Terms Of Use | Privacy Policy



Process and Guidelines


Part A: Project Prioritization

The Biostatistics Core is a shared resource available to all Stephenson Cancer Center members. Requests for support are initiated by submitting the Online Request Form. In general, requests for support are prioritized according to proposal deadlines and the level of support required as determined by the Core Director. However, in the event that the demand for support exceeds the Core’s capacity, the following schema will be used to help prioritize projects:

Stephenson Cancer Center priority application (as determined by the Stephenson Director)
Large Center or Multi PI projects (P Awards, U Awards, etc.)
NCI projects
NIH projects
Other projects

Part B: Grant Preperation

With rare exception we require grant proposals that involve biostatisticians to be submitted to the Core at least three weeks before grant submission deadline.

For grant applications that include a BSE faculty member, salary support will be included. The minimum percent effort funded on a grant is generally 10 percent for clinical studies and 5 percent for basic science studies for faculty biostatisticians.

Core Information Systems

We require a minimum of two weeks (for simple analysis) and four weeks (for complex analyses) to deliver initial statistical analysis results for abstracts and manuscripts.

Biostatisticians are expected to be co-authors and the decision on authorship should be based on scientific contribution, independent of funding consideration, as per the International Committee of Medical Journal Editors guidelines.

Part D: Initial Consultation Visits

Investigators should prepare responses to the following questions for initial consultation visits:
Statement of research questions:

Primary objectives, which in turn will eventually determine:

Choice of statistical strategy
Calculations of sample size, effect size, or statistical power
Secondary objectives

Details regarding primary research question:

Outcome of interest:
Level of measurement (nominal, ordinal, interval, ratio) to focus and limit the realm of useful statistical tools, models, or approaches 

Design issues related to primary question:
Comparison groups, Based on intervention, Based on interest in stratification by important cofactors, covariates, or confounders
Hierarchies, clusters, or correlations among the outcome measures such as repeated measures over time (i.e., measures taken each week), spatially correlated data (i.e., measures from left and right eye), natural groupings (i.e., families, clinics, etc).

Plans to measure continuous or categorical covariates that will permit:
Exploration of subgroup associations (statistical interaction or effect modification)
Adjustment for confounding of the primary association of interest by factors known or suspected to be associated with the outcome from previous literature

Preliminary data or published papers that provide estimates of:
The magnitude of the outcome of interest (such as group means or proportions)
The magnitude of the variability you expect in your outcome (usually standard deviation)
These will be used to calculate sample size, effect size, and power

Details regarding the participants in your study

Inclusion criteria to help define the population(s) of interest from which group(s) will be selected or sampled
Exclusion criteria that, by restricting the sample, limit the influence of potential confounding variables or known sources of variation
Sampling plans (random, stratified, paired)

Requirements associated with delivery or sharing of study data

Data must be de-identified so that it has no direct identifiers such as:
Street name or street address or post office box (i.e., not including city, state, or ZIP code)
Telephone and fax numbers
Email address
Social security number
Certificate / license numbers
Vehicle identifiers and serial numbers
URLs and IP addresses
Full-face photos and other comparable images
Medical record numbers, health plan beneficiary numbers, and other account numbers
Device identifiers and serial numbers
Biometric identifiers, including finger and voice prints
The study protocol must accompany data.

A data dictionary should accompany data and should include:
Variable names
Units of measurement for each continuous variable
Definitions for levels of categorical variables with underlying continuous measures
Definitions of codes for nominal categorical variables like race, ethnicity, county, etc.

If data is not de-identified then send via the following link. You will need to Zip the file(s) to send via the link.

Contact & Links


Contact Information

Stephenson Cancer Center
800 NE 10th St., Room 5037
Oklahoma City, OK 73104
E: scc-biostat@ouhsc.edu

Core Hours

A Biostatistics Core faculty member will generally be in the office on Tuesdays, but please schedule an appointment.

Faculty Expertise

Sara Vesely, PhD
View Profile

Daniel Zhao, PhD
View Profile

Kai Ding, PhD
View Profile

Usage Acknowledgement

If research supported by the Stephenson Cancer Center core facilities results in a publication or news release, please acknowledge this support in your manuscript. Following publication, please send us one of your reprints for our records.