By Ethan Van Norman, Lehigh University
Many topics on this blog are dedicated to helping early career faculty identify, develop, and sustain a productive program of research. As with most things in academia, navigating this process seems to become easier with time as we learn strategies and habits to remain productive and free up our time. However, there is one arguable advantage that researchers just out of graduate school hold over more experienced faculty. Recent graduates are likely to have received the most up-to date training in cutting-edge statistical analyses. The more seasoned veteran may have not have run a given statistical test for well over half a decade or are using an expired SPSS license while the newest hire is explaining how R works to the rest of their colleagues. Fortunately, there are ways to ensure that your statistics muscles do not completely atrophy. The purpose of this post is to describe potential strategies, and associated considerations, to keep up to date with your quantitative training.
Paid Seminars and Classes
Demands from associate editors and reviewers for sophisticated quantitative analyses seems like an ever rising tide. Given that many researchers lack training in analytic approaches only recently made accessible via commercially available software for personal computers (e.g., Bayesian analysis, multilevel modeling, item response theory, etc.), many companies have stepped in to offer paid trainings on these analyses in a condensed class format. Often, you may required to pay upwards of multiple hundreds to thousands of dollars to enroll in a class to learn a given set of analyses. Course formats can be face-to-face, online, or hybrid. One such provider charges $1,295 to participate in a three day face-to-face training. Depending on your geographic location, there may be other options that are less expensive. For instance, there are training options that last two days face-to-face and start around $900. Sometimes you can receive a discount (sometimes upwards of $300) if you attend the training online rather than face-to-face. Another option is to check university websites as summer approaches, as many institutions offer statistical training over a 1-2-week period during breaks. Briefer trainings are frequently available in conjunction with annual meetings and conventions.
In terms of paid online courses, many of the providers listed above will offer access to previously taped trainings or offer live-streaming synchronous options if you are unable to attend in person. A google search provides other online-only options for training.
Things to Consider
First, it is advisable to consider costs other than registration fees to attend such training, such as airfare, hotel, meals, and if applicable missed time from work when making your decision. Further, if the analyses you want to learn require specialized software, find out whether the software will be provided at the training or if you are expected to have purchased and downloaded the software prior to arriving. Some software packages cost more than the laptops you’re reading this post on. Using start-up funds to cover these costs may be a wise investment if the software is not something your unit, college, or university provides. Also, be sure to verify the level of familiarity with the software needed to participate in the training.
It goes without saying that paid seminars and classes are expensive! Therefore, it is in your best interest to make sure that the training is worth your money and time. The best advice is to talk to colleagues about their impressions of trainings they have completed (please comment below!). In addition, it may be worthwhile to search reviews or message board posts about attendees’ experiences. Be sure to seek out information from websites unaffiliated with the company as a common marketing tactic is to selectively post positive testimonials from customers. Relatedly, make sure that the training you are considering is appropriately geared towards social scientists or educational researchers. Statistics training is a rather nebulous term and any topic can be approached with different levels of breadth and depth. While it may be nice to understand the matrix algebra behind a new-to-you analysis or to delineate each assumption associated with a given probability density function, this information may not be crucial to your needs. Ensure that you will be equipped to carry out the research you wish to pursue after completing the training.
For any training that you pay to attend, make sure that course materials will be made available to you after the training has ended. In the midst of the all-day training you may feel extremely confident that you’re absorbing all of this new information without a problem. Once you come back to reality and are pulled in 100 different directions, it may be difficult—if not impossible—to retain all that new knowledge. Being able to revisit lecture notes, activities, and videos from that seminar will be invaluable to ensure you retain everything you have learned.
Auditing/Sitting in Classes
Another option may be to sit-in on a class being taught by a colleague at your university. This may be beneficial in that the content of the course is not condensed into a few rapid-fire daylong classes. As one can imagine, this option is dependent on you having a positive relationship with the instructor of the course. In addition, there may be other institutional considerations within the department or university at play. Generally, you can be clear this up by having a discussion with the instructor of the course.
Things to Consider
It goes without saying that one of the biggest perks of our jobs as instructors is academic freedom. To that end, it is imperative that if you are to sit in on someone’s class you (a) ask them beforehand, (b) not take up space or resources that would otherwise be used for students paying tuition, and (c) not act in an evaluative manner regarding the instructors teaching. A little bit of perspective taking can go a long way. Ask yourself how you would feel if a colleague sat in on your course for an entire semester and act accordingly. If you are a more senior faculty member, recognize the power differential that may be at play if you make the request to a newer colleague. While they may have the option to decline, they may feel uncomfortable doing so to someone that holds a vote as to whether they will keep their job in the next few years.
Another consideration relates to the method of delivery. Be sure that you have the time to dedicate to the course throughout the entire semester (to both make the scheduled class time and work needed outside of class). The optics of treating a class that your colleague has invested substantial time and energy into as a drop-in service can be a very bad look. In addition, if you truly want to get the most out of the course, be prepared to do any reading and assignments given to students. While it may not be possible to receive graded feedback from your efforts (again, think about how you would feel if you had another paper to grade from a non-student), actually practicing the skills being demonstrated is critical.
A third option is to engage in self-study. There are no shortage of textbooks covering novel statistical analyses. The most straightforward method would be picking up a book and working through it. Building upon the previous point, many faculty are willing to share a reading list or syllabus regarding an analytic method if attending a class is not feasible. Having an annotated bibliography can also help in conducting a self-assessment to determine where you may have more or less conceptual and procedural knowledge about a given approach.
Things to Consider
The feasibility of engaging in sustained self-study is almost entirely driven by your level of personal accountability. Although you may buy a book with the best of intentions, without a clearly defined schedule for completing readings—and intentionally blocking out time each week to complete those readings—it is very likely that it will end up on your shelf collecting dust. In addition to finding a book geared toward a social science or educational audience, it may be beneficial to seek out textbooks written with the intent of self-study. That is, some authors have written resources in the prologue or foreword with recommended reading order and practice exercises with answers to assist in mastering the content. However, in this case, it is critical that you actually do the exercises. Building upon the broader open-science movement in psychology and educational research, some authors will record and distribute lectures from classes that the textbooks are based on for free. Again, creating and sticking to a pre-determined schedule to watch each lecture on a specified date and time will increase the probability of following through with the endeavor.
Another strategy could be to engage in a book club in which colleagues with similar interests also read and work through the book on a common schedule. Having regular conference calls or discussions can help maintain accountability.
All of the options discussed thus far have a common shortcoming: each lack an expert instructor to actively monitor your mastery of the content and provide corrective feedback. The reality is, absent actually enrolling in graduate-level statistics classes, you are unlikely to be exposed to such a highly structured environment with frequent feedback to learn a new statistical analysis at this point in your career. For the vast majority of the readers of this blog, their PhD is their terminal degree. Therefore, before undertaking any endeavor to broaden your statistical toolbox, it would be wise to critically evaluate how acquiring this new skill will help further your career. That is, you should weigh the potential return on investment of learning a new statistical strategy. Will this new set of skills help maintain or propel your research agenda? Or is the desire to learn a new strategy driven by a one-off study? If the reason is the latter, your time may be better spent seeking collaborations with colleagues that have expertise in that analytic strategy. It goes without saying that being able to read, understand, and execute the statistical analyses underlying your research is crucial. However, it is important to remember there are multiple ways to contribute to a research project. What have been your experience with the approaches discussed here? Are there other factors to consider? Feel free to discuss these points in the comments section below!