Biostatistical Consulting

Clinical Research Design, Epidemiology and Biostatistics Consulting

The CRDEB core provides consulting throughout the entire span of the research process. This includes:

  • Study design
  • Sample size calculations
  • Advice on data management
  • Statistical analysis
  • Interpretation of results
  • Dissemination of methods and results
  • Assistance with interpretation of statistical methods found in the literature

These services are by appointment only, and require the completion of the online request form. After requesting a consultation, a CRDEB team member will contact the investigator to schedule an appointment.

Please request assistance at least 5 business days before your deadline (for an abstract submission, etc.) in order to ensure your needs can be met by that deadline.

Clinical Research Design, Epidemiology and Biostatistics Authorship Policy

It is the policy of the Biostatistics Department that authorship should be based on level of involvement in the study and not whether the statistician was reimbursed financially for his or her efforts.   As recommended in published guidelines (Parker RA, Berman NG: Criteria for authorship for statisticians in medical papers. Statistics in Medicine 17: 2289-2299 (1998)), "The basis of financial support should be the time/effort spent on a project and the basis for authorship should be whether the statistician has made a scientific contribution to the project."
In determining whether scientific contributions were sufficient to warrant co-authorship, we adhere to the guidelines set forth by the International Committee of Medical Journal Editors:

“Authorship credit should be based on:

  1. Substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data
  2. Drafting the article or revising it critically for important intellectual content
  3. Final approval of the version to be published.

Authors should meet conditions 1, 2, and 3.”

For most consulting arrangements, the statistician will perform activity (1), by assisting in study design and/or analysis of data.  The expectation is that CRDEB consultants will participate in activities (2) and (3) unless otherwise negotiated. The manuscript will be strengthened by properly written statistical methods, clear reporting of results through carefully crafted tables and figures, and proper interpretation of the results. If a consultant within CRDEB performs these tasks, authorship has been earned and should be provided.

In those rare cases in which all parties involved agree prior to the commencement of any scholarly work that assistance with (2) and (3) is unnecessary, an acknowledgment of the efforts of the statistician is sufficient.

For questions regarding this policy, please contact the CRDEB Core Director, Dr. Tom Hulsey.

Clinical Research Design, Epidemiology and Biostatistics IRB Policy

For all research involving human subjects where actual data sets are reviewed or analyzed, the CRDEB Core requires information regarding approval or exemption from the WVU IRB.  Consulting for clients not affiliated with the WVU HSC will be reviewed individually to determine whether IRB approval is required and from where approval must originate.

Proof of IRB approval will be necessary before data will be analyzed or reviewed by the Department:

  • Copy of the official approval or exemption letter from the IRB
  • Email or letter from the PI with the approval/exemption date and IRB study number

Information regarding research integrity and the WVU IRB process can be found at the following links below:

Office of Research Integrity and Compliance

Database Guidelines for Statistical Analysis

These guidelines are not required.  They only serve as recommendations that will lead to a more timely and efficient consultation.  

Recommendations for efficient import of your data to statistical software:

  • Use standard software to enter and manage data (Excel, Access, SPSS, etc.).  It is very likely that the CRDEB consultant will use SAS to analyze your data.
  • Variable names should be included in the first row
  • One column per variable
  • A variable/column should be either numeric or character (do not use both in the same column!)   Example: Do not put "NA", "Missing", "< 0.3", etc. in a numeric column
  • Missing data should be left as blank cells
  • Delete all unnecessary rows/columns.  This includes summary statistics, any notes, etc.
  • To the extent that it is possible, check for data entry errors beforehand.

Recommendations for efficient data analysis:

  • One row per observation.  See templates below.  
  • Keep variable names short and unique.  Start with a letter and do not use special characters (only letters, numbers, and the underscore) or spaces.
  • Be consistent with character variables (Do not enter "Hisp" and also "HISP" for Hispanic)
  • It is helpful for yes/no variables to be denoted by 1's and 0's, respectively.
  • Highlighting, use of colors, special fonts, etc. will be ignored.

Example Dataset (One Observation per Subject)

ID

GENDER

AGE

RACE

HEIGHT

WEIGHT

SBP

1

M

19

WHITE

72

190

112

2

M

15

AA

65

162

120

3

F

25

HISP

54

114

100

4

M

31

AA

70

221

98

5

F

22

WHITE

60

150

90

6

F

16

WHITE

55

130

105


Example Dataset (Multiple Observations per Subject)​

ID

WEEK

GENDER

RACE

AGE

WEIGHT_LB

WAIST_IN

1

1

M

WHITE

19

250

48

1

2

M

WHITE

19

245

48

1

3

M

WHITE

19

242

47

2

1

F

AA

21

200

40

2

2

F

AA

21

199

40

2

3

F

AA

21

191