|
Department of Biostatistics, Bioinformatics, & Epidemiology at |
Frequently Asked Questions
I NEED HELP WITH MY GRANT!
Q. I am in the early planning stages of writing an R01. How long should I wait before contacting one of the HCC statisticians to get involved?
A. Right away! We can have the biggest impact and maximize the efficiency of your study by getting involved at the early stages of planning. When helping with grant development our jobs are to assist in study design (including sample size calculations) and also to provide statistical analysis plans. Please contact one of us soon!
HOW MUCH DO YOU CHARGE TO HELP WITH MY GRANT DEVELOPMENT?
Q. I would really like some help with my grant because I have no idea how many mice to include in my experiments, or how to formally describe the comparisons I plan to do. But, I don't have any funding to pay a statistician. Is there some way to include a charge-back in the budget?
A. NO NEED! Statistical support for cancer research grant development for HCC investigators is FREE!!!!! We expect that, when appropriate, HCC biostatisticians will be included as co-investigators on the grant to cover the time they will spend on the grant after it is funded. However, we provide services for grant development free of charge. Please contact one of us soon!
AT MUSC, DO WE HAVE ANY STATISTICAL SOFTWARE PACKAGES AVAILABLE?
Q. I know that MUSC has a bunch of software available to all MUSC faculty and staff. Is any of it statistical software?
A. Yes, MUSC makes SPSS available. Contact your IT person and then can have it automatically loaded onto your computer. We will be adding pages to our website with some SPSS tutorials that are specific to cancer research biostatistics.
WHICH TEST SHOULD I USE?
Q. I've had some statistical training and know how to perform tests
like the t-test and chi-square test. But, my problem is that I can
never remember which test I am supposed to use when. Are there a set of
rules that define what statistical test to use and when?
A: There are some guidelines that have been developed. Here is a link that should help: WhatTestToUse?
HOW DO I SOLVE A SYSTEM OF EQUATIONS?
Q. I have a system of equations to solve. It has been so long since I had algebra! Is there an easy way to remind me how to do this?
A. Well, we could try to re-teach you algebra. Perhaps a more efficient solution would be to use a web-based program that can solve a system of equations for you: QuickMathSolutions
HOW DO I CALCULATE A P-VALUE?
Q. I have a t-statistic that I want to find the associated p-value for. The way I learned it, we had to look these up in tables in statistics books. Isn't there a better way to do this that will allow me to get the exact p-value?
A. There is an easy way to do this online. For t-statistics, chi-square, Z-scores, F-statistics, and many more, you can simply input your statistic (and its degrees of freedom or other required information), and the p-value will be calculated. Try it: p-value calculator
IS THERE A FREE SAMPLE SIZE CALCULATOR I CAN USE?
Q. Sometimes I would like to be able to use a sample size and power calculator, but I do not want to invest a lot of money in software. Are there any freely available web-based sample size calculators?
A. The "Cancer Research and Biostatistics" (CRAB) group has a very nice page with a number of useful sample size calculators. They are, for the most part, easy to use. Of course, you have to have some ideas a priori about the effect size, the variance, and other parameters before you can do a sample size calculation. For example, if you are planning a randomized trial with time to progression as the outcome, you will need to have an expected accrual rate, and have an idea of the expected time to progression in the two arms of the study. Here is the link to the CRAB calculators: CRAB calculators.
WHERE CAN I FIND AN ONLINE REFRESHER COURSE ON STATISTICS?
Q. Every once in awhile, a statistical topic comes up that I'd like to get a refresher on. Is there an archive for some of the common statistical topics in cancer research? For example, I know what power is, but I'd like a resource to provide some more detail and remind me of the statistical issues involved.
A. A professor at Rice University has put together a collection of pages on some basic biostatistics topics, including power, sampling distributions, confidence intervals, ANOVA, hypothesis testing, and quite a few more. Check out his page: HyperStat. But, if there is a topic you'd like more information about, please feel free to contact one of us.
WHAT TYPES OF STATISTICAL FREE-WARE IS OUT THERE?
Q. There are a few types of tests that I need to do frequently, and I'd like a simple interface to be able to do them. Are there free web-based statistical calculators for doing things like making contingency tables, calculating correlations, and performing sample size calculations?
A. There is a very large number of different web-based computer tools. StatPages has compiled a list with over 600 pages and 380 calculators: StatPages.
WHAT IS A DATA SAFETY MONITORING BOARD?
Q. I am planning a clinical trial and am not sure if I need to include a data safety and monitoring board. Do I need one? If so, who should I include and what are the responsibilities of the DSMB?
A. It depends on the type of trial design you have. In general, for phase III studies for which you are comparing two or more groups of patients, your study requires a data safety and monitoring committee. For early phase trials (e.g., phase I and phase II single arm), local monitoring of the trial through the clinical trials office will be sufficient. Dr. Jay Johnson has put together a very nice slide presentation about DSMBs explaining what they are and what they do: DSMB.pdf.
Back to Statistics Help Main Page