One of the responsibilities that we have in working with students on research projects is to show them how to create a study to tackle questions that emerge from the literature in our discipline. I work with a number of undergraduate students in my lab each term, and many of these students are just beginning to understand the scientific method and operationalizing variables in the lab. They typically sign up for independent study credit or are volunteering to gain research experience to help inform their possible future directions. When undergraduates are learning how to conduct research, training them to follow lab protocols for participant recruitment and experimental sessions requires our careful attention. For those studying gerontological research questions in perception, cognition, or social psychology, we inevitably turn to stimulus presentation packages to implement our research designs. I have been training my student assistants to use a number of these platforms, and I have been impressed by how quickly they can acquire some coding skills to aid in the design and management of research projects without having any prior experience (for lab course example, see Hamilton & Geraci, 2004).
Most stimulus presentation platforms include sample experiments that students can review to learn about the basic rules and flow of the coding language or graphic user interface of the packages. For instance, common platforms like E-Primeopens in new window, PsychoPyopens in new window, and PsyToolkitopens in new window provide libraries of experiments designed by platform users. Moreover, samples of code for classic experiments in psychology are often cataloged on websites that support engaging students in experimental psychologyopens in new window, and manuals exist with how-tos for basic designs (e.g., Peirce & MacAskill, 2018 for PsychoPy). Although training students to use these platforms can seem daunting at first for faculty and students alike, students quickly catch on if they have consistent access to the platform to practice and if a mentor sets up assignments with reasonable incremental goals.
In the past, I might start the training process by reviewing a journal article with the students that includes a task that I have used in the past. We draw out a storyboard for the task. Then I show the students the program that was used to collect data. During our first pass through the program, I focus on the main functions of the program, like presenting an image and then asking for a response. Then I build outward and talk about the ways that the software cycles through trial lists to pull relevant stimulus features from the list to plug into placeholder variables to control the display. We make small changes to the code so they can see how trials change within the task. After reviewing the program, I ask the students to consider a specific change to the code that we have not yet discussed and let them explore the features of the stimulus presentation platform. I also ask students to rebuild the program from scratch, referring to the example program to debug their work.
As professionals, when we code, we know that there is trial and error involved. I don't hide my initial mistakes from the students as we work on a project together because the creation process is quite rewarding despite the occasional roadblock that we face in trying to get the program to do what we want. Moreover, my students will work on the same code together across multiple lab visits, discussing possible solutions to errors that emerge. When they do this, they are also learning about some of the limitations that we face when we design tasks (e.g., conditional statements and loops that do not work, refresh rate problems, etc.).
In the face of the COVID-19 pandemic, work with my students has shifted from face-to-face training sessions in the lab to virtual training sessions where we use screen-sharing to review code, to make changes to the code together, and to explore the data that emerge from the stimulus presentation platform. At this time, we are using PsyToolkit (Stoet, 2017) because it is a free resource that all of my research assistants can access remotely. It also can be used to collect data from participants remotely (e.g., Sona Systems, Mechanical Turk, or Prolific). Students use the library of experiments and self-directed tutorials to learn some of the basics. Over time, as students develop code, they learn about new features of the platform that I have not yet discovered, so I have also been learning from their successes as we work together on projects.
Inasmuch as undergraduate research assistants are vital to our research productivity and program success (Stoloff et al., 2012), we are vital to their professional training (Perlman & McCann, 2005). Focusing students on one or two examples of a research protocol can serve as a means to provide an orientation to using stimulus presentation software for data collection. Creating a small to-do list of possible modifications to an existing task program can build students' experience and allow them to work together to solve coding roadblocks. Lastly, by engaging students in the process of experiment coding, they will learn more about operationalizing variables to address a research question. The coding skills that we are cultivating in our students will help empower them to ask their own research questions for years to come.
About the author
Andrew Mienaltowski, PhD, is an associate professor of psychological sciences at Western Kentucky University. He is the current undergraduate program coordinator for the department of psychological sciences, and his research focuses on changes in the way that emotions are perceived in adulthood.
References
Hamilton, M., & Geraci, L. (2004). Converting an experimental laboratory course from paper and pencil to computer. Teaching of Psychology, 31(2), 141-143. doi: 10.1207/s15328023top3102_7
Peirce, J., & MacAskill, M. (2018). Building experiments in PsychoPy. London: Sage.
Perlman, B., & McCann, L. I. (2005). Undergraduate research experiences in psychology: A national study of courses and curricula. Teaching of Psychology, 32(1), 5-14. doi: 10.1207/s15328023top3201_2
Stoet, G. (2017). PsyToolkit: A novel web-based method for running online questionnaires and reaction-time experiments. Teaching of Psychology, 44(1), 24-31. doi: 10.1177/0098628316677643
Stoloff, M. L., Curtis, N. A., Rodgers, M., Brewster, J., & McCarthy, M. A. (2012). Characteristics of successful undergraduate programs. Teaching of Psychology, 39(2), 91-99. doi: 10.1177/0098628312437721
