What is Statistical Consulting?
Statistical consulting is an essential part of the Quantitative Sciences Group's (QSG) objective to empower researchers with quantitative tools for all aspects of research from inception to completion. Researchers are encouraged to meet with a statistician during all phases of the research project including the design, data collection, analysis, interpretation and presentation phases. Statistical consulting is most effective when clients work with statisticians from the inception through the conclusion of a study. However, with an accurate understanding of the study design, consultants can become involved after a study has been started. Statisticians cannot provide consulting on analyses that are inappropriate for the study design.
Common consulting topics
- Consultation on study design at the start of a program of study, sampling design, scope of inference, and other issues relevant to future statistical analyses.
- On-going consultations on statistical techniques during the analysis phase of a program.
- On-going consultations on interpretation and inference, and presentation of results.
Statistical Consultants can also:
- Review text concerning statistical methods and study design.
- Guide clients to existing literary resources and course offerings that will provide background that the client will need to address statistical issues in their program of study.
- Conduct statistical analyses for researchers under a mutually agreed upon arrangement. Usually this arrangement is either contractual or involves authorship. This service is not available for any work related to a degree program.
- Provide programming assistance with R, SAS, and other statistical software, as well as guidance on reproducible data analysis with git and bioinformatics using HPCs.
Who is eligible for statistical consulting resources?
- All faculty and staff of the College of Forestry. This includes all professorial faculty whose home department is in the College of Forestry as well as all professional faculty, FRAs, Research Associates, Research Scholars, Visiting Scholars, Instructors, and adjunct or joint faculty members.
- Graduate students in any College of Forestry degree program or whose major professor qualifies under any of the criteria above.
- Any T/TT professors who are the major professor of a student in a College of Forestry graduate degree program.
- Undergraduate students working on a research project under the direct supervision of any College of Forestry faculty or staff.
- Computer users who have accounts on the College of Forestry computer network and who have subscribed to the statistical consulting resources.
What does it cost?
- Statistical consulting is available at no charge to all faculty, staff, and students listed above. **This does not apply to courtesy faculty.**
- Statistical consulting is available to College of Forestry Computer network users for an annual fee, related by the Extended Service Fee in the Fee Book. Anyone interested in fee-based consulting should contact Dusty Gannon.
How does it work?
An eligible user should contact the Research Consulting Statistician by email (dustin.gannon@oregonstate.edu), phone (541-737-1491), or in person (PFSC 317). Meetings will usually last 30 to 60 minutes and initial meetings are largely focused on the motivation of a study and less on statistics. Statistical consulting is most effective if the consultant understands the system and scientific motivation of the research. Following initial meetings, the consultant will organize the notes from the meeting and have the client confirm that the consultant understood the objectives of the research and details of the system correctly. Statistical recommendations and considerations will follow in written format and be reviewed in a follow-up meeting.
What to bring to meetings
The content of meetings will vary depending on the client's motivation for seeking statistical consulting, but the effectiveness of initial meetings can be enhanced by bringing materials such as:
- Well-formed scientific question(s) that are testable and precise.
- Diagrams or written descriptions of any prior data collection protocols.
- Patience. Consultants need time to get up to speed with your awesome research before giving effective statistical advice.
Consultants
Dustin Gannon, Research Consulting Statistician
Email: dustin.gannon@oregonstate.edu
Phone: 541-737-1491
Website: dusty-gannon.github.io
Education:
MS Statistics, Oregon State University
PhD Ecology, Oregon State University