One reason I love to do what I do is because a). I am curious, which I’m guessing you are as well; but also b). I’m fairly shy — or at least I used to be. Not sure I ever really outgrew it. Shy and curious is a tough combination: there are answers you want to know but you’re not always comfortable asking the questions. These days, the internet is a big help — you can learn a lot from the comfort of your keyboard. But there are still occasions where you really need to ask another human being a question. Sometimes a sensitive question. That’s one reason I became a writer: it gives you permission to ask.
I once interviewed Ted Kaczynski, the Unabomber, at his supermax prison in Colorado. One question I asked him was whether he considered himself clinically insane. I’m not saying this was a good question; but it certainly wasn’t the kind of question I’d feel comfortable asking outside of an interview context. Kaczynski didn’t seem at all bothered by the question. He was happy to explain why he wasn’t insane. But I will admit, in my normal, everyday life, it’s still hard to ask sensitive questions. And apparently it’s hard for you too. We did a callout to Freakonomics Radio listeners:
“ROSIE”: Hi there, I’m Rosie, which is a fake name. This is also a fake accent.
GREGORY: Hi, my name is Gregory, and I work for a small construction firm .
SHELLEY: My husband is an atheist. And I’ve always wanted to ask his parents, who are deeply religious, if they think it would have been better to have aborted him, thereby sending him to heaven, rather than have him live the life that he’s currently living, in which their religious beliefs dictate that he’s likely going to hell if he dies.
“ROSIE”: My sensitive question that I always want to know — I work in a large bank. I always want to know what everyone earns. I think that this information is really important so that there’s not unfair pay. But if I try and ask someone, it creates an awkward moment.
GREGORY: Our boss, the guy who runs the construction company, is getting up there in age. And none of us know what would happen if he were to die or to get Covid. Literally, my job hangs in the balance. But I have no idea how to ask. Yeah, that’s my sensitive question.
“ROSIE”: So I would like that sensitive question to be answered for everyone.
Wouldn’t it be nice to know more about sensitive questions? For instance: exactly how strong is our avoidance, and is it generally a good thing or a bad thing? When you do ask a sensitive question, how uncomfortable does it make the other person feel? And what are we out on by swallowing so many of our sensitive questions? The good news is that academia — being the wonderfully weird place it is — has people who have been studying these very questions.
Einav HART: Hi. I’m Einav Hart.
Hart is a cognitive scientist who teaches in the School of Business at George Mason University.
HART: And what I care about is how people navigate difficult conversations and how we handle conflict.
Stephen DUBNER: In what context do you mean that? In a personal context, a business context, a political context?
HART: A bit all of the above. But what I care most about is situations where both parties have an influence on each other’s outcomes. So, you can think of negotiations, you can think of even everyday conversations, and conversations in business contexts as well.
Hart grew up in Israel; her Ph.D. is from Hebrew University in Jerusalem.
HART: I tend to be a bit more blunt than potentially advised in polite American society.
DUBNER: Can you think of a question that is not considered sensitive in the U.S., but is in Israel?
HART: Potentially politics? Although nowadays, nothing is off the table anywhere.
Today on Freakonomics Radio: the science of sensitive questions. We’ll talk about the questions we usually ask each other:
Eric VanEPPS: What are your job responsibilities? What do you think about the weather?
The questions we want to ask each other:
HART: Have you ever committed a crime?
VanEPPS: Have you ever had an affair?
Maurice SCHWEITZER: What are your views on abortion?
And: whether we should all be trying to move from the first kind of question to the second.
* * *
DUBNER: So what’s the difference between a sensitive question and an offensive question?
HART: That’s a good question. I think “offensive question” bakes in the consequence. Whereas a sensitive question can be something to do with a topic that is potentially combustible or that we don’t talk about a lot, but it’s not necessarily offensive, per se. So if you ask someone how much they’re paying in rent, that’s a potentially sensitive topic.
DUBNER: What if I said, “God, why do you pay so much in rent?” Is that the offensive version of the sensitive question?
HART: Potentially. I don’t think that’s the most offensive you could go, but—.
DUBNER: All right, how about like, “You sure you make enough money to pay that rent?” Is that offensive? Do you have a horn you can blow when I get to offensive? How about this? “How did an idiot like you make enough money to pay that rent?” Am I there yet?
HART: That could work, right? The more curse words or derogatory remarks you throw in, that could work.
DUBNER: I see. Okay, so “What a f***** idiot you are to pay so much f***** money for a s***hole of an apartment.” And that wasn’t even a question. But it was offensive, yes?
HART: Yes, yes, you could definitely offend people.
Okay, Einav Hart is a good sport, but our goal today isn’t to ask questions that are outright offensive. It’s to figure out what turns a regular question into a sensitive one, and whether it’s still worth asking. We should probably start by thinking about why we ask questions. There are many reasons! You want information. You’re curious; maybe you’re bored. Maybe you’re really interested in the person you’re talking to — or at least you want to signal that you’re interested. Or maybe your question is meant to signal that you yourself are an interesting person. Asking questions about questions: this is something Einav Hart and her academic collaborators enjoy.
HART: Yeah, apparently.
Her collaborators in this case are:
VanEPPS: I’m Eric VanEpps. I’m an assistant professor of marketing at the University of Utah.
SCHWEITZER: I’m Maurice Schweitzer, professor at the Wharton School.
Unlike Hart, VanEpps and Schweitzer both grew up in the U.S. — Schweitzer in California, VanEpps mostly in Nebraska.
VanEPPS: I think Nebraska fits into this category of “Midwest nice,” where we’re often pretty uncomfortable sharing a lot of information or asking a lot of questions.
HART: When you move across even places in the U.S., let alone overseas, there are different norms of conversation and what tends to be sensitive and potentially offensive.
SCHWEITZER: We had long conversations about ways to study different norms where, in many parts of the world, asking questions like: “When are you planning to have kids?” “Why aren’t you married yet?” “How much money do you earn?” — questions that for Americans might seem very sensitive, in other places would not be.
HART: There’s very little research that actually looks at what questions people are, in fact, willing to ask.
As a researcher, this represents a gorgeous stretch of virgin territory. With important consequences.
HART: How do we navigate these tradeoffs between the information we want and potentially not offending or annoying someone else?
So Hart, Schweitzer, and VanEpps set about to conduct a study that could examine the types of questions that people feel comfortable and uncomfortable asking, what it would take to get people to ask more uncomfortable questions — and what would happen if they did. The first thing they had to do was determine what, exactly, is a sensitive question.
VanEPPS: Yeah, we use the definition that sensitive questions are questions about topics that are uncomfortable to discuss, are inappropriate for the social context, or are about information that respondents would rather keep private.
Okay, so that’s a definition. But which questions would a given person consider “inappropriate,” for instance? To find out, the researchers ran surveys on the Mechanical Turk crowdsourcing platform, better known as MTurk. It is hardly the most representative sample in the world — but this is how a lot of research is done these days. The researchers asked participants to rank a variety of questions, from least to most sensitive. Let’s see how you’d rank them. Here’s one batch of questions:
HART: What do you think about the weather?
SCHWEITZER: Have you ever mowed a lawn?
HART: How do you get to work?
SCHWEITZER: Are you a morning person?
And, here’s another batch.
HART: Have you ever had financial problems?
VanEPPS: Have you ever had an affair?
SCHWEITZER: How much is your salary?
VanEPPS: Do you or have you ever gone to therapy?
HART: I imagine most people listening have this cringey feeling about either asking these or being asked these. And that’s what we’re trying to get at.
Once the researchers had a pool of questions ranked by sensitivity, they were ready to test them, with what turned out to be a series of five experiments. The first one, again using participants from MTurk, sorted 360 people into pairs that would have online conversations, by text. Half of them would ask questions, the other half would answer. The askers were then split into three groups, and had to pick their questions from the pool of questions that had already been ranked by sensitivity.
SCHWEITZER: Essentially we forced them to pick questions that were either mostly sensitive, mostly not sensitive, or something of a mix in the middle.
The research subjects were then asked to predict how their conversation partners felt about being asked these questions.
SCHWEITZER: And then we also had their counterpart, after they had the conversation, rate how uncomfortable the conversation was.
This was something the researchers would do throughout the whole series of experiments.
HART: We wanted to see what do people think would happen if they ask sensitive questions? And by what would happen, I mean: what would the other person feel and what would the other person think about them?
In this experiment, people were not particularly uncomfortable asking sensitive questions, and the respondents’ comfort levels were fairly similar across the three conditions. But remember, these are anonymous, online text conversations between MTurkers who didn’t know each other and would likely never communicate again. The second experiment, still using MTurk, added a pair of wrinkles. The first was the use of financial incentives — although they were so tiny, less than a dollar, that you have to wonder how powerful they’d be. The second wrinkle was the introduction of what are called “impression-management concerns.”
HART: So, a lot of times when we approach interactions — be they with strangers, with friends, with bosses — we want to leave a good impression. You want the other side to think nice, positive things about you and you don’t want them to think that you’re a terrible person or that you’re really annoying.
In this experiment, there were three groups of question-askers. The first group would be paid more if their conversation partner thought better of them by the end of the experiment. The second group would be paid more if they made their partner uncomfortable. And the third group was the control condition, which was not assigned a conversational goal.
HART: What we see is that people, when they were incentivized to create discomfort, chose almost all sensitive questions. You’re asking about abortion and about sexual preferences, about bank accounts, everything.
And how about the group that was paid to leave a good impression?
HART: They stuck to neutral, safe topics.
And what about the control group?
HART: People in the control condition were very close to the good-impression condition, which suggests that people, even if not given any specific goal, are thinking very much about the impression and not so much about the interesting or sensitive conversation.
This result may not surprise you. Most of us want other people to like us. And it would seem obvious that we’re more likable if we don’t ask sensitive questions. On the other hand: these were anonymous, virtual conversations; you might think it’d simply be more interesting, more fun, to ask the so-called sensitive questions. But this experiment suggests that most people don’t think that way, or perhaps that we’re so conditioned to not ask sensitive questions that even when allowed, we don’t.
HART: There are a lot of things that could affect whether people choose sensitive questions. People can have a lot of goals when they enter into a conversation. So potentially had we incentivized people to have a more interesting conversation or to get as much information as they can, that would also potentially lead people to prefer sensitive questions over neutral ones. Or if you want to amuse yourself or amuse the other person, that could also affect what questions you ask.
For the next experiment, the researchers recruited a couple hundred students from the University of Pennsylvania. All the participants had to do was select the questions they were going to ask a conversation partner — again, from the pool of questions ranked by sensitivity. Five of these questions were ranked as sensitive. The wrinkle here was that some participants were told their conversations would happen in person, and some online. Once again, there were financial incentives: 30 or 40 cents for each sensitive question they chose, versus just 5 or 10 cents for a non-sensitive question. Again, the money is so tiny, it’s hard to imagine — at least for me — that it would matter much.
SCHWEITZER: The fact that they could have earned, you know, a few extra dollars — yeah, I think that’s a modest incentive in this case.
Still, you do get paid three or four times as much for picking a sensitive question. So what happened?
VanEPPS: Importantly, in both conditions, people are leaving money on the table, so to speak. Most people are not choosing to ask all five sensitive questions and maximize their payments. They’re choosing some combination of sensitive and non-sensitive questions, ostensibly because the payment isn’t enough to overcome some of the problems that they might have with asking particular sensitive questions.
SCHWEITZER: It could be we don’t want to make other people uncomfortable. It could be we just feel embarrassed to do it. But I think what’s so interesting is that even with a very explicit economic incentive there’s still an aversion to this.
Again, we should acknowledge just how small the stakes were here. You can imagine a lot of situations in real life where you might overcome this aversion to asking sensitive questions. For instance, is the founder of the startup where I might take a job as big a jerk as I’ve heard? Is the person I’m about to marry 100 percent over her ex? Is my teenager really going to a friend’s house tonight to study? As we like to say around here: incentives matter, and scale matters. There are a lot of things you would not do for $10 that you might for $100. And the list gets much longer if I offer you a million dollars. Anyway: for the next experiment in the study, Hart, Schweitzer, and VanEpps wondered if they could alleviate the concern that people have about sensitive questions by deflecting responsibility away from the question-asker.
VanEPPS: Maybe we can get you to be more likely to ask sensitive questions if the responsibility for question selection is instead given to the computer.
This experiment was done in-person, face-to-face, with more than 250 students at, again, the University of Pennsylvania. Just as the people who use MTurk are not a particularly representative population, nor are Ivy League students. Still, this was what the researchers had access to. Once again, the participants were split into groups: question-askers and respondents. Then the respondents were further split: half of them were told that a computer had picked the questions they were going to be asked. The askers knew the respondents were given these instructions. But in reality, all the askers picked their own questions. The researchers suspected they’d be more willing to ask sensitive questions if the respondents thought the computer had done the picking.
VanEPPS: We actually didn’t find that. We found across both conditions, people are choosing a similar number of sensitive questions to ask, and the number is quite low. And so this reluctance to ask sensitive questions is robust even to a manipulation where we kind of absolve you of the responsibility for asking the questions, suggesting that you’re actually quite worried about how comfortable your partner is rather than completely selfishly worried about whether it’s going to reflect badly on you.
This raises perhaps the most interesting question of all: how comfortable is the partner? How does the person being asked a sensitive question feel about being asked that question? That’s what the researchers were really trying to get at by now, as the study reached its crescendo. Because in our daily lives, most of us aren’t just asking questions of strangers; it’s people we know. So in the fifth and final experiment:
SCHWEITZER: We contrasted talking with friends versus strangers.
VanEPPS: We recruited student participants from a university lab pool, and we asked them to come with a friend to the lab.
HART: But they were either chatting with their friends or chatting with a stranger. Someone else’s friend.
So there are two conditions: friend-to-friend, stranger-to-stranger. In both cases, the participants all had the same list of sensitive questions to ask.
SCHWEITZER: Here we found that when you’re having a conversation with friends, you’re a little bit more receptive, a little more open to asking them sensitive questions.
VanEPPS: Because I know my friend better, I have a better prediction of how they will respond.
How much more willing are people to ask their friend a sensitive question? Here’s how the researchers answer that question.
VanEPPS: It is a somewhat smaller overestimation for friends.
Meaning, an overestimation of whether a sensitive question will bother the friend.
VanEPPS: But even still, it is an overestimation of how uncomfortable they will be, and how bad an impression I am making.
So here’s what this series of experiments shows. Most people think that asking a sensitive question will upset the person being asked. And that it will make that person have a worse impression of the asker. But neither of those, it turns out, are substantially true.
VanEPPS: People are less likely to ask sensitive questions because they think sensitive questions would offend their conversation partner.
SCHWEITZER: And in fact, they were far more comfortable than the askers both anticipated and then thought afterwards.
HART: People are generally less offended and care less about being asked a sensitive question than we think they would be.
SCHWEITZER: As an asker, I think I’ve created a really unfavorable impression. But then the reality is, the respondents, they’re actually happy to have these conversations.
In other words: according to this study by Hart, Schweitzer, and VanEpps, our reluctance to ask sensitive questions is just wrong! Or, at least wildly misplaced.
SCHWEITZER: We find that people often don’t ask the important questions they should be asking. I think part of it is figuring out what kinds of questions we should be asking, to make sure that we ask them and make sure that we get the answers. Questions like “how much money do you earn?” Or “how does your commission work?” Questions that, if I care about the impression that I’m creating on you, I might avoid asking but could be really valuable.
* * *
There are questions many of us would like to ask but don’t. Here are a few, sent in by our listeners.
JON: Hello, my name is Jon. I work for a very large, multinational company.
JAMES: My name is James, from Des Moines, Iowa.
D’Juan HOPEWELL: My name is D’Juan Hopewell and I live in Bronzeville, a very historic Black neighborhood on the South Side of Chicago. My question is, are white people deep, deep down actually okay with the world that we have?
JAMES: I want to ask my wife why she doesn’t remove her upper lip hair. She meticulously manages the shape of her eyebrows and is well-groomed otherwise.
JON: With the recent elections, it’s very clear that our nation is divided in half. So whether it’s someone on my own team or someone within a department next to us, we know that there must be someone that’s voted for someone that we have very strong opinions against. I would want to know if there’s a way to really ask a coworker who they voted for. But then, how do you not become, you know, somewhat biased towards them on their beliefs? And how do you professionally navigate that?
Now, that last question illustrated the potential downside of asking sensitive questions. Because you may not want to turn your workplace into a non-stop political debate. Some sensitive questions are plainly thornier than others. In a recent paper called “The (Better Than Expected) Consequences of Asking Sensitive Questions,” the researchers Einav Hart, Maurice Schweitzer, and Eric VanEpps ran a series of studies that examined how people ask, and respond to, sensitive questions. Here, again, is Hart; she teaches at George Mason University:
HART: We show that people are pretty much averse to asking sensitive questions, because they fear leaving a bad impression on the counterpart, which is actually a biased view of the world, in that respondents in our studies really don’t care as much.
DUBNER: You also write that our aversion to asking sensitive questions is costly.
HART: If you had not asked me about how much I pay in rent or potentially my salary, you wouldn’t have learned that you should ask for more yourself or pay less in rent. And so there is that informational gain as well as relationship-building potential.
DUBNER: So you mentioned that we overestimate the impact of a potentially sensitive question. How much is that related to the fact that humans are pretty bad at predicting the future generally, and that maybe this doesn’t have so much to do with sensitive questions per se?
HART: There’s definitely part of that. But this is a very specific direction. You could imagine, “Oh, I’m really bad at predicting what will happen. So let me just ask these questions and hope for the best.” And that’s not what we see. We see this misprediction even after the conversation.
DUBNER: I’m curious if that relates to what psychologists call the spotlight effect: this idea that we tend to overestimate how much other people pay attention to us and how much we as individuals actually matter in the scheme of things.
HART: I think it could very well be related to that. We didn’t test that directly. But what’s interesting is that we also find that these misperceptions occur both when we think about having the conversation beforehand and even after the conversation itself.
SCHWEITZER: I think we go through life really focused on ourselves.
And that is Maurice Schweitzer, who teaches at the University of Pennsylvania.
SCHWEITZER: We’re focused on our own actions. We think our actions have an exaggerated influence on other people. We’re trying to forecast and take perspective, and both of those things — thinking into the future and putting ourself in somebody else’s shoes — are extremely hard. Those are two things that humans have some capacity for, but we constantly fall short. And I think the one consolation is that it’s just human.
And humans, being the complicated and opaque creatures we tend to be, are full of cognitive biases and errors. So how big of a deal is this new research? Does this mean we should all start asking every question that pops into our head? Maybe not yet.
HART: This is the tip of the iceberg. This is an initial foray into this area of what questions are people actually willing to ask and why.
And, while it’s a very nice and deeply interesting study, it’s also got a long list of limitations. As we noted earlier, the research subjects — college students and MTurk users — they represent narrow slices of the population. Since many of the experiments were done online, with anonymity, they don’t represent the real-life environments we most care about. Although, as Hart points out:
HART: A lot of our conversations, especially now during Covid, are happening over chats and with strangers. And so I don’t think this is an experience that’s that far removed from a lot of our interactions.
Nor did the researchers focus on individual differences that might be useful to know, like ethnicity or gender differences.
HART: What I can say is we actually did not find a difference between men and women I think that’s also because we’re all strangers online and so, you don’t know also if your counterpart is a man or woman, and they don’t know that about you.
DUBNER: Do you know anything, or could you suspect anything, about age? For instance, I think of teenagers as being really willing to ask questions that some people interpret as sensitive. I also think that some older people are like, “Screw it, it’s too late. I’m just going to ask it.” Do you see any age relationships?
HART: We don’t. But also our population was very limited in age. To your point, I think, yes, kids are known to ask a lot of questions off the top of their head. Kids are still developing their prefrontal cortex, that guides controlled behaviors and adherence to other norms in society. In that respect also, elderly people, their prefrontal cortex usually tends to degrade as well. And so to your point, yeah, their concern over what other people think of them —probably lower. At least if my grandma is a good case.
Nor did this study get into the area that first got Hart, Schweitzer, and VanEpps thinking about the subject: how people in different cultures ask and answer different kinds of questions. But here’s what Hart thinks:
HART: While the set of sensitive topics or questions can vary quite a lot across cultures, I think the overall mechanism is the same, where we just don’t want to ask things that we think our counterpart will perceive as sensitive.
So Hart and her co-authors, having hit the tip of this interesting iceberg, now have a long list of follow-up research to do:
HART: How do we behave across different groups and cultures and with different norms? Would I be more willing to ask questions if I think you can also ask me them in return? And who are we talking with? Is it someone that we know? Don’t know? How much power do they have over us? An initial project I’m looking into involves sensitive questions for specific groups. So you can think of say, marginalized groups. And what questions are we willing to ask or when or why? What questions do we ask of disabled people? What questions do we ask of people of different racial groups? Do we just want to know where the group lines are? And how does our curiosity and value of information play out in these settings?
There are also questions to be asked about answering sensitive questions. While Hart’s research shows that people generally overestimate the discomfort level, plainly there are limits. Especially when it concerns a private or intimate question. Let’s say you’re suffering from deep depression, and you call a doctor’s office. How candid will you be? What if the intake screener is actually a chatbot? They are increasingly being used in medical settings just like this, and elsewhere.
Ryan Schuetzler, an assistant professor of information systems at Brigham Young University, has been studying chatbots, and how people respond to them. In one study, he and his co-authors found that the more human a chatbot seems, the less likely someone is to answer a sensitive question truthfully. This would seem to be an important finding to consider as chatbot use grows — not just in the screening of patients, but even to treat depression. But I think the biggest takeaway from the research by Hart, Schweitzer, and VanEpps is that we should all push ourselves a bit more for conversations that may inevitably include some sensitive questions. Because there’s a significant opportunity cost of not asking such questions. How significant?
HART: given the range and variety of sensitive questions and topics, I don’t think you can put a number to it. I wish I could. But there is data to suggest that if you don’t know what your salary expectations are, and also don’t negotiate them, you can miss out on about 10 percent maybe?
DUBNER: When I first read this research, what intrigued me so much was thinking about it, not even in the labor markets, but just how much any person can miss out on opportunities in life. Like, “Oh, you look like a really interesting person. Can we be friends?” I assume that’s a question that most people don’t ask. Maybe it’s not a sensitive question per se, but I’m just thinking about the broadest implications of our oversensitivity. Do you have any insights into that?
HART: So I think you can think of these potential gains or, in this case, misses. If we continue just talking about the weather, we don’t actually learn that much about people and who we like. And who your good friends are, in times of trouble. Imagine a good friend of yours had a personal tragedy. You might be hesitant to ask them about it because it’s a sensitive topic. And they might be unlikely to talk about it because, again, they also potentially don’t want to burden you. But actually both sides might want to talk about it, right?
DUBNER: I have a very dear friend who had a tragedy involving the loss of a child very young. And he made the note that no one would ask him about it, because, he assumed, they thought he didn’t want to talk about it, when in fact that was all he wanted to talk about.
HART: Avoiding the question also means that we don’t learn that this is actually fine or even a good idea to ask.
SCHWEITZER: We’re missing often some really important information.
That is Schweitzer again.
SCHWEITZER: So we can think about asking peers for their salary. When I go to negotiate my salary — suppose I just got promoted. I’m going to be at the managerial level. I need to walk into that and know what kinds of things should I be asking for. What sort of salary is the right range? I might, without that conversation, totally miss opportunities to say, “Hey, I should be asking for my own parking spot. I should be asking for my own assistant. I should be asking for coaching.” There could be things that I fail to think about because I haven’t had those conversations.
DUBNER: Let’s say I read your research and believe it, that sensitive questions are a good thing to engage in for the asker and the answerer. And then I say to you, “Look, Maurice, we just met on the line here a little while ago. I want to demonstrate the value of your argument, or I want you to back up the value of your argument. Ask me a sensitive question that you think will provide good evidence that this argument is worth listening to.”
SCHWEITZER: So I might ask you, you know, can you describe the pleasure and pitfalls of writing your first book?
DUBNER: What makes you categorize that as a sensitive question?
SCHWEITZER: Well, I think maybe if I focused on the pitfalls, or if I were to say, “Where did you really struggle writing your first book?” — I think that might cast you in a negative light or cause you to reveal some weakness or shortcoming. I think actually, it could be an interesting launching point for developing a deep understanding, a deeper conversation, around something that could be quite useful.
DUBNER: And how would you measure the efficacy of asking the sensitive question versus revealing a sensitive piece of information about yourself?
SCHWEITZER: When we ask questions, we’re almost invariably inviting a reciprocal reply. So I ask you how your weekend was.
DUBNER: Mine was good, end of story.
SCHWEITZER: That’s it, right. I’m also coming back to that idea that the questions that we ask convey information. I’m demonstrating what I’m interested in. I’m giving you a sense of how assertive I’m going to be, what sort of knowledge I have.
DUBNER: Yeah, I guess I’m thinking about seeding the conversation with more information than just what’s contained within the question. Here’s a for-instance. Let’s say a child of mine goes off to college, they’re a freshman, and I want to know how they’re doing — like really how they’re doing — emotionally, academically, socially. But if I just say, “Hey, how are you doing? Is everything okay?” — which I feel are the pro forma questions that will get pro forma answers — what if I say, “You know, child of mine, I remember my first couple of months of freshman year and I was miserable. I didn’t know how to register for classes. I didn’t know how to make friends. I didn’t know how to do laundry. How’s it going with you?” Is that useful or is it more useful to just straight-up ask a question that, as you said, contains some contextual information?
SCHWEITZER: Well, so, what I like about your approach is that it demonstrates some vulnerability. And it’s likely to make it easier for somebody else to demonstrate some vulnerability. What I was thinking of is, having some college-age kids myself, you can ask a question like, “When’s the last time you went out on a date?” That’s a more direct sensitive question, right? That’s going to give me some insight, like, “Hey, what’s happening?” But I liked your approach, where I’m making myself vulnerable. I’m sharing information. And I’m effectively changing the dynamic of the conversation that lowers the bar for people to be more forthcoming.
DUBNER: What can you tell us about how people feel about being asked questions generally, whether sensitive or non-sensitive questions? I’m curious, because I ask a lot of questions. And I always assume that it’s taken generally as a sign of curiosity in the person, and that most people really like to talk about themselves and their work and their lives. But on the other hand, some people seem to just turn right off.
SCHWEITZER: Yeah, there is some recent research that in general, people love being asked questions, particularly if it’s about my very favorite subject in the whole world, which is me. Asking people questions, it demonstrates interest, it demonstrates concern, it can build rapport and relationships and people generally like being asked questions.
VanEPPS: Just calling somebody up out of the blue and asking them a sensitive question might be really aggressive or really uncomfortable.
And that, again, is Eric VanEpps, from the University of Utah.
VanEPPS: I have a 2-year-old son, and in the months leading up to his birth, I decided I wanted to do interviews with my parents to hear their oral history, so that we would have a record of that. In that context, I found myself quite willing and comfortable asking pretty sensitive questions. I was having conversations with my dad that I have never had before, because we had created this framework of, “We’re going to talk about our lives and we’re going to get things down on paper that we otherwise wouldn’t talk about.” And so to the extent that we can create opportunities to be an interviewer, to assign ourselves the role of being an asker of questions, I think that generally makes people more willing to ask a variety of questions, including sensitive questions.
SCHWEITZER: I think the best advice would be this:
And Maurice Schweitzer again.
SCHWEITZER: If you’re on the fence about asking a sensitive question, go for it. If it seems like something that’s just crazy, don’t do it. But I think these findings should give people a small push in the direction of going for it and asking that question.
DUBNER: So we know from the literature in psychology and economics that what’s called loss aversion is real and substantial, that people suffer more from a loss than they take pleasure in an equal-sized gain. So from a loss-aversion perspective, wouldn’t it make sense to avoid sensitive questions? Because if I ask the question, there’s a chance I might blow up a personal relationship or ruin a work relationship, whereas the gains to be made are, in some cases at least not life-changing.
SCHWEITZER: I think that account helps explain why people might not do it. And I’d push that further to say, first of all, loss aversion is a bias. So the fact that we feel more painful when we lose $20 than when we gain $20, that might cause us to be far more risk-averse than we should be. And second, what we’re finding is that that perceived harm is likely to be far, far smaller than we expect. That is, I think I’m gonna make you extremely uncomfortable. But in fact, it’s the launching pad for an interesting conversation. And we end up in a much better place as a result.
Did we end up in a better place — I mean here, today, in this episode? Your feedback is always welcome; we’re at firstname.lastname@example.org. If you did like this episode, there’s another one from our archive you might enjoy. It’s Episode 192, and it’s called “That’s a Great Question!” It explores why people say that so often, and why the question they were asked is often not great at all. If you don’t like the sound of that episode, there are more than 400 others to choose from. You can now get the complete archive of Freakonomics Radio, free, on any podcast app. And check out the other shows in the Freakonomics Radio Network — No Stupid Questions and People I (Mostly) Admire.
* * *
Freakonomics Radio is produced by Stitcher and Renbud Radio. This episode was produced by Mary Diduch. Our staff also includes Alison Craiglow, Greg Rippin, Mark McClusky, Matt Hickey, Zack Lapinski, Daphne Chen, and Emma Tyrrell; we had help this week from Jasmin Klinger. Our theme song is “Mr. Fortune,” by the Hitchhikers; the rest of the music was composed by Luis Guerra. You can subscribe to Freakonomics Radio on Apple Podcasts, Stitcher, or wherever you get your podcasts.
- “The (Better than Expected) Consequences of Asking Sensitive Questions,” by Einav Hart, Eric M. VanEpps, and Maurice E. Schweitzer (Organizational Behavior and Human Decision Processes, 2021).
- “Chatbots and Conversational Agents in Mental Health: A Review of the Psychiatric Landscape,” by Aditya Nrusimha Vaidyam, Hannah Wisniewski, John David Halamka, Matcheri S. Kashavan, and John Blake Torous (Canadian Journal of Psychiatry, 2019).
- “Online Panels in Social Science Research: Expanding Sampling Methods Beyond Mechanical Turk,” by Jesse Chandler, Cheskie Rosenzweig, Aaron J. Moss, Jonathan Robinson, and Leib Litman (Behavior Research Methods, 2019).
- “The Influence of Conversational Agent Embodiment and Conversational Relevance on Socially Desirable Responding,” by Ryan M. Schuetzler, Justin Scott Giboney, G. Mark Grimes, and Jay F. Nunamaker Jr. (Decision Support Systems, 2018).
- “The Influence of Conversational Agents on Socially Desirable Responding,” by Ryan Schuetzler, Justin Scott Giboney, Mark Grimes, and Jay F. Nunamaker (Hawaii International Conference on System Sciences, 2018).
- “Social Incentives for Gender Differences in the Propensity to Initiate Negotiations: Sometimes It Does Hurt to Ask,” by Hannah Riley Bowles, Linda Babcock, and Lei Lai (Organizational Behavior and Human Decision Processes, 2007).
- “Ageing and the Brain,” by R. Peters (Postgraduate Medical Journal, 2006).
- “That’s a Great Question! (Ep. 192)” by Freakonomics Radio (2015).