A Q&A with an Eller researcher digging into the science of 'sorry'

Image

If your job entails working with other humans, opportunities will occasionally come up when an apology is in order. And maybe, some of those times, you were the one receiving the apology and it left a bit to be desired.

Delivering effective apologies can be an important part of repairing, maintaining and strengthening relationships with colleagues.

The language we use is a key component of constructing a successful apology, says Sarah Doyle, associate professor in the Department of Management and Organizations in the Eller College of Management. She was part of a research team that measured apology effectiveness by studying celebrity apologies on X, formerly known as Twitter, and by collecting data from people who participated in scenarios in which they received an apology from an accountant or nurse.

For the social media study, researchers examined 87 apology tweets from celebrities including rapper and singer Lizzo, actor Tyler Posey and television personality Kendra Wilkinson. For the "everyday apology" studies, individuals rated apologies delivered to them by someone playing the part of an accountant or nurse who had made a mistake.

In both cases, the same trend emerged: When the person giving the apology "violated" gender stereotypes, the apology was seen as more effective.

In this Q&A, Doyle details what her research found about the language we use when saying sorry and how to craft an effective apology at work.

Why were you interested in doing this research?

When it comes to apologies, most people would probably agree that quality matters more than quantity. Given this, it was interesting to us that the bulk of attention surrounding apologies from men and women has been on the relative frequency with which they apologize, with women being commonly advised to apologize less and men to apologize more. We felt that these recommendations were oversimplified, at least in terms of helping people to effectively navigate difficult situations and effectively repair broken trust.

In many situations, a lack of an apology can be really harmful, and therefore for women, simply apologizing less is unlikely to be a successful strategy. For men, suggesting that they apologize more, without any direction about how to construct that apology, could lead them to deliver apologies that come across as forced or insincere. Given all of this, we were motivated to do this research out of a desire to shift the focus away from how often men and women apologize and move towards building a greater understanding about how to deliver a better, or more effective, apology.

What did you learn from studying celebrity apologies on social media?

Even though we were focusing primarily on workplace apologies, examining celebrity apologies served as a useful starting point to our research as it provided us with a lot of data, which was helpful in terms of allowing us the opportunity to detect any initial trends. The ability to examine people's reactions to celebrity apologies on social media was really instrumental, as the large amount of data gave us greater confidence that any potential trends observed were less likely to be spurious findings. It also provides some initial insight into the generalizability of our findings, suggesting that the reactions we observed in our studies may not be constrained to the workplace but could extend to people's relationships more broadly.

In this initial study, we found that, for women, apologies that used language that was more task-oriented or conveyed stereotypically masculine qualities such as assertiveness and independence were more positively received by the general public. Since celebrity apologies are often directed, at least in part, toward fans, we measured that in terms of the number of likes the apology post received and the degree of positive or negative sentiment reflected in the comments made by readers of the post. For example, positive sentiment might include expressions of gratitude for the apology or mentioning how genuine or sincere it sounded. Whereas this initial study supported the benefits that women receive from contradicting gender stereotypes, subsequent studies revealed the same benefits for men who used warmer and more relationship-oriented communal language, which is typically associated with more stereotypically feminine qualities of interpersonal sensitivity or empathy.

What defines "masculine" and "feminine" language?

We relied on decades of previous research for this. Historically, research has shown that men are expected to be more self-focused, signaling stereotypically masculine qualities such as assertiveness or independence. Within apologies, this means focusing more on the task at hand and demonstrating a commitment to problem-solving, perhaps by providing a reason for the offense and extending a solution for how to fix the problem. For example, "The mistake that I made was a product of my lack of understanding of the software I was using."

On the other hand, women are often expected to use communal language, which tends to be more relationship-oriented, expressive and sympathetic, signaling stereotypically feminine qualities such as interpersonal sensitivity or warmth. In the context of an apology, the language used may be perceived as more emotional or more focused on the damaged relationship. This might be saying something like, "I know what I did harmed you and I care about how you are feeling."

Were you surprised by any of the results of your research?

The predominant perspective within much of the research in this area is that there tends to be a negative backlash when people violate gender stereotypes at work, particularly for women. If you think about the context of an apology, you are apologizing because you have violated trust. Of all situations, this seems like one where it would be reasonable to assume that it is better to avoid further violating expectations. When we found that using unexpected language and countering stereotypes in your apology actually helped both male and female apologizers, it was pretty surprising at first.

When we started to dig into the data, it began to make more sense. Violating expectations can show that you've taken the time to really craft an apology that is going to matter to the person and that you are committed to effectively dealing with the trust breach.

Based on the results of this study, when someone needs to deliver an apology to a colleague, what should they be considering?

First, we hope our results don't lead people to overthink things to the point where they delay giving an apology. If you wait too long to apologize, it's not going to be as effective in repairing trust. That said, it can be useful, instead of having a knee-jerk reaction, to take a pause and think about how you can communicate with this person and show them that you are able to address the breach of trust.

For women, this could be taking a step back and thinking about communicating that they have an adequate understanding of what went wrong and how to fix the issue. For men, it could be making sure to show that you care about the harm you caused and repairing the relationship.

It's important to note that we're not saying that women and men don't already have these understandings. The research just shows the effectiveness of communicating those understandings.

What further research would you like to do in this area?

It would be interesting to look at the gender of the person receiving the apology. This is just speculation, but I would think that apologizing to someone of the opposite gender would raise the importance of violating gender stereotypes. In our experiments, we found that the effect for women was stronger in a stereotypically more masculine occupational context and the effect for men was stronger in a stereotypically more feminine occupational context, suggesting that it's possible that the gender of the person receiving the apology may play a role.

Another interesting topic would be the mode through which you're apologizing. We focused on written apologies, but there may be different nuances in verbal apologies or differences in live apologies versus a voicemail. It would be very useful to understand how all those different methods of communication are received.

Read more about Doyle's research on the University's news website.