Throughout the COVID-19 pandemic, the term “social distancing” has been at the center of public conversation, but this phrase is a bit of a misnomer. Taken literally, the phrase seems to endorse social separation. But it’s not “social” distance we are trying to promote. It’s physical separation. In fact, preserving social ties—even at a distance—is essential for both mental and physical health. The results of epidemiological meta-analyses, for example, indicate that a lack of social support is on par with smoking cigarettes as a risk factor for morbidity and mortality and is even more harmful than other stressors, such as obesity and air pollution. Given this empirical fact, how might we best stay connected to others while maintaining physical distance? Would we better off e-mailing a friend? Making a phone call? Setting up a video chat? Modern technology has provided us with many tools at our disposal. Not all tools foster social connection equally, however. And often such seemingly small choices can make a big difference between cultivating stronger social connections and giving in to growing social distance.

In a paper soon to be published in the Journal of Experimental Psychology: General, Nick Epley and I tested whether the media through which people interact affects their sense of connection—and how expectations about certain technologies impact the communication media they choose to use. Note that these expectations can be misguided. Although voice-based interactions (such as phone calls) can produce stronger connections, text-based media (such as e-mails) are often preferred because of mistaken beliefs about how social interactions will unfold. Any interaction we have can come with advantages and disadvantages, and decisions about how to connect tend to be based on expectations of these potential costs and benefits. When people overestimate the cost, or underestimate the benefit, of voice-based communication, it can create a misplaced bias for text-based media.

In one experiment, for instance, we asked participants to reconnect with someone that they hadn’t interacted with recently, either through e-mail or over the phone. Participants first made predictions about what it would be like to get in touch if they reached out in these two ways. More specifically, these participants predicted how connected and how awkward they would feel in each situation. In this experiment, they did generally intuit that they’d feel more connected when interacting via the phone than over e-mail. But they also predicted that talking on the phone could be more uncomfortable than dashing off an e-mail. Participants additionally indicated which option they preferred. Although these participants believed that talking encouraged stronger bonds, most of them said they’d rather send an e-mail than call the person up. Fears about awkwardness, it seems, push individuals toward text-based methods for communicating.

Our results demonstrated that, contrary to participants’ expectations, worries about awkwardness are largely unwarranted. In the next part of the experiment, we had participants actually reconnect using one randomly determined mode of communication—either e-mail or phone—and then followed up with them after they had done so. As expected, we found that people do form meaningfully stronger bonds when interacting over the phone than over e-mail. Importantly, though, there was no difference in the amount of discomfort when reconnecting on the phone. The human voice appears to provide benefits for connection without the expected costs.

In another experiment, we had individuals connect with each other by asking and answering a series of relatively personal questions (for example, “Is there something you’ve dreamed of doing for a long time? Why haven’t you done it?”). These conversations took place by texting in real time during a live chat, using only audio, or engaging in a video chat. Once again, participants first made predictions about how they believed they would feel and then actually had a discussion with someone else. We again measured awkwardness and connection, in this case using statements such as how much they would get to know their conversation partner, how much they would like that person, and how strong of a bond they would feel. Here, participants didn’t expect that the media through which they communicated would matter. But when they actually interacted, people again felt significantly more connected—and notably, no more awkward—when they communicated by talking rather than typing. Interestingly, visual cues didn’t add more to what voice-based media already provided. Media containing only audio, as in a phone call, created as strong a sense of connection as audiovisual media, and both produced higher-quality social interactions than text-based media.

Our work suggests that these miscalibrated expectations can affect how people choose to connect with others and therefore how well they do so. Misunderstanding the costs and benefits of different interactions can lead individuals to choose inferior methods for connecting, leading them to text, e-mail, or send a message on Slack instead of picking up the phone, which results in a more positive interaction. In light of this evidence, it is critical that we not only focus on the content we are trying to convey but also the context in which it is conveyed. The next time you think about how best to connect, consider calling or setting up a video chat. You’re likely to feel better as a result. E-mail or text messaging can sometimes be useful for sending attachments or scheduling a time to talk, to be sure. Feelings of social connection, however, are optimally facilitated by one’s voice rather than their keyboard.