Last Update: April 30th, 2024
"I have nothing to ask when it is the Q&A session." Do you resonate with that? In previous years, I struggled to develop questions to ask during seminars. I wouldn't say I like the silence after the speaker or the host asks if there is any question. However, when I am more used to the academic setting in the United States, I realize the benefit of asking questions and admire people who can ask good questions during a seminar. I set a goal to ask one question for each seminar I attended in 2024. I am writing this blog post to summarize my attempt and what I learned during this period. This blog post might be biased toward the field of robotics or engineering because those are the seminars I have the most experience with.
People reading this post understand the advantage of asking questions. Still, I will list a few down here to convince myself that this is an essential skill I need to keep working on.
People will appreciate you breaking the silence and encouraging others to participate.
Increase your visibility not only to the speaker but also to other participants (students, faculties).
When I plan to ask questions, I focus more on the talk and get more out of it because I want to ask something relevant and not something they already mentioned.
Learn how to communicate with people in the field and engage in research discussion.
Interestingly, by asking questions, you also learn how to answer the questions by actively observing how the professionals are doing!
For people who are not used to asking questions, make sure to set up the environment to encourage yourself to ask questions.
Sit in the middle front of the room so people will clearly see when you raise your hand.
Sit around the people you feel comfortable with, and you know you won't be judged by the questions you ask.
Prepare 10 minutes before the seminar starts.
From reading the talk title and the abstract, think about what you expect and imagine the talk will be about. If it ends up being different, ask the speaker about their opinion on the scope you thought of.
Look up some key terminology beforehand in the talk title that you need help understanding, especially if the talk is not in your primary research focus. For example, I am unfamiliar with the manipulation area, and English is not my mother language, so I have to search "dexterity" and "affordance" online to ensure I grasp the idea correctly.
Prepare your mindset. There is no wrong question! If you don't understand something, someone else might also not understand, and they will appreciate you asking about it.
Tell yourself: "I am going to ask a question today" before you join the seminar.
Now comes the timing to ask the question. I highly recommend being the first to raise your hand after the host says, "We are open to questions." Another good timing is right after someone just asked a related question you want to ask, you can follow up on the previous question and extend the topic.
I want to give some examples of the structure of the question. This structure might be obvious to some people. Still, it helps me focus on asking the question, and I don't need to think about whether I am doing it correctly. My structure choice is [complement, relate to the section, ask the question]. I would start the question by stating, "Great talk," to compliment the speaker. Next, mention which part of the talk is related to your question by saying, "You mention in your talk that …". Then ask your question.
Example 1: "Great talk! You mentioned that your work takes in the history information as prior and predicts the future movement. I wonder if this kind of temporal information can help find repeating patterns of the environment?"
Example 2: "Thank you for the amazing talk. In your demo, the robot cuts vegetables from top to bottom. However, when humans cut vegetables, there might be a rocking motion in the real world. I want to know if your framework can expand the states to such a motion."
Here are some questions I came up with. Once again, these questions might be biased toward the field of robotics or engineering because those are the seminars I have the most experience with.
Three categories of questions:
Details about the work.
Why did you choose this approach? If you know there is another way, ask their opinion on it.
What's the limitation of this work? How does it apply to this other situation? Will the result be the same or different?
Can you explain more about <something in the talk>? For example, a specific term, what does the x-y axis represent in the figure, why the result is like this.
What are the assumptions in the work? Think of a real-world situation; how is the problem simplified? Is there anything missing? How can we make it more aligned with the real-world situation?
How is the scalability of the framework? How can we scale up the framework to handle more complex situations? For example, more states.
Expand the domain.
How can this work be applied to <specific area you work on>?
Can your work overcome this other challenge that is not the main focus? For example, deformable objects, occlusion in the scene.
How can your framework combine with this other framework? For example, foundation model.
Ask about their personal opinion. These questions are useful when the topic is general, and you know the speaker is the authority in this area.
People say, "SLAM is a solved problem." What's your opinion on that?
There is <a trend> in this area of research. What's your thoughts on that? For example, data-driven methods instead of model-based approach.
What is your vision of robotics, especially the area you work on, in the next five years? What do you think robots can achieve at that time?
I am a strong advocate for having a growth mindset. I view asking questions as a skill I want to continue working on. For now, I can still hear my heart rate rise whenever I am preparing to ask the question. I believe I can come up with better questions more effortlessly in the future, but for now the above summary is my thoughts from my short practice period.
I also want to credit Silke Janz, who guided me and gave me tips on this environment setup and mindset.
Another good thing to do is to send a follow-up email to the speaker to make a connection. I will write another post about this topic. If you have a great example you want to share, please email me!