This column is written for statisticians with master’s degrees and highlights areas of employment that will benefit statisticians at the master’s level. Comments and suggestions should be sent to Megan Murphy, Amstat News managing editor.

Jessica Lavery is a biostatistician in the Health Outcomes Research Group at Memorial Sloan Kettering Cancer Center in New York. She has an undergraduate degree in statistics from Loyola University Maryland and a master’s degree in biostatistics from The University of North Carolina at Chapel Hill.

 
While statisticians are formally taught methodological skills, tactful project management and establishing credibility early on are less often taught, but equally important to successfully contributing to a team. In some settings, such as the medical field where traditional hierarchies drive authority, it may be particularly difficult as an early-career statistician to establish credibility and manage projects with confidence. To build expertise and command respect from the get-go, or to re-work the current dynamic on a project team, several subtle tactics may be used.

At the Women in Statistics and Data Science (WSDS) conference in October 2017, I shared tools picked up during my first few years as a statistician that have proven useful in being taken seriously by senior collaborators. I share them again with you here.

Managing Expectations

At the beginning of collaboration, it is important to establish that the scope of a statistician’s responsibilities often spans beyond producing analyses. The level of involvement by the statistician certainly varies across institutions and may vary across projects and investigators, as well. For this reason, it is most beneficial to establish what your skills are and how you expect to use them at the beginning of each project and on an ongoing basis so your collaborator is aware of how you expect to be involved. Do you expect to be consulted at the beginning of a project, or on an ad hoc basis as the project progresses? Do you anticipate writing part of the manuscript, or to only review the methods and results sections? Do you want to be involved in all project meetings, or only meetings that are statistics oriented? These are examples of details the collaborator should be informed of from the beginning to set the stage for a successful collaborative relationship.

After you have explained your role, it is important to maintain respect for your expertise throughout the project. Often, intending to be seen as having mastered my role, I’m inclined to say, “Sure, no problem, this task will be easy.” This was something I struggled with in my first job out of graduate school. Of course I can tackle any task you present me with; I just got my master’s degree! Not only does this approach minimize the amount of effort required and results produced, resulting in unrealistic expectations in terms of how long it takes to complete a task, it also conveys the idea that anyone can do my job, devaluing my role on the team. Additionally, if a task comes easily, often that’s only because I went through six years of school to learn how to master that skill.

Instead of saying something is easy, it is more beneficial to say something along the lines of, “This project will involve x, y and z, and I’m looking forward to getting started on it.” This conveys competence, confidence, and enthusiasm without minimizing your contributions.

In addition to wanting to seem so skilled at my job that everything was easy, another challenge I encountered was figuring out the right time to send results to an investigator. Naively, my initial approach was to drop what I was doing if it was a quick request and to always send results the moment they were produced. Of course, this sense of urgency would be necessary in the presence of a deadline, but it can lead to setting up unreasonable expectations in the absence of one.

Like saying something is easy, constantly having a short turnaround time implies your job is quick and effortless and you are constantly on call, ready and waiting to immediately handle all requests. Even investigators with good intentions will learn they can wait until the last minute to ask for something, since it takes such a short time to produce. Instead of immediately replying with results, acknowledging you have received the request and intentionally delaying sending the results for a short time helps manage your collaborators’ expectations and provides buffer time for you to digest the results before disseminating.

Collaborating and Communicating

In the day to day, back and forth of collaborating, several subtle language modifications can be helpful in enforcing the idea that you are an integral part of the team.

As statisticians, we are often taught to speak the same language as our collaborators, and this cannot be emphasized enough. Giving an example or explaining an analysis in the same content area as the project prevents your clinical collaborator from having to translate from an abstract content area to statistics, and from statistics back to their content area. As an example, working in oncology, I would want to avoid explaining a methodology based on agriculture plots, even though this is how I was taught a lot of statistics. My oncologist collaborator is then translating ideas from agriculture to statistics to oncology, opening a lot of opportunity for things to get lost in translation. While this may sound obvious, it is something that serves us well to remember.

Using the term “we” instead of “you” ingrains you into the research team. Asking, “What question are we trying to answer?” conveys you are equally invested in the research.

I used to frequently insert “just,” as in, “I’m just emailing to …” or “I just wanted to …” The intention is to be polite and deferential, but what this does is indicate we are in a position of inferiority when we deserve to be an equal collaborator. If I am emailing to follow up with a collaborator who ignored my last email for a week, inserting “just” makes me feel like I am conveying understanding that they are assiduously working and may not have had time to respond. As equal members in the professional field, it is my collaborator’s responsibility to respond to an email, and I should not feel shy about reaching out. I now scan all emails after drafting them and remove unnecessary instances of “just” (which is most of them) and re-read, noting how much clearer the email sounds.

Upspeak is something I only recently learned about, and I could not stop noticing how frequently I did it once I did. The term upspeak refers to ending what should be a statement as a question. As a simple example, if someone asks when a meeting is and you reply, “The meeting is Monday at noon?” Ending this in a question leaves the group with no more information than before you answered. It is especially important to avoid doing this when sharing results with a group, and to instead speak a sentence as a declaration, portraying more confidence in the answer. It is much easier for a collaborator to be dismissive of an idea or response if it is presented as a question.

Authorship can be a touchy subject, especially when the project team is large. Additionally, clinical collaborators are sometimes unaware that authorship is important to a statistician’s career. It is your responsibility to establish early on what your expectations are regarding any abstracts, as well as the final manuscript. Instead of presenting this as a question—“Can I be listed as second author?”—it is more effective to present a statement such as, “Generally, when a statistician cleans the data, runs the analyses, and writes the methods and results, they are awarded second author.” This politely establishes a norm and requires justification by your collaborator if there is going to be a shift.

Collaborating with medical doctors and doctoral-level health services researchers, I often felt conflicted about how to refer to them. Dr. Jones? Ms. or Mrs. Jones? Jane? The trick that comes in handy here is to start formally and then mirror the way someone signs emails. If a collaborator signs off using his or her first name, I take that as an invitation to refer to them as such, connecting on a more personal level and further building a collaborative relationship. Always defaulting back to the more formal name reinforces the idea that you are a subordinate. With that said, if a collaborator indicates a preference for a more formal communication, then it is certainly appropriate to respect that preference.

Last, but certainly not least, it is important to remember that silence is okay. Often, when I present a result or explain a method and a collaborator is processing what I just told them, I continue nervous-talking—speaking rapidly and quickly trailing off into gibberish as I worry something I explained did not make sense. I often fall back on prematurely asking, “Did that make sense?”, pre-emptively suggesting I was incoherent. Instead, it is more constructive to give the collaborator a few moments to process, and then invite feedback by having the collaborator reiterate their understanding of what you communicated.

Thoroughly explaining your role and consistently communicating the value of your contributions both explicitly and implicitly increases your influence on the project team. Of course, not all tips will be applicable in all scenarios, but I hope you will find some helpful. This is an area in which I am constantly learning, and I look forward to hearing from others about tactics that have proven useful.