to improve health. Interventional trials
determine whether experimental treatments or new ways of using known therapies are
safe and effective under controlled environments. Observational trials address health
issues in large groups of people or populations in natural settings.

What are the different types of clinical trials?

Treatment trials test new treatments, new combinations of drugs, or new approaches to
surgery or radiation therapy.

Prevention trials look for better ways to prevent disease in people who have never had the
disease or to prevent a disease from returning. These
approaches may include medicines, vitamins, vaccines, minerals, or lifestyle changes.

Diagnostic trials are conducted to find better tests or procedures for diagnosing a
particular disease or condition.

Screening trials test the best way to detect certain diseases or health conditions.

Quality of Life trials (or Supportive Care trials) explore ways to improve comfort and the
quality of life for individuals with a chronic illness.

What are the phases of clinical trials?

Clinical trials are conducted in phases. The trials at each phase have a different purpose
and help scientists answer different questions:

In Phase I trials, researchers test a new drug or treatment in a small group of people (20-
80) for the first time to evaluate its safety, determine a safe dosage range, and identify
side effects.

In Phase II trials, the study drug or treatment is given to a larger group of people (100-
300) to see if it is effective and to further evaluate its safety.

In Phase III trials, the study drug or treatment is given to large groups of people (1,000-
3,000) to confirm its effectiveness, monitor side effects, compare it to commonly used
treatments, and collect information that will allow the drug or treatment to be used safely.

In Phase IV trials, post marketing studies delineate additional information including the
drug's risks, benefits, and optimal use.

What is a protocol?

A protocol is a study plan on which all clinical trials are based. The plan is carefully
designed to safeguard the health of the participants as well as

answer specific research questions. A protocol describes what types of people may
participate in the trial; the schedule of tests, procedures,
medications, and dosages; and the length of the study. While in a clinical trial,
participants following a protocol are seen regularly by the research
staff to monitor their health and to determine the safety and effectiveness of their

What is a placebo?

A placebo is an inactive pill, liquid, or powder that has no treatment value. In clinical
trials, experimental treatments are often compared with placebos to assess the treatment's
effectiveness. In some studies, the participants in the control group will receive a placebo
instead of an active drug or treatment.

What is a control or control group?

A control is the standard by which experimental observations are evaluated. In many
clinical trials, one group of patients will be given an
experimental drug or treatment, while the control group is given either a standard
treatment for the illness or a placebo.

Do you know about our center? How do you know?

Yes. I know you are the Center for Biostatistics in AIDS Research (CBAR), is a Center at
the Harvard School of Public Health. You are doing statistical research in clinical trials
and other public health research in HIV disease, developing methods for medical
interventions and study design, and providing education and training relevant to statistical
aspects of HIV disease research.

I know of your center from my advisor Dr. John P. Klein, who encouraged me to apply
for your positions.

Your knowledge in

Basic statistics
clinical trials
survival analysis
sequential analysis
experimental design
sample size determination
linear models
multivariate analysis
categorical data analysis
nonparametric analysis
bayesian analysis
explain the results-- interaction term
epidemiology ( sensitivity, specificity, incidence, prevalence, crossover, case-control,
prospective, retrospective, cross-sectional, cohort)

Interview Tip 1: Plan Ahead - Do a little homework! Research the company and the
position if possible, as well, the people you will meet with at the interview. Review your
work experiences. Be ready to support past career accomplishments with specific
information targeted toward the companies needs. Have your facts ready!

Interview Tip 2: Role Play - Once you have finished studying, begin role playing
(rehearsing). Use the general questions provided below in the Interview Preparation Area.
Write down answers if it helps to make your presentation more concise. Try to keep your
answers to the information your new employer will want to know.

Interview Tip 3: Eye Contact - Maintain eye contact with your interviewer. Show you
want the job with your interest.

Interview Tip 4: Be Positive - In particular, avoid negative comments about past

Interview Tip 5: Adapt - Listen and adapt. Be sensitive to the style of the interviewer. Pay
attention to those details of dress, office furniture, and general decor which will afford
helpful clues to assist you in tailoring your presentation.

Interview Tip 6: Relate - Try to relate your answers to the interviewer and his or her
company. Focus on achievements relevant to the position.

Interview Tip 7: Encourage - Encourage the interviewer to share information about his or
her company. Demonstrate your interest. Some suggested questions to ask the interviewer
are provided in the "Questions You Could Consider Asking the Employer" section.

                                     Old Questions:

Why did you do in the Children's hospital?
Why did you use factor analysis and cluster analysis?
Do you like teaching or research?
If we let you teach an introduction course to undergraduate students, how do you teach?
Why do you think you're a good fit to this position?
 We mostly do research, what do you prefer, teach or research?
Tell me something about you thesis?
Do you run on some real data? Do you find something significant?
What's your teaching philosophy?
Have you applied to other positions yet?
How gibbs sampling works?
How many alleles do you use?
Do you incorporate some prior consideration?

Do you have some experiences in experimental Design?
Do you know anything about our company?
Harvard CBAR
Tell me about yourself?
Pick one of these projects you did and tell me about it.
What kind of research do you like to do?
Do you plan to teach here?
Do you have any more questions?
How do you explain to scientists?

1.8. What are your plans for integrating students into your research?

If possible I would advice students. I would first give them some papers to read. Then I
let them do some basic research that is related to my research.

1.9. Please explain why you feel you would be the best person to help us develop this
program. Include some specifics in terms of how you would work with our
Department to contribute to our M.S. Option in Biostatistics.

 I have built a solid pure statistics background with a total of 7 years' study for my B.S.
and M.S. in Probability and statistics at Peking University.
With the 5 years' study for my Ph.D. in the Dept. of Biostatistics at Medical College of
Wisconsin, I gained quite a lot of experience in applying statistics methods to biological
and medical sciences. Also I taught mathematics and statistics courses for 8 years. I think
I am well qualified to help you to develop the M.S. program.

1.11. Please read the additional information included below and briefly comment
about how you see yourself fitting into our Department.

 I had much experience in undergraduate and gradute teaching. In my classes the students
were also quite diverse with respect to age and educational
background. From the additional information I think I will be able to well fit into the


To top