00:00:00
the first line result is that men and
00:00:02
women performed equally well on average
00:00:04
in fact if you look across all of our
00:00:06
subjects women performed maybe a half a
00:00:09
question better but the men rated
00:00:12
themselves much more favorably on all of
00:00:14
these scales and so on the zero to 100
00:00:17
scale men gave themselves ratings over
00:00:19
25 percent higher than the ratings that
00:00:22
women gave themselves welcome to the
00:00:25
ripple effect the podcast that takes you
00:00:27
on a journey through the minds of work
00:00:29
and faculty I'm your host Dan Loney and
00:00:32
in each episode we'll be diving deep
00:00:33
into the inspiration behind the
00:00:35
groundbreaking research that Wharton
00:00:37
professors have conducted and exploring
00:00:40
how their findings resonate with the
00:00:42
world today we'll be covering a diverse
00:00:44
range of topics bringing you the latest
00:00:46
insights and knowledge that you can
00:00:48
apply to your life into work so get
00:00:51
ready to dive into new ideas with the
00:00:54
ripple effect
00:00:55
so Judd the annual performance review is
00:00:59
a thing thing that a lot of people dread
00:01:02
what was it that led you to study this
00:01:05
topic in the first place
00:01:07
yeah we were very interested in a
00:01:09
question that economists in particular
00:01:11
had not really studied that much which
00:01:14
is how do people subjectively describe
00:01:16
their performance it's one thing to
00:01:19
answer a question like how how many uh
00:01:22
units did you sell this year or how many
00:01:24
new clients did you sign on but a lot of
00:01:27
communication about performance and
00:01:30
ability so even before you join a job
00:01:33
maybe you're asked about how good you
00:01:35
are at something a lot of the
00:01:36
communication around that is subjective
00:01:39
you there's no right answer you have to
00:01:42
describe something that's kind of hard
00:01:44
to pin down and economists had not
00:01:47
really looked at that but we think that
00:01:49
it's something that's quite important in
00:01:51
the way people perceive you and so we
00:01:53
wanted to really look at that style of
00:01:55
question being asked something
00:01:57
subjective about ability or performance
00:02:00
and apparently your research showed to a
00:02:03
degree that women systematically rate
00:02:06
themselves lower than men on work
00:02:09
performance even when their work may be
00:02:12
viewed as being objectively better can
00:02:16
you account for the difference there
00:02:19
so this was one of the things that
00:02:21
motivated us two subjective questions
00:02:24
and the way people answered them because
00:02:26
we had this hypothesis that not only is
00:02:28
this important and understudied but it's
00:02:30
something that we might see a gender gap
00:02:33
where men conditional on the same
00:02:36
performance men would describe that
00:02:38
performance more favorably so my
00:02:41
co-author on this work is Christine
00:02:43
Exley at Harvard Business School and you
00:02:45
know she and I wrote this paper and if
00:02:47
you asked us about the same paper I
00:02:49
might say oh it's phenomenal and she
00:02:51
might say oh you know it's pretty good
00:02:52
and that difference of the way that we
00:02:56
talk about the work even though it's the
00:02:58
same underlying project that we're both
00:03:00
a part of that was the kind of dynamic
00:03:02
that we thought might be interesting to
00:03:04
explore
00:03:06
so when you talk about the office are
00:03:09
there subtle ways that that this might
00:03:11
show up more so than not
00:03:14
so we were interested not just in the
00:03:17
performance reviews that you know you
00:03:19
might get asked once a year where the
00:03:21
employer will sit you down and kind of
00:03:23
work through a set of questions we were
00:03:25
also interested in the kinds of more
00:03:29
commonplace and often sometimes subtle
00:03:33
interactions that you have with your
00:03:34
colleagues and with your supervisors
00:03:37
where you're just talking about work and
00:03:39
and people are getting an impression of
00:03:41
how well you do your job and whether you
00:03:43
might be able to take on more difficult
00:03:45
challenges and if men systematically
00:03:49
talk about their prior performance and
00:03:51
their underlying ability in more
00:03:53
positive terms that might change the way
00:03:56
that people perceive them relative to
00:03:58
their equally capable female female
00:04:00
colleagues and so that was uh the kind
00:04:04
of what pushed us a little bit more
00:04:05
towards these more subjective questions
00:04:08
and the ways in which people typically
00:04:10
communicate that is you know with
00:04:13
adjectives or uh we we still we're
00:04:16
economists who wanted to study this we
00:04:18
still had quantitative questions so we
00:04:20
asked things like how much would you
00:04:22
agree with the statement I performed
00:04:24
well on a zero to 100 scale so we can
00:04:27
still get a quantitative measure but
00:04:28
it's clearly a subjective of question
00:04:30
that doesn't have a correct answer
00:04:33
is the potential impact for women having
00:04:35
this difference at times noticeable
00:04:38
yeah so it's a little tricky because
00:04:41
what we don't want to do is say that
00:04:45
women should necessarily self-promote
00:04:47
more that that they should necessarily
00:04:49
talk more favorably about their own
00:04:50
performance they are in a situation
00:04:53
where it's possible that there will be
00:04:56
backlash and perhaps differential
00:04:58
backlash uh for speaking too positively
00:05:02
about their own ability and performance
00:05:04
it's possible that the gender difference
00:05:08
that we observe is in fact a response to
00:05:11
kind of social training that people go
00:05:14
through where when they do talk
00:05:15
favorably about how well they did you
00:05:17
know that they might be met with kind of
00:05:19
harsher uh responses and so they've kind
00:05:21
of learned not to not to push it on
00:05:24
those adjectives so the study that we
00:05:27
did which I'm excited to tell you more
00:05:29
about took away the possibility for
00:05:32
backlash so the there wasn't anyone who
00:05:35
was going to respond to what the S the
00:05:37
subjects in our study said uh in uh
00:05:40
necessarily a negative way in fact they
00:05:42
didn't even know if there was anyone
00:05:44
responding they didn't know the gender
00:05:45
of the person
00:05:46
um but it doesn't mean that those same
00:05:49
forces weren't influencing the way that
00:05:52
men and women talked about how well they
00:05:54
did so the advice is not to the women or
00:05:57
the men about how they talk about
00:05:59
themselves any takeaways from the
00:06:01
research have to be on how we elicit
00:06:04
information about people's uh
00:06:07
performance and ability and and
00:06:08
potentially not relying so much on the
00:06:10
way people talk about themselves
00:06:13
all right so take us through the
00:06:15
research a little bit and what you were
00:06:17
able to decipher
00:06:19
yeah so this uh this was a project I was
00:06:22
really excited to be a part of and it uh
00:06:24
ended up with a lot of parts but the
00:06:26
first thing we we did was we had our
00:06:30
study subjects uh so there were about 4
00:06:32
000 of them who were recruited from an
00:06:34
online labor market platform uh and then
00:06:36
I'll tell you about some youth that we
00:06:38
uh did a similar version of the study
00:06:40
with uh I'll tell you about that later
00:06:41
but the uh 4 000 folks the first thing
00:06:45
they did was they took a math and
00:06:46
science test so 20 Questions taken from
00:06:50
the armed services vocational aptitude
00:06:52
battery test which is a test that is
00:06:55
used by folks like me as measure of
00:06:58
cognitive ability but the questions we
00:07:01
picked were were math and science
00:07:02
questions
00:07:04
the subjects took that test and then we
00:07:07
asked them you know how well did you
00:07:08
think you did on the kind of normal
00:07:11
out of 20 questions you know tell me the
00:07:13
number that you think you got correct
00:07:14
and that's something that economists
00:07:16
have looked at for a while uh sort of
00:07:18
just we call it we think of it as
00:07:20
confidence of how how many questions you
00:07:22
you got correct then we asked our
00:07:24
subjective question so describing the
00:07:27
performance on the test with an
00:07:29
adjective that range from very poor to
00:07:32
exceptional and also these questions on
00:07:35
the zero to 100 scale of agreement with
00:07:38
statements like I performed well on the
00:07:40
test I took
00:07:42
what we found was that this the first
00:07:45
line result is that men and women
00:07:47
performed equally well on average in
00:07:49
fact if you look across all of our
00:07:50
subjects women performed maybe a half a
00:07:53
question better but the men rated
00:07:56
themselves much more favorably on all of
00:07:59
these scales and so on the zero to 100
00:08:01
scale men gave themselves ratings over
00:08:04
25 percent higher than the ratings that
00:08:07
women gave themselves
00:08:08
so the first hypothesis that we had was
00:08:10
maybe this is just uh confidence just
00:08:13
remanifesting so we looked at that
00:08:16
question of how uh how many of the 20
00:08:18
questions that you got asked did you get
00:08:20
correct and sure enough even though men
00:08:22
and women on average each got about 10
00:08:24
questions correct
00:08:25
men said they got 11 correct and women
00:08:28
said they got eight correct so there was
00:08:30
clear clear evidence that men and women
00:08:33
had different perceptions of their
00:08:35
underlying ability and we wondered
00:08:39
whether that was what was causing the
00:08:41
gender gap in self-promotion so what we
00:08:44
did then was we told men and women after
00:08:48
they took the test we told them exactly
00:08:50
how many questions I got correct so now
00:08:52
we're comparing men and women say who
00:08:54
each answer 10 questions correctly and
00:08:56
have been told that they answer 10
00:08:58
questions correctly and we asked the
00:09:00
same questions about uh you know stating
00:09:04
agreement with I performed well on the
00:09:06
test and even
00:09:08
with the same performance and knowing
00:09:09
they have the same performance there was
00:09:11
still a very large gender gap so the
00:09:13
gender gap shrinks a little bit when you
00:09:15
give performance information but it but
00:09:17
just a little bit it's still quite big
00:09:19
and so that told us there's something
00:09:20
Beyond just the not being sure how many
00:09:23
questions I got correct there's
00:09:24
something fundamentally different about
00:09:26
the way men and women interpret the same
00:09:29
underlying score
00:09:30
but the challenge I think in terms of
00:09:32
giving that self-review that
00:09:35
self-evaluation that I think in many
00:09:37
cases is probably a very unique element
00:09:40
for people to have to do in general
00:09:42
whether they be men or women
00:09:45
so this was uh one of the kind of
00:09:48
hypotheses that we had and we wanted to
00:09:50
push the result that I just described a
00:09:53
little bit further so we did a few
00:09:55
different things one thing we did was we
00:09:57
told men and women the average
00:10:01
self-evaluations of other people who had
00:10:03
scored the same as them so now the men
00:10:05
and women both know they answer 10
00:10:08
questions correctly and are now both
00:10:11
told what the average self-evaluation is
00:10:13
of folks who get 10 questions correct
00:10:15
still we see the gender gap it's
00:10:18
unchanged from that so we had another
00:10:20
hypothesis which is maybe women just
00:10:23
have higher standards than men maybe you
00:10:25
know it it's just the case that if you
00:10:27
ask a woman how a certain score is that
00:10:30
they just kind of always think that it's
00:10:31
it's less good than than a man would so
00:10:34
to get at this we did two things uh both
00:10:37
I think are informative one thing we did
00:10:39
was we asked people not to describe
00:10:42
their own performance but we asked them
00:10:44
to describe somebody else's performance
00:10:46
so now you've taken the test you you
00:10:49
don't know that you've answered 10
00:10:50
questions correct but we're gonna look
00:10:52
at folks say who men and women who both
00:10:54
answer 10 questions correctly and we're
00:10:57
going to ask them to describe the
00:10:58
performance of somebody else who
00:11:00
answered 10 questions correctly so it's
00:11:02
their performance they don't know it's
00:11:03
their performance but now it's not about
00:11:06
themselves it's about somebody else
00:11:07
when we do that we see no gender gap so
00:11:10
when describing a third party men and
00:11:14
women have the same at use the same
00:11:16
adjectives for the same score it's only
00:11:18
when they're talking about
00:11:19
their own performance that men and women
00:11:21
differ
00:11:23
the other thing we did was and this is
00:11:25
based on Research from economics in
00:11:27
other fields suggesting that gender gaps
00:11:30
are related to the kind of stereotypes
00:11:33
surrounding the task so
00:11:36
the unfortunate stereotype around Math
00:11:38
and Science tasks is that they're kind
00:11:40
of more male we call a male type that
00:11:42
men are expected to do better than women
00:11:44
even though in our case of course women
00:11:46
do do better than men
00:11:48
but
00:11:49
we picked a task that is less male type
00:11:52
that's kind of more neutral and if
00:11:53
anything uh folks think that women might
00:11:55
outperform men and that was a verbal
00:11:57
task so same kind of setup but instead
00:12:00
of doing a test with 20 math and science
00:12:02
questions uh the subjects are doing a
00:12:05
test with verbal questions
00:12:08
and when we do that we again see no
00:12:11
gender gap in the way that men and women
00:12:12
talk about their performance when the
00:12:15
performance is equally good so the thing
00:12:19
that's happening here is that there's
00:12:20
something unique about this kind of math
00:12:22
and science domain where the stereotype
00:12:25
is that men do better where
00:12:27
in that domain only women are
00:12:30
saying that they do less well even
00:12:32
knowing that they have you know a
00:12:35
particular performance and that that
00:12:36
performance is the same as a man who
00:12:38
says they did great so going back to
00:12:40
what you said a couple of moments ago I
00:12:42
think it brings up the topic of
00:12:43
self-promotion and how we think about
00:12:45
how we are answering questions and how
00:12:48
we view our perception of how we're
00:12:50
doing in comparison to what we believe
00:12:53
our colleagues may be doing as well
00:12:56
and I'm wondering if there is a
00:12:58
potential concern of whether or not
00:13:00
there can be too much self-promotion for
00:13:02
either men or women
00:13:05
yeah so it's hard to get at this because
00:13:08
I mean there certainly could be and
00:13:10
that's uh earlier when I was talking
00:13:11
about backlash that
00:13:13
um you know maybe women don't say that
00:13:16
they're phenomenal in math and science
00:13:18
because they're worried uh
00:13:21
that doesn't fit stereotypes of women
00:13:23
and people will kind of react negatively
00:13:25
to them saying that they're uh that
00:13:28
they're very good at this and in fact
00:13:29
there is a lot of evidence of backlash
00:13:31
uh for actions that are not that don't
00:13:35
fit stereotypes that could you know
00:13:39
could be playing out here as well so
00:13:41
there's certainly that element I think
00:13:43
you know it's not
00:13:46
obvious without more research about
00:13:49
exactly where the line is for kind of
00:13:50
what the what's the appropriate amount
00:13:52
of uh
00:13:54
positivity to have about you you know
00:13:56
how much can you uh maybe over inflate
00:13:59
your your own sense of self I will say
00:14:01
one other thing one other kind of
00:14:03
studies that we did uh to kind of start
00:14:06
to think about this we
00:14:08
in much of what I've described the
00:14:11
person who is answering the questions
00:14:13
about how well they did these subjective
00:14:15
questions we told them that their uh
00:14:19
performed that sorry there's their
00:14:21
answer to the self-promotion questions
00:14:23
these subjective questions would be
00:14:26
given to another study subject who would
00:14:30
use that answer and only that answer so
00:14:33
they wouldn't learn about how well the
00:14:34
person actually performed they would use
00:14:36
the self-promotion answer to decide
00:14:38
whether to hire the person and how much
00:14:40
to pay them
00:14:41
so that's where the self-promotion kind
00:14:43
of language comes from because this is
00:14:45
an answer to a question that could be
00:14:46
used to determine your your pay from the
00:14:49
study well and that was going to be part
00:14:52
of my question next because I think you
00:14:54
also have to look at this from the
00:14:56
employer's perspective and how they
00:14:58
understand these differences in the
00:15:01
process of hiring and promotion through
00:15:03
a company
00:15:04
yeah so that's uh one of the things that
00:15:07
I found interesting about the the
00:15:09
different versions that we did because
00:15:11
when the uh
00:15:14
studies subjects knew that the employer
00:15:16
would see the
00:15:18
self-evaluation that they gave about
00:15:21
their own performance uh they rated
00:15:23
themselves more favorably so they did
00:15:25
respond to the fact that there was an
00:15:27
incentive uh to to say that they did
00:15:29
well but when we took the employer
00:15:31
incentive away
00:15:33
we still saw a big gender gap so men and
00:15:36
women both of them were kind of
00:15:38
inflating how they described themselves
00:15:40
in the presence of an employer when the
00:15:42
employer was removed so when they're
00:15:44
just describing themselves to us the
00:15:46
experimenter
00:15:47
they still they both come down and level
00:15:50
but they're still a massive Gap so this
00:15:52
to me says for the employer perspective
00:15:54
it's not about exactly the way that you
00:15:57
incentivize the self-promotion questions
00:15:59
that's going to matter it's not like you
00:16:01
can say oh we're asking these questions
00:16:03
but you know we promise not to use them
00:16:05
to determine bonus or you know we're
00:16:07
only going to use them in certain ways
00:16:09
it's not gonna removing the incentives
00:16:12
to promote
00:16:13
is not necessarily going to change the
00:16:15
gender gap which was there even absent
00:16:18
the the employer at all it really is
00:16:21
something about the way that people
00:16:23
describe their
00:16:25
their ability and performance even when
00:16:27
you know it's not being used for
00:16:29
anything it's just yeah they they're
00:16:31
just saying it to us the researcher so
00:16:34
can you theorize what might be the way
00:16:37
for women to be able to close that
00:16:39
gender gap a little bit
00:16:41
so our takeaway is that well it's
00:16:45
twofold one we think it's not the
00:16:47
responsibility of the women it shouldn't
00:16:49
be the responsibility of the women to
00:16:51
change the way that they act for
00:16:53
multiple reasons one it's it's unfair to
00:16:55
put it on them two it's not clear that
00:16:58
we know how to help people change the
00:17:01
way they think about their own ability
00:17:03
and performance oh one result I didn't
00:17:05
tell you that speaks to that is the
00:17:07
study on youth so uh we did a study with
00:17:10
10 000 middle and high school students
00:17:13
where we had them do essentially the
00:17:15
same thing except we did it without
00:17:17
employers and we did it with uh shorter
00:17:21
math and science tests and we saw gender
00:17:23
gaps in the way that people talked about
00:17:26
their ability and performance in every
00:17:28
grade that we looked at from uh as young
00:17:32
as sixth grade all the way up to
00:17:34
um to to seniors and that was pretty uh
00:17:39
pretty clear evidence to us that this
00:17:41
was not something saying that you know
00:17:43
was was coming at a particular point in
00:17:45
time this was something that uh you know
00:17:48
seemed to be throughout at least Middle
00:17:50
School and High School uh this pattern
00:17:54
um so if there is intervention to be
00:17:55
done it might need to happen before
00:17:57
sixth grade
00:17:59
um
00:18:00
so that's the second thing is that it's
00:18:01
not clear how we would change the way
00:18:03
people perceive their own ability and
00:18:05
performance and then third is the thing
00:18:06
I've mentioned already which is the
00:18:08
potential for backlash so it's not clear
00:18:10
that we want to encourage women to
00:18:11
promote more because that could uh you
00:18:14
know that could it could be optimizing
00:18:16
given the incentives that they face
00:18:18
what that means then is that it pushes
00:18:21
the the the kind of fixing the problem
00:18:23
back to the employers who at interview
00:18:25
stages uh at application stages at the
00:18:29
performance review stages
00:18:32
you know we would argue should rely Less
00:18:34
on these subjective self-evaluations and
00:18:37
should rely more on objective measures
00:18:41
to the extent that they exist in
00:18:43
assessing ability and performance
00:18:46
so where do you think the next step in
00:18:48
this research is where do you want to
00:18:50
take this next
00:18:51
yeah so uh we are starting to think more
00:18:55
about how the social dynamics
00:18:58
of say the workplace or in educational
00:19:01
institutions how that interacts with the
00:19:04
gender gap so one thing that we started
00:19:06
to look at in our research was whether
00:19:09
people perceive the gender gap so
00:19:11
whether I anticipate that women are
00:19:14
going to talk less favorably about their
00:19:16
own ability and performance
00:19:18
and it does not look like the study
00:19:21
subjects who are participating so the
00:19:23
people who are giving us the answers
00:19:25
that display the Gap they don't appear
00:19:28
to anticipate that the Gap is there so
00:19:31
that doesn't answer a key question which
00:19:34
is do employers who are doing the hiring
00:19:37
and promotion and deciding on bonuses
00:19:40
whether they've kind of learned over
00:19:41
time that uh that women kind of talk
00:19:44
differently about their performance it's
00:19:47
possible that good employers have
00:19:49
recognized that if they ask this
00:19:51
question they get the the gender gap and
00:19:54
and maybe they even correct for it to a
00:19:56
certain extent but but that I think is
00:19:58
the next line of work to figure out
00:20:00
is that observation there and are the
00:20:03
corrections actually being made when
00:20:05
they need to be
00:20:06
Judd great to have you with us today
00:20:08
thanks very much for your time it's
00:20:10
great to be here thanks so much
00:20:12
thank you for listening to the ripple
00:20:14
effect we hope you found this episode
00:20:15
informative and engaging don't forget to
00:20:18
subscribe and leave us a review so that
00:20:20
we can continue to bring you the best
00:20:22
Insight from the Wharton School
