The mission of the Biostatistical Collaboration and Consultation Core (BC3) is: (1) to be a high-quality resource for robust, reliable, and reproducible statistical support; and (2) to educate and train the next generation of collaborative statisticians.
The BC3 offers a wide range of statistical support. We work with investigators at every stage of the research process: proposal and grant development; hypothesis generation; study design; data validation and cleaning; statistical analysis; interpretation and visualization of results; and even addressing referee comments for a resubmission. We support projects either through short-term periodic consultation meetings, or through long-term collaboration with our faculty and students. Learn more about the types of support we offer here.
Initial consultation meetings are free of charge. We love hearing about all kinds of research, and are happy to offer statistical advice. The end goal of a consultation meeting is to decide what type of statistical support is best suited to your project. Fill out our short project initiation form, and we’ll reach out to set up a time.
Learn more about the BC3:
- Types of Support
- Values and Policies
- Student-Run Consulting Lab
- How to prepare for your initial meeting
- Previous Work
Initial consultation meetings are free of charge. In the initial meeting, we will work together to define the type and scope of the statistical work needed for your project, and agree on a plan of action.
The BC3 offers 4 types of support for your project:
- Grant support: We provide grant preparation support at no cost, though it is expected that the statistician is invited to be included on the grant for an appropriate percentage effort.
- Long-term statistical partnership: Departments or investigators that have regular need for statistical support can “subscribe” to a portion of a BC3 statistician’s time, like keeping a lawyer on retainer. Over the course of the relationship, the partnering statistician can collaborate or consult on many different projects. We can work with you to arrange a percentage effort and subscription period that meets your needs.
- Short-term statistical support: For shorter projects, we offer statistical support at an hourly rate. We can work with you before starting the project to provide an estimate of the number of hours of work involved. We can reduce or waive the hourly rate on a case-by-case basis.
- Consulting Lab: Finally, there is the option to work with the students in our Consulting Lab, which we provide at no cost. Small teams of graduate students will work under the supervision of our faculty to provide statistical support. The Consulting Lab runs year-round, and is offered as a for-credit course in the Spring semester. If you know you want to work with the Consulting Lab on your project, you can fill out the Consulting Lab form here.
Values
- Consultation meetings are free and welcomed: Initial consultations for all projects are free of charge. We can advise on the statistical feasibility of a study, answer any immediate statistical questions, and discuss options for working together on a project. The end goal of a consultation meeting is to determine what type of statistical support is best suited for any given project.
- Team Science: Statistical thinking is relevant to decisions made at every stage of research, and collaborative statisticians are therefore most effective when they are deeply embedded in a project from its inception. We aim to be able to contribute meaningfully to projects at all stages of development, including: honing research questions; designing the study; collecting and cleaning the data; conducting statistical analysis; interpreting, communicating, visualizing, and disseminating the results.
- Equity: Data Science can be a powerful tool for fighting healthcare inequity and addressing social determinants of health. However, it is also important to acknowledge that some early naive applications of statistical methods in healthcare have had the effect of multiplying discrepancies. We will seek to be attentive to these issues, to prioritize questions of equity in our analyses, and to be diverse in the perspectives we gather.
- Rigor & Reproducibility: Projects that we work on will embrace principles of open, rigorous science. Before doing any substantial analysis, we will prepare a detailed analysis plan (which we recommend pre-registering); after completing an analysis, we will share code and de-identified data (when possible). Careful analysis takes time, and we will only work with collaborators who can accommodate this. (See policy on timeline for requests.)
- Clear Communication (including data visualization): Clear and concise oral, written, and visual communication of statistical results is an important skill. We consider it to be a fundamental component of the statistician’s job.
- Practical experience: There is no substitute for learning by doing. We have an educational mission and we seek to make practical research opportunities available to as many early career statisticians as we can.
- Self-study and iteration: We aim to be nimble and iterative as we evolve. We will collect feedback from collaborators early and often, and apply the same scientific standards we use in research to understand how we can grow and improve.
Policies
- Timeline for Requests: Our preference is to be involved as early as possible in the research process. Grant proposals should ideally be discussed at least 6 weeks prior to the deadline (but we recommend coming to us as early as possible!). For projects with less than 6 weeks of lead-time, our support may be limited, but we still recommend having a consultation meeting to discuss what options are available. For data analysis projects, timing varies depending on the level of complexity, but projects typically require at least a few months from start to finish.
- Rigor and Reproducibility: As much as possible, projects will begin with pre-registration of a Statistical Analysis Plan (SAP), and will end with publicly sharing de-identified data and analysis code.
- Working with early career scientists: Training the next generation of biostatisticians is fundamental to our mission. As much as possible, we will work with Master’s and Ph.D. students, and other early career statisticians.
- Authorship: We follow guidelines established by the International Committee of Medical Journal Editors (ICMJE) for authorship criteria (LINK). All statisticians who satisfy these criteria must be offered authorship on the resulting manuscript.
In the Consulting Lab, our graduate students (under the guidance of our faculty) work in small teams to provide statistical support for active research projects. The Lab runs all year, but in the Spring semester, it operates as a credit-bearing course, where students learn about team science and put it into action on research projects. New projects for the Lab are always welcome! There is no charge for projects that are submitted to the Consulting Lab.
To inquire about participation in current active projects, or for any other questions, contact Evan Wardell at evan.wardell@nyu.edu. Investigators may submit their projects to be scheduled for these Consulting Lab sessions using our request form.
Current Projects
27 of our graduate students are currently providing statistical support on 9 exciting studies by investigators across the NYU community (with even more to come). Below are some of these projects:
- "Trauma Services: Understanding Time from Referral to Service Initiation", with lead investigator Kate Guastaferro, PhD of the NYU GPH Department of Social and Behavioral Sciences
- "Sun Protective Behaviors Study", with lead investigator Christine Olagun-Samuel of the NYU Langone Department of Dermatology
- "Long-Term Continuation Rate of Botox for Overactive Bladder", with lead investigator Anjali Kapur, MD of the NYU Langone Department of Urology
- "Investigating the need for a Health Equity Curriculum within the IM program", with lead investigator Alyssar Habib, MD of the NYU Langone Department of Internal Medicine
- "Exploring the association between gestational diabetes mellitus and food insecurity", with lead investigator Veronica Pasha of the NYU Rory Meyers College of Nursing
- "Understanding loss to follow-up in a large phone survey cohort in Kenya", with lead investigator Corrina Moucheraud, ScD of the NYU GPH Department of Public Health Policy and Management
- "HealthRhythms", with lead investigator Anna Van Meter, PhD of the NYU Langone Department of Child and Adolescent Psychiatry
- "Validation of the IN-HOME Tool", with lead investigator Jasmine Travers, PhD of the NYU Rory Meyers College of Nursing
- "Understanding and modeling collaboration networks across different academic fields", with lead investigator Wen Zhou, PhD of the NYU GPH Department of Biostatistics
Products of Past Consultations
Several publications have resulted from projects in collaboration with the students in our consulting lab; you can see some of them here.
These meetings will start with us learning as much as we can about your research question. We will ask you to explain the premise of the study (the motivation and background science), the research question, the study design and limitations, and what methods have been employed. It’s important for a statistician to understand the problem in detail in order to be most effective.
It is normal not to have all of these topics clearly defined in the early stages of a research project, and we can work with you during the consultation to hone the research question. If you have any materials, such as a short slide presentation or a draft of your aims page, those will be very helpful to share in advance.
After this introductory meeting, we will work with you to make a plan for how to proceed. Click here to learn more about what types of statistical support are available through the BC3.
- “Kupewa: Optimizing Strategies to Implement Provider Recommendation of HPV Vaccination for Adolescent Girls and Young Women with HIV in Malawi”
- “Elucidating the Link Between Oral Health and Dementia Subtypes: A Multifaceted Study on the Biological Pathways and Social Determinants"
- “The Stroll Safe Randomized Control Trial: Program Effects on Outdoor Falls Self-Efficacy”: https://osf.io/g8ncj
- “Risk factors associated with Barth Syndrome”: https://osf.io/5wf4m
- Alex's OSF page: https://osf.io/8kz3j/
- "Power Simulation for Factorial SMART Study Designs": https://osf.io/k9epq/
- Alex's GitHub page: https://github.com/alex-dahlen