|
|
|
General Policy
The policies listed herein are intended to allow for high productivity, timely results, clear expectations,
and successful experiences.
See the charging policy
for more specific information about partial cost-recovery of the Biostatistics Consulting Group (BCG).
- Funding Submissions
- Researchers are encouraged to seek advice from the BCG early in the grant proposal development stage. Requests should be submitted at least 5 weeks before the grant deadline. More time may be necessary to facilitate mandatory review and/or other requirements from the researcher's home department.
- Consultation will only be provided for projects where a BCG consultant will do the statistical work.
- The BCG will provide design advice, sample size calculations, and an analysis plan. The BCG consultant will write the sample size and data analysis sections of the grant proposal, if requested. Edits to these sections need the approval of the BCG consultant involved to ensure that alterations do not jeopardize the statistical integrity.
- The BCG is a fee-for-service unit and a cost quotation for statistical work will be provided for inclusion with the funding application. It is not appropriate to indicate BCG involvement or any BCG staff involvement in any funding application without obtaining a BCG cost quotation.
- By including a BCG cost quotation in a funding application, the researcher agrees to the terms of the quotation.
- Researchers must inform the BCG of the outcome of funding applications. If funded, a work schedule and timelines will be determined.
- Study Protocols
- A study protocol/proposal should be provided to the BCG. BCG consultants can provide input on several aspects including study design, data collection, and statistical analyses.
- Changes to the study protocol should be discussed with the BCG consultant, before implemented. Such changes might have statistical implications as well as budget implications.
- All interim analyses must be pre-planned and stated in the study protocol. Interim analyses will not be conducted unless they are pre-planned.
- Forms used in the collection of data (e.g., case report forms) need to be provided to the BCG consultant as early as possible. The BCG consultant can provide feedback and advice on such forms during the development process.
- Data Issues
- Researchers are expected to have the data available in an appropriately designed database. Researchers are encouraged to access database experts to ensure an appropriate database is designed, data are cleaned prior to delivery to the BCG, and data are in an appropriate format for analysis.
- Double entry techniques should be used to enter the data.
- Data files provided to the BCG must not contain subject identifiers (e.g., subject name, initials, person health number, address). A unique study ID number should be provided for each subject.
- A data dictionary is to be provided for each data file. The data dictionary should include for each field: Name, Description, Code Levels (eg., 1=Female) and/or Appropriate Ranges (e.g., age > 18 years).
- Data files not suitable for data analysis because of data quality or data organization issues will be returned and/or the cost
quotation will be revised.
- Data from surveys/questionnaires will only be analyzed if the survey/questionnaire has been validated.
- Trainees
- A part of the trainee experience is to gain knowledge in research methods. Statistical methods are a key part of quantitative research.
- The trainee's mentor/supervisor is required to attend at least the initial meeting with the BCG consultant.
- Trainees are expected to actively conduct the data analysis, including numerical and graphical summaries, basic statistical tests (e.g., t-test, chi-square), confidence intervals,
and linear regression. Advice on the appropriate methods can be obtained from BCG consultants.
- Ethics
- Proof of Ethics approval is required before statistical analysis can proceed.
- The BCG will not participate in any project that amounts to a misuse of statistics and/or an abuse of statistical ethics.
- Manuscripts, Presentations, Publications
- It is expected that projects will lead to presentations and peer-reviewed publications.
- The BCG will provide numerical and/or graphical result summaries for publication.
- Far too often abstracts are submitted and a subsequent manuscript is not submitted to peer-reviewed journals.
- Researchers are required to give the BCG consultant at least 4 weeks notice of their intention to submit an abstract.
This time frame will provide the BCG consultant an opportunity to assess if the timeline is feasible to provide valid information for the abstract.
- Within 1 year of abstract submission, the researcher is required to have a manuscript submitted to a peer-reviewed journal. Ideally this manuscript will be drafted by the time the abstract is submitted.
- If manuscript submission does not occur within the time frame, work on the project will cease and not be reinstated until a draft manuscript is produced. In addition, work may cease on all other projects with the same principal investigator and new projects may not be accepted with the same principal investigator.
- Researchers need to provide the BCG with electronic or hard copies of each presentation and/or publication which received support from the BCG.
- Authorship
- The BCG follows the International Committee of Medical Journal Editors statement on authorship (International Committee of Medical Journal Editors, Uniform requirements for manuscripts submitted to biomedical journals. JAMA, 1997. 277(11):927-34).
- Each author should have participated sufficiently in the work to take public responsibility for appropriate portions of the content. One or more authors should take responsibility for the integrity of the work as a whole, from inception to published article. Authorship credit should be based only on:
- substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; and
- drafting the article or revising it critically for important intellectual content; and
- final approval of the version to be published.
Conditions (I), (II), and (III) must all be met.
- As an author, the BCG consultant must provide final approval of any manuscript, abstract, etc before submission.
- BCG consultants and/or BCG Director meeting the authorship criteria are expected to be included in the list of authors. Authorship criteria of individual journals should also be discussed with the BCG consultant
- The name of a BCG consultant must not be given any manuscript, abstract, etc. without the consultant's consent. The same policy applies if the consultant is
to appear in the acknowledgements.
- BCG consultants reserve the right to refuse authorship if the document does not meet standards judged to be acceptable.
- Payment or other compensation for statistical consulting activities does not exempt the BCG consultant from authorship consideration.
- Typically, the most senior statistician on the project is the second author.
- Reporting
- BCG consultants will provide memos on various stages of the project. It is the responsibility of the researcher identify any issues or concerns with the information in the memos. Researchers may be required to sign-off on memo content before work proceeds.
- Generally, an early memo concerns data issues that need to be clarified/corrected (data cleaning). Work will not proceed until those issues are resolved.
- Statistical Attestations
- Some journals require statistical attestations to be signed by statisticians for manuscript submissions. These attestations may be based on the soundness and validity of the quantitative methods and/or the validity/integrity of the results.
- Statistical attestations will only be signed by BCG statisticians if the data analyses have been conducted by the BCG and the results and interpretations are judged to be valid.
At the initial meeting, you will be asked to sign and date a copy of this web page to indicate that you agree with the above policies.
| Printed Name: |
| Signature: |
| Date: |
|
