Welcome to the Statistical Consulting program of the Department of Radiology Research Support Core. The purpose of this program is to provide accessible statistical services to Radiology faculty, fellows, residents and staff, which include assistance with
study and experimental design
sample size calculations/power analysis
choosing and applying appropriate statistical methodology
interpretation and presentation of results
grant and manuscript review
other statistical questions.
These services may be utilized from the early planning stages of a new study to the preparation of grant proposals to the analysis and presentation of results for publication. In particular, meeting to discuss a new study before any data is collected is recommended, if feasible. This can help identify the most efficient way to collect the data and not miss important variables.
Unless you have previously met regarding your present project, it is highly recommended that you first schedule a meeting in order to sufficiently cover the goals and background. This is to ensure the highest quality advice, which depends strongly on context.
Meetings times are generally flexible with Tuesday and Thursday being the most open days.
Actual availability varies week to week, so please confirm your desired meeting time through email. Unless you have made other arrangements, such as your individual office or a teleconference, the following locations will typically be used for meetings:
UWMC – BB 308
HMC – Radiology library
SLU – Any available meeting room
Other arrangements can be accomodated on a case-by-case basis.
Depending on demand, the degree of hands-on assistance that will be provided will vary. To prioritize requests, deadlines, complexity of analysis, software availability, and skills of the investigator will be taken into account. Making requests early is recommended.
Co-Authorship and Acknowledgment
If substantial contributions (e.g., study planning based on statistical estimation, statistical planning, data analysis, etc) were made by the consulting statistician, and if that is acknowledged by the Principal Investigator, the consulting statistician should be entitled to a co-authorship in a publication and will have the commensurate responsibility for the publication as a whole. If the contributions were useful but do not rise to the level of a co-authorship, we simply ask for acknowledgement in the publication.
You may send data to firstname.lastname@example.org when you are ready to get help analyzing it. Typically an initial meeting is required to explain the project and variables meanings. When you send data, please indicate any confidentiality issues and if the data should be destroyed after a particular time. In addition, please remember to remove any protected health information (PHI) like patient names, addresses, etc.
Other questions, comments, and feedback may be directed to the Vice Chair, Research, Dr. Paul Kinahan PhD (email@example.com).