A key piece of the graduate seminar class is writing discussion questions. Your cohort will probably complain about this until they are done with classes (at least I will), and every professor will have wildly different expectations for how they want these questions written or addressed, but there are some underlying themes that can help you write good questions for your social science graduate seminars.
The purpose of discussion questions is to stimulate conversation among a diverse group of scholars. You and your classmates are all studying different topics and ideas, so it’s important that questions help you talk to each other, rather than past each other.
To me, a good discussion question has a variety of answers that can be debated and considered. They should not be simply clarification questions or basic critiques (such as “the study had a small sample size” without further justification on why that might be important for this type of study or research question).
You should be prepared to answer the questions you pose. Students are often hesitant to engage in discussion (out of fear of being wrong, but that’s why good discussion questions have many answers!), so be a good colleague and help start the conversation!
Below, I break down different types of questions with some examples that I have used in past courses. The questions posed here should guide your thinking on research and help you come up with more specific examples. I also made a PDF cheat sheet with just the types of questions so you can reference while reading for your seminar (scroll to the bottom of this post)!
Category 1: Assumptions, Definitions, and Conceptualizations
One way to start a discussion is to ask about an assumption, definition, or conceptualization posed by the author(s) of one piece. You can also juxtapose concepts across papers. For example, one author might understand the word “affordance” differently from another, and you could ask about how that changes the interpretation of their results.
Assumptions: Questions about assumptions consider what the author(s) take as truth in order to conduct their study or interpret their results. This can be tricky to think about, and my example question details one way to approach this type of question, but asking a question about an assumption the author is making usually prompts discussion about the relevance of a study to the field or society at large.
What does the author assume must be true in order to conduct their study or interpret their results?
Definitions/Conceptualizations: Comparing and contrasting definitions and conceptualizations across readings can help you get deeper into an area of research and start identifying some of the crucial, yet sometimes opaque arguments happening in your field.
How does the author define key constructs in their study? Are their definitions cohesive with others in the field? What is assumed or lacking from this definition?
Do two authors conceptualize the same concept differently? Why might this be an important distinction? How might we interpret a set of results differently if we change the conceptualization of a construct?
Example Question for Category 1
One issue that has come up in our readings throughout the quarter is the idea that many Americans do not read the news (e.g., Wojcieszak et al., 2021; Wojcieszak et al., 2022). The underlying assumption with this concern is that engagement with news is beneficial to society at large. Though this assumption is somewhat substantiated, how can we empirically test this idea? Or is it necessary to test this assumption before continuing to conduct research on interventions to increase news consumption?
Category 2: Study Design
Asking about a study’s design can take many forms. You might raise questions about the sample, measures, or the study design generally.
Measures: If you read empirical articles, you can ask about measures from one or more papers. These questions usually refer to the appropriateness of a measure to assess a construct.
Does the author's conceptualization of a construct match how they measured it (operationalization)?
Was the measure used appropriate for the population studied? How could this measure be improved?
How might the authors use additional measures to gain more information?
Sample: Questions about sample sometimes come off as surface-level, but they can also be posed in a thoughtful way. If you choose to ask this type of question, it can be important to point to other research on why there might be differences in findings based on sample, and you should be prepared to discuss what differences might be found. You might also ask if differing samples explain differing results from two or more articles. Again, these questions should have theory or other research brought in to help guide the conversation.
Was the sample appropriate in terms of size and demographics (such as age and location) for the research question?
Who is not included in this study? Why might data from this population be important?
What might have been the challenges of recruiting this sample? How did the authors work within these limitations?
Study Design: General questions about the study design can prompt great discussion. As social scientists, we are often trying to understand human behavior, but we put that human behavior into artificial contexts, such as in lab experiments, or when we do examine human behavior in more naturalistic settings, we may overlook confounding variables. These types of questions are important to discuss, and they can better inform your own research.
Was the setting of this study appropriate for answering the research question? What additional studies could be conducted to help us understand the results?
Were the conditions of this study commensurate? What could have been changed to better control for confounding variables?
Thinking about two or more studies, what do we learn about the phenomena studied? Do these studies inform each other or confuse our understanding further?
Example Question #1 for Category 2
Chein and colleagues (2011) use fMRI to examine how reward centers in the brain relate to risk-taking behaviors for teenagers in the presence and absence of peers. What are some offline* measures that could be used to understand these relations? What are the limitations of using fMRI in understanding these relations?
*Here offline refers to measures that do not involve neuroimaging.
Example Question #2 for Category 2
Category 3: Analysis and Interpretations
One of my favorite types of questions to ask deals with results and interpretations. Asking about if analyses were appropriate or if the author is interpreting their results correctly can spark important conversation about statistics, theory, and even method.
Analysis: Asking about analytical approaches often requires a working understanding of the methods in your field (mine is primarily quantitative), but if you and your peers have this knowledge, you can spark thoughtful discussion.
Were analyses appropriate to the research question? Or did the analyses answer a different question?
Looking at the raw data or correlations (if provided), what might the authors be missing from their analyses?
Interpretations: Recently I've started to ask questions about how me might understand results under different theories or under different assumptions. This type of question allows people to talk about their own area of research to discuss potential explanations for a finding (or series of findings).
How might we interpret these results under another theory?
Based on the authors assumptions, they interpret their results in one way, but if we make another assumption, how might we interpret the results instead?
Example Question #1 for Category 3
Grinberg and colleagues (2019) discuss findings around sharing of fake news on Twitter during the 2016 presidential election. Findings demonstrated that minuscule amounts of users were exposed to the majority of fake news and even fewer accounted for the majority of shares of this news. These users, however, were more likely to be conservative, older, and highly engaged with this type of political news. Grinberg et al. do not assess why these groups might be more likely to see and share fake news. In the discussion, they speculate about the potential of cognitive decline, low media literacy, stronger motivated reasoning, or potential cohort effects. Guess and colleagues (2019) also find that older individuals are more likely to share fake news on Facebook. In contrast to Grinberg, however, users on Facebook who shared the most content were also less likely to share fake news (potentially demonstrating high levels of media literacy).
How can theory help us answer this question (why older and more conservative individuals are more likely to share fake news? Are there existing communication theories that we could use to test this phenomenon?
Example Question #2 for Category 3
Category 4: Implications and Discussion
Question about implications and discussion can take many forms. Policy or practical implications fall under this category, but so do question about designing future studies. These questions can be particularly relevant when two or more studies have similar findings.
What are the policy implications of this research? How can research in one area be readily applicable to current events?
The authors provide an explanation for their findings, but if they do not explicitly test this explanation, how might other researchers design a study to test these relations?
Example Question for Category 4
Eccles and colleagues (1993) outline developmental mismatches between middle school learning environments and adolescent developmental needs. This article was published 30 years ago, but many educational practices remain the same, even if there are minor curricular differences. What are the major cultural and systemic barriers preventing major education reform? What research arguments would be most persuasive to driving this type of reform?
Category 5: Strictly Theoretical Pieces
To me, the hardest articles to write discussion questions for are those that are purely theoretical or do not contain any empirical studies. These papers can be very important in terms of guiding the field, but they can be hard to discuss, especially as a first year graduate student!
Assumptions and Framing. Once again, you can ask about assumptions of a theory. You can also ask about how the theory frames our understanding of the world.
What does the author assume to be true in order to present their arguments?
What or whom does this theory center? What would be different if the author centered something or someone else? Does another author address this with their theory?
Does this theory explain phenomena on a micro or macro level?
Boundaries. Another issue with regard to theories is to point out boundary conditions. These are scenarios or contexts in which the theory might not apply.
What are the boundary conditions of the theory? How would we test these conditions empirically?
Has something new happened that could affect this theory? What about our current historical context limits our understanding of this theory?
Example Question for Category #5
Bennett and Iyengar (2008) also discuss shifting audiences, such that how mass media has been disseminated and consumed has changed and displaced by personalized content. How can communication scholars theorize about these somewhat recent advances in technology (such as algorithms tailoring personalized feeds of content to individuals) as well as longstanding technologies (such as radio, e.g., Adena et al., 2013) in tandem?
Of course, you are not limited to the questions I have outlined here, and your instructors may require something else from you, but I find these questions to be helpful in guiding my thinking and preparing for my graduate seminars. Good luck with your courses!
PDF Cheat Sheet of Discussion Questions:
Comments