Skip to content

Week 2

Preparing for Biostatistical Consulting

➤ This week we will explore the role of the statistician in a research collaboration, including how to be prepared for a statistical consultation.

Before Class

Understanding the Research Question

One of the jobs of a study biostatistician can be to help translate a researcher's research question into a statistically testable hypothesis. Sometimes, though, researchers may be uncertain about their own research questions. Our statistical consulting core group recommends researchers use the PICOT framework to construct their research questions. As a data analyst, it will be beneficial if we know these components of our client's research question.

📖 What Is Your Research Question? An Introduction to the PICOT Format for Clinicians (15 minutes)

  1. Use chat GPT to request the following information:

  2. Generate a hypothetical biostatistical consulting scenario.

  3. Use the PICOT framework to break down the researcher's question.

  4. Ask if there are any other frameworks that can be used for the statistician to better understand the client's research question.

  5. Post a short summary of the research question, written in PICOT format, to the course Slack.

Tips for Consulting

📖 4 Tips for Statistical Consulting (7 minutes)

  1. What did Chanel say were the main reasons why she struggled with her consulting project?

Supplemental Readings

📖 Textbook Chapter 1

In Class

Consulting Vignettes

🎥 Realistic Timelines for Work

  1. In this video the researcher wanted immediate assistance. What is one thing the consultant did to reassure the researcher?
  2. If a client came to you with a thesis draft deadline in two days, how would you respond to them?
  3. Some statisticians may find it useful to, say, triple the amount of time you think it will take to complete a project, in order to account for unforeseen circumstances. What considerations may be needed for deciding on a reasonable time frame?

🎥 Beginning and Ending A Meeting Effectively

  1. What did the consultant say was the most important goal in a first meeting?
  2. Why did the consultant wrap up the discussion 10 minutes before the end of the scheduled time slot?
  3. What is one pitfall that could occur if the statistician gave the client specific statistical advice during this first meeting?

🎥 Using Statistical Results to Identify Questions

  1. List 3 things you think the client should have known before entering this consultation.
  2. Given your current or hypothetical future position, would you take on a project like this knowing the grant proposal did not have a budget for statistical consulting/collaboration?
  3. The researcher believed that “statistical magic” could be used to determine which variables are the most important. What challenges could arise from this view? What might you do to discuss this topic with potential clients?

Reflection

Consider the consulting strategies we saw in class today. In light of this, which aspects of statistical consulting do you feel more comfortable with? Which aspects are you less comfortable with?