"Don't tell me you're sorry cause you're not": Some Surprising Science of Apologies

A few days ago, an email popped into my inbox from a current psychology PhD student. She had read an academic journal article that I published in 2013 about all the existing ways to either induce or measure people's abstract thinking during psychological experiments, and she felt compelled to thank me for it. I chuckled reading the email. The "Tour Guide for Abstraction" was the most painful project I worked on in grad school and I wished it good riddance the moment it was officially published. I briefly considered centering this week's blog on the paper, not because I wanted to write more about abstract thinking, but because it tested every ounce of my will for nearly 3 years. Not only did it set an unofficial record for the most times the current editor of the journal had sent back an official "revise and resubmit" decision (9 times), it was also immensely boring to write (and rewrite). The topic of abstract thinking, though it was my PhD advisor's passion, was never mine. 

As I started to outline a blog in my head, it dawned on me that instead of writing about the most boring work I did in grad school, it would be much more fun to talk about the most interesting work I did. So, as the title indicates, I'm going to share a little bit of what we -- myself and four coauthors -- learned about apologies.


Let me set the stage a little bit: Around the same time my advisor suggested I write the "Tour Guide for Abstraction," I approached my close friend and classmate, Dr. Gili Freedman, about collaborating on a project. I desperately wanted to work on something other than abstract thinking and her research area fascinated me -- understanding the causes and consequences of social rejection from the perspective of the rejector. (Most of the existing research has been about what it feels like to be rejected, so she was exploring a new frontier). We soon hatched a plan to understand what people found appropriate or inappropriate to say when rejecting others, with no particular hypotheses in mind. Our multi-year collaboration -- including begging hundreds of people in line at the South by Southwest film festival to participate in one of our studies, and purchasing large quantities of hot sauce -- is one I was both sentimental and relieved about coming to an end. Though the studies were in Dr. Freedman's domain, and she was the lead author on the paper, when I think back on grad school, our collaboration is what I am most proud of. So without further ado, here's what we learned.

What people say when they have to reject someone

At the outset, apologies were not the central focus of our work. But our first study made it clear that we had landed on something counterintuitive and interesting. We crafted four different scenarios in which a person might reject another and asked study participants (some of whom were those moviegoers mentioned earlier) to imagine themselves in that situation:

  • a coworker wants to join your standing Friday lunch group, but your group does not want to include other people
  • someone you met at a party wants to be your friend, but you are not interested
  • a person you have exchanged messages with on an online dating site asks to meet in person, but you are not interested
  • your lease is ending and your roommate asks to live together again next year, but you have already decided not to live with them again

For each of these scenarios, we asked the participant to write out exactly what they would tell the person as a "good way of saying no.” Dr. Freedman and I read all 1300+ responses and looked for common themes. We found a few key patterns: people tended to say something nice about the person they were rejecting, and often included some form of alternative to the ask (e.g., “maybe you and I can go to lunch together on another day”). We also found that 40% of people apologized in some way for the rejection. A typical rejection response including all of those features might have read: "I'm sorry, I think you're a really nice person, but I don't have time for new friends right now. Hope to see you at a party again sometime." Things got even more interesting in the next phase.

The surprising effect of apologies

After Dr. Freedman and I documented these patterns of language, we next wanted to know what language made rejections feel better and worse. Specifically, we wanted to know what language, if any, impacted how hurt the rejectee felt and how socially adept the rejector seemed. Because the written rejections were responses to hypothetical scenarios, we couldn't measure real hurt feelings of the rejectee or perceptions of the rejector. We did the next best thing. Undergraduate research assistants -- unaware of any hypothesis -- read each rejection and rated both how hurt they would feel if receiving it, and how adept the rejector seemed. When we combined the ratings from all the research assistants, the results surprised us. A few elements of rejections, such as saying something nice about the rejectee, tended to soften the blow of a rejection. However, apologizing had the opposite effect! People not only felt more hurt about rejections containing apologies, they also saw the rejector as less socially adept. So even though a large percentage of people apologized in their rejections, it seemed to backfire.

Because our findings were limited to hypothetical scenarios, we knew we needed to follow up with a real-life rejection scenario. That's where the hot sauce came in...

The field of psychology is known for all kinds of creative ways to simulate real-life behavior in experimental labs. One pretty famous method, usually aimed to measure aggressive behavior, uses hot sauce. Two participants come into the study, though one participant is actually a confederate -- a research assistant pretending to be a participant. Both participants are told the study is about taste testing foods, and fill out forms about their preferences for spicy and sweet. The confederate admits to hating spicy foods. At some point in the experiment, the confederate does something that may or may not provoke the real participant. Then during the taste testing portion of the study, the experimenter tells the real participant that they get to choose how much hot sauce the confederate, who ostensibly hates spicy foods, has to drink. Of course, the confederate doesn't actually drink the hot sauce, and in the end, the true hypotheses of the study are revealed to the participant. The amount of hot sauce the participant poured is then measured, with the idea that the more hot sauce poured, the more aggressively the participant responded to the confederate's behavior.

We adapted this famous method to study how apologizing for a rejection influences how rejectees respond. In two separate studies, an experimenter told the participant that the confederate got to choose if the pair worked on a task together or separately. The confederate would respond with either "No, I don't want to work with you" or "No, I'm sorry, I don't want to work with you." In both studies, the results were clear -- participants who heard the words "I'm sorry" doled out significantly more hot sauce to the confederate than those who didn't. 

Why do apologies backfire?

Now that we had pretty clear evidence that apologies were making rejections worse, we wanted to know why. We had several different hypotheses, including the "Don't tell me you're sorry cause you're not" line from Rihanna's song Take a Bow. (Side note: for years I wanted the title of our paper to be that line, but it was it too cutesy for an academic journal and was not the only reason apologies backfire). Our final study involved online participants imagining themselves as a person named Taylor who was then rejected by Jamie. They watched a pre-recorded video where they envisioned they were having a conversation with Jamie (an actor on the video) about rooming together for another year. Jamie would either say (into the camera as if speaking to the participant) "I’ve actually already found a different place" or "I'm sorry, I’ve actually already found a different place." Not only did people who heard apologies find the rejections less sincere, they also felt obligated to express forgiveness by saying "it's okay" even if they didn't actually forgive Jamie.

So putting all of our studies together, we found that although people likely feel that "I'm sorry" is an obligatory part of rejecting another person, they are actually making the rejector answer with obligatory forgiveness, making the rejectee feel more hurt and think less of the rejector. 

Some caveats (because one academic paper only tells one piece of a larger story)

As I wrap up, I will caution that we did not exhaust all the possibilities of why saying sorry hurts, and it's possible that there are several reasons. We also only studied rejection situations that involved somewhat casual relationships. Something like a romantic breakup would warrant more than the few sentences our participants wrote in response to our situations, and it is possible that an apology could actually go a long way.

Despite our study only being preliminary evidence of the negative effect that apologies have on social rejections, I have become very mindful of my own behavior in last few years when I've had to say no to someone, and have helped several friends and family craft "good" rejections. And, as a rejectee (cause we all get rejected sometimes), I have learned to truly (rather an obligatorily) forgive people who apologize when they have to reject me. :)

If you're curious to hear more about this topic while having a few laughs, check out Dr. Freedman's interview with comedian Paula Poundstone on NPR's Live from the Poundstone Institute. Or if you'd like to read the actual paper, you can find the full text here

Much love, Erin