The Washington Post - USA (2022-02-13)

SUNDAY, FEBRUARY 13 , 2022. THE WASHINGTON POST EZ BD B3

are a sufficient signal of perceived closeness.

(Presumably there’s a point at w hich a r esponse

seems too quick — rudely abrupt and possibly

off the point. But we did not study that phe-

nomenon.)

As mentioned, we did find some people

whom you might call super-connecters — those

who tended to respond more rapidly to others,

no matter their partner, and who therefore

instilled feelings of connection. Could it follow

that people without this natural ability could

simply decide to speed up their response time

and have the same effect?

Not really. Because humans can’t just edit

down gaps in their conversations (unless you’re

a researcher working with an audio file). The

only way to respond quickly is to understand

where the other person is coming from and to

anticipate where they are going. Sometimes

that happens effortlessly — we f all immediately

into step, minds dancing together in a kind of

flow state. Other times it requires slowly mov-

these contributions matter equally? The an-

swer, we found, is no: The degree to which

people felt connected was much more depend-

ent on how q uickly t he other person responded

to them. How quickly they themselves re-

sponded made much less difference.

Our studies also found that outside observ-

ers, too, perceive shorter gaps in conversation

as an indicator of social connection. To explore

this question, we took snippets of six different

conversations between strangers and manipu-

lated the size of the gaps with audio software.

We assigned participants on a crowdsourcing

platform to listen to one of three versions of

each conversation: one edited so that the gaps

were doubled, o ne in which gaps w ere one-fifth

the original length and the original exchange.

Across all the conversations, the shorter the

gaps, the more the outside observers thought

the two partners seemed connected. Since the

only things that varied in these conversations

were the gaps, we know that short gaps alone

A

s schools made the switch to remote teaching in 2020, Andrew Bacher-

Hicks, an assistant professor of education at Boston University, kept

hearing an alarming prediction: Cyberbullying would spike. With students

living more of their lives online, wouldn’t this pernicious form of harassment —

which some researchers have found is even more closely correlated with suicidal

ideation than in-person bullying — take off? (According to one 2019 study from

the Centers for Disease Control and Prevention, some 20 percent of American

high school students reported being bullied in person in the previous year; 16

percent reported being bullied online.) Bacher-Hicks and four colleagues at B.U.

— associate professors Joshua G oodman, Jennifer G. Green and Melissa Holt —

tackled the question using an unusual approach: T hey first established, by

examining past data, that Google searches for such terms as “bullying” and

“cyberbullying” closely track real-world trends, as measured in surveys. They then

looked at what happened to search trends during the pandemic. The results,

which appeared in a National Bureau of Economic Research working paper in

December, weren’t what many people expected. This interview has been edited

for length and clarity.

Q: Online is a space where people bully one

another, so some people predicted things

would go off the rails when schools went re-

mote.

A: Exactly. Those predictions were based on

research that was done before the pandemic.

It shows when schooling is constant — when

there’s not a major disruption to schooling —

there’s a correlation between students who

spend more time online and the amount of cy-

berbullying that they experience. That e vi-

dence did suggest that there was some cause

for concern when nearly all K-12 students were

spending a lot more time online.

Q: You end up establishing that, at least in the

pre-pandemic world, Google Trends data — on

searches related to bullying — tracks pretty

well with real-world, state-level survey an-

swers regarding bullying behavior. Just to

quickly summarize what you found: Google

searches for topics related to bullying drop off

markedly during the summer. They are espe-

cially high at the start of a school year, then

fade a bit. Most remarkably, you’ve got biannu-

al surveys of students by the Centers for Dis-

ease Control and Prevention — state-by-state

data on self-reported bullying. The Google

Trends data tracks variation in those survey

answers — state by state and over time —

which is pretty remarkable.

A: Yeah. We did think, when we first saw it,

that the tracking of the school calendar was

pretty amazing. There’s a very clear spike [in

searches] in the beginning of the school year, a

dip over periods when school was not in ses-

sion, to some degree — over the winter recess,

for example. Then a large drop-off over the

summer. And this pattern happens year after

year after year. So there’s a clear connection

between Internet searches for bullying and the

school calendar. And we did connect these bi-

ennial surveys from the CDC and found a

strong correlation between areas where stu-

dents self-reported more bullying and search-

es for bullying that happened in that area. And

the same thing for cyberbullying. So we were

able to validate the link between Internet

searches and self-reported survey information

— for both those topics — in the time period

before covid.

Q: Some schools went all virtual. Some schools

were hybrid. Some schools stayed open. You’ve

got data on where those schools are. When

schools went all virtual, how much did search-

es for school bullying and cyberbullying drop?

A: T he first analysis that we conducted looked

at the period directly following March 2020,

when schools across the entire country

abruptly switched from in-person instruction

to remote instruction. And during that time,

we found that searches related to school bully-

ing and cyberbullying both dropped by 30 to

35 percent.

Q: You also looked at the next school year,

where approaches to schooling were a little

more mixed, and had similar findings, right?

A: We found similar findings overall. But we

had a nice opportunity in the fall of 2020 be-

cause there was a lot of variation across differ-

ent types of schools. Some were quick to open

back up for in-person instruction, where oth-

ers remained remote. And what we saw is that

bullying dropped no matter what: In both

forms of instruction, school bullying and cy-

berbullying remained lower than in prior

years. However, the drop in both forms of bul-

lying was much larger in the areas that re-

mained remote.

PANDEMIC LAB BY CHRISTOPHER SHEA

The social science research emerging from the outbreak

When schools went virtual,

online bullying declined

JAYME GERSHEN/BLOOMBERG NEWS

As students shifted to remote learning during the pandemic, many observers predicted that the increase in time

spent online would lead to more online bullying. But new research shows just the opposite.

S

ometimes you meet someone and feel

instantly in sync. Conversation flows ef-

fortlessly, as if you can finish each other’s

sentences. It is hard to predict when it will

happen, and, in fact, such connection is so rare

and mysterious that at times it can seem

illusory — perhaps even more so in these

masked and socially distanced times.

In a study we recently published, however,

we showed that t here is a n objective measure of

when two people are connecting, a sort of

“tell”: The gaps between their conversational

turns shrink.

Though most of us rarely think about it,

two-way conversations represent an impres-

sive feat of coordination. People are able to

respond to each other remarkably quickly, in

about 250 milliseconds, or a quarter of a

second, on average. Each person absorbs what

the o ther says a nd formulates a r esponse in the

blink of an eye. Shrinking that gap further

requires even more attentiveness to what the

other person is saying. These speeds are too

short to be under conscious control, meaning

they are what b iologists call a n “honest s ignal.”

Our central finding w as t hat people reported

greater enjoyment, and a stronger sense of

connection, when the gaps in their conversa-

tion were shorter. To put it another way, close

conversations naturally feature shorter gaps.

But there were other intriguing findings, too.

Some people, for example, seemed to have a

particular gift for swiftly absorbing other peo-

ple’s points and responding in kind. They

responded more quickly than their peers,

across numerous partners — and they left a

feeling of warmth in their wake. Unfortunately,

there’s unlikely to be a shortcut for learning

how to emulate these conversational rapid

responders.

How did our study work? First, we asked

pairs of participants — d rawn from a 6 6-person

sample — to have 10-minute conversations

about whatever they wanted. Most had never

met each other. We recorded the conversations

and measured the intervals between each

speech turn. In a variation on this experiment,

we brought some of the same people back but

paired them with friends of theirs.

Afterward, in both cases, the participants

entered separate rooms and w atched the inter-

actions on video while moving a slider to

indicate how connected they had felt, moment

by moment. Even within a single conversation,

the instances when people felt especially con-

nected were characterized by shorter gaps than

at other times. Few people will be surprised to

hear that conversations have this kind of ebb

and flow.

In any conversation, two people are contrib-

uting to the average gap between turns. Do

When do we ‘click’ with someone in a conversation? This test tells us.

When we bond

with someone

we’re talking

with, the gaps

between our

turns speaking

shrink, say

researchers

Thalia

Wheatley and

Emma M.

Templeton

OLI SCARFF/AGENCE FRANCE-PRESSE/GETTY IMAGES

Researchers found

that the shorter the

pauses in a

conversation, the

more connected the

participants felt,

and the more they

enjoyed their talk.

ing toward that point through active listening

and getting to know each other (gradually

learning new dance steps). You can’t simply

will conversational connection into existence.

When we talk about i ntimacy, we tend t o use

distance words: We say we are “close,” “tight,”

“joined at the hip”; we say that someone is part

of our “inner circle.” This set of studies shows

that how we talk to each other also telescopes

interpersonal space. Responding quickly to

our p artners makes them feel closer. B y shrink-

ing the gaps between our turns, we also shrink

the psychological distance between us.

Twitter: @ThaliaWheatley

@emma__templeton

Thalia Wheatley is the Lincoln Filene professor

in human relations at Dartmouth College and

external professor at the Santa Fe Institute.

Emma M. Templeton is a graduate student in

psychological and brain sciences at Dartmouth

College.

Q: I f I understand it right, in your interpreta-

tion online bullying is largely an outgrowth of

in-school bullying.

A: Absolutely. I think that bullying that occurs

online may in many cases just be an extension

of bullying that started in person. We know

from prior research that many of the same in-

dividuals — that is, both the victims and the

aggressors — are involved in in-person and in

cyberbullying. So there are clear links between

the two. And I think an important contribu-

tion of our paper is to show that when you dis-

rupt one form of bullying, there is a clear re-

duction in both forms.

Q: Your paper suggests there might be lessons

for this and for the post-pandemic world.

What might those be? Because kids eventually

will be all back in these chaotic physical envi-

ronments again.

A: T he lessons might come out of this question:

Why do we think that even when schools re-

opened in fall 2020 during the pandemic, bul-

lying was lower in those schools than we would

have predicted? One reason is that schools put

additional structures in place to prevent the

spread of covid-19. And many of those struc-

tures likely helped to reduce bullying when stu-

dents were back in person. We know from prior

studies that a lot of bullying occurs during un-

structured time — that is, time spent passing

other students in the hallway, time at lunch,

etc. During the pandemic, there has been a lot

less flexibility in offering that type of unstruc-

tured time. And there is a lot more supervision

during the school day. I don’t think that we

should necessarily maintain all of these new

structures moving forward, but I think it does

suggest there’s something that we can learn

about how providing additional structured

time might reduce bullying.

Q: All of this really complicates the whole

mental health picture for students. Because

what we hear is that students are isolated. Stu-

dents are depressed. All of that’s true. On the

other hand, you can’t help but think that the

20 percent of students who are bullied might

have found some relief during the pandemic.

A: I think that’s fair. I will say, you know, near-

ly all of the educational research that I see to

date has focused on the negative consequenc-

es that you just mentioned. This study sug-

gests that at least some aspects of the educa-

tional experience have improved for some set

of students. This does not mean that we

should maintain this crazy system that we

have right now, but it does suggest that there

might be a small subset of students who really

suffered from bullying in the past, for whom

reducing in-person time could be beneficial.

Q: What else should we take away from this

study?

A: If you look over time in that biennial survey

by the CDC — at the national numbers show-

ing how prevalent bullying is — it’s been fairly

static over time. And so to see a large shift

[down] in bullying, I think, is a good sign. It

suggests that it’s not some static concept. It’s

not something that we have to accept. It shows

that changes can occur. That suggests that

there is room and scope for really making a

dent in what I think is a serious challenge that

clearly affects a lot of students.

Twitter: @cshea4

Christopher Shea is an assistant editor for

Outlook and PostEverything. Before joining The

Washington Post, he ran the Perspectives section

at Vox.com.