Popular in Economics

Central Monetary Authorities

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Central Monetary Authorities

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Economic History

Golden Fetters

UC Berkeley Barry Eichengreen

Economic Inequalities

Visions of Inequality

City University of New York Branko Milanovic

Economics

School Finance Equalization Increases Intergenerational Mobility

Yale University Barbara Biasi

Economics

Olivier Blanchard on U.S. Pandemic-Era Inflation

Peterson Institute for International Economics Olivier Blanchard

Strategic Conformity or Anticonformity to Avoid Punishment and Attract Reward

View more insights in Social Sciences - Economics - Behavioral economics

View more figures from University of Konstanz

View the page of Urs Fischbacher

Read the Study

Next in Behavioral economics

A Family of Norms: Related Beliefs about Female Labor Force Participation

Elif Incekara Hafalir | University of Technology Sydney

Strategy-Proof Multi-Issue Mediation: An Application to Online Dispute Resolution

Onur Kesten | University of Sydney

How do the expectations of rewards or punishments influence our decision to either conform or stand out?

Image courtesy of interviewee. November 22, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video discusses a research study centered around two key questions: the heterogeneity of conformity and the interaction of conformity with rewards and punishments. Conformity is defined as individuals aligning their behavior with others, while anticonformity describes behavior that diverges from both the group and personal preferences. The study finds that anticonformity is relatively rare.

The research examines environments dominated by either rewards or punishments. In reward settings, individuals may strive to stand out, while in punitive contexts, they may conform to avoid negative consequences. The study’s findings indicate that evaluators (those enforcing rewards or punishments) often prefer conformity in punishment contexts and may encourage anticonformity in reward environments, depending on how informative social feedback is.

Different experimental designs were employed to measure these behaviors, including knowledge questions and art taste assessments. Results showed that conformity was most pronounced in knowledge tasks and less so in artistic selections, with minimal anticonformity observed overall.

The researchers concluded that the context determines the necessity for conformity or anticonformity. For instance, in high-stakes environments (like piloting), conformity is crucial, whereas creative fields (like art) may benefit from anticonformity. The study emphasizes the relationship between environmental demands and individual behavior, suggesting a clear link between the nature of the task and the tendency toward conformity or creative expression.

Language:

Export as PDF

Export as PDF
Here are five key points from the video:

1. **Heterogeneity in Conformity**: The study focuses on two main questions regarding conformity, specifically exploring the concept of anticonformity, where individuals act against both the group's norms and their own preferences. This behavior is under-researched and was found to be rare in the findings of the study.

2. **Influence of Reward and Punishment**: The research examines how different environments, distinguished by dominance of reward versus punishment, affect conformity behavior. Rewarding environments, like in art and science, encourage individuals to stand out, while punishing environments, like those involving safety and performance (e.g., pilots), foster conformity to avoid negative outcomes.

3. **Social Feedback Effects**: The study indicates that social feedback influences decision-making by providing individuals with information about the evaluator’s preferences. The degree of social feedback available can lead individuals to conform or behave anticonformity, depending on the convideoual dynamics of potential reward or punishment.

4. **Experimental Design and Methodology**: The authors implemented a novel experimental design using a triplet choice method, allowing for the observation of conformity and anticonformity without and with social feedback, distinguishing these behaviors from mere independence. This design included multiple domains such as knowledge questions and artistic preferences.

5. **Findings and Trends**: The study reveals that in environments where punishment is prevalent, individuals tend to conform to the majority; conversely, in rewarding situations, individuals might lean towards anticonformity. The findings highlight that the amount of observed anticonformity was low, particularly in areas where participants felt less restricted and had the freedom for more creative expression, like color creation tasks.

Export as PDF
Here are some key quotes extracted from the video, focusing on the main ideas and findings:

1. **On Heterogeneity in Conformity and Anticonformity:**
- "We were specifically interested whether there are also people who specifically are anticon conform."
- "In our convideo, conformity means that people do something in the direction of others and deviating from their own taste."

2. **On the Nature of Environments with Reward and Punishment:**
- "There are environments in which reward is dominant, and there are environments in which punishment is dominant."
- "When there is this risk of being punished, people try to hide while when there is a reward possible, people try to stand out."

3. **On the Influence of Evaluators:**
- "This rule has the implication that people try to get the information from social feedback."
- "When they learn what others do, they get information about how the taste of the evaluator could look like."

4. **On Conformity vs. Anticonformity:**
- "We wanted the design in which we can distinguish conformity and anticon conformity from independence."
- "We define conformity as a shift of these individual decisions towards the decision of a majority."

5. **On Experimental Design:**
- "We ask people about their choice between A and B, and then we ask them about A and C with social feedback."
- "We have different domains in which we conduct these experiments."

6. **On Results Related to Conformity and Anticonformity:**
- "In the case of punishment, both go in the same direction towards conformity. In the case of reward, there is a conflict between the two."
- "There is a tendency towards conformity. There is also a difference between the different domains."

7. **On Personal Taste vs. Social Pressure:**
- "In the salience situation where salience matters in the reward treatment, the own taste was more important than salience."
- "We do not observe that much anticon conformity... so there is a tendency towards conformity."

8. **On Implications of the Findings:**
- "What are environments in which reward is used and what are environments in which punishment is used that's actually related to conformity and non-conformity."

These quotes collectively highlight the study's exploration of conformity and anticonformity, the role of reward and punishment in decision-making, and the design of experiments to understand these behaviors better.

X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:90000

1
00:00:00.065 --> 00:00:02.045
We were interested in two questions.

2
00:00:02.305 --> 00:00:06.605
One, refer refers to heterogeneity in conformity.

3
00:00:06.945 --> 00:00:09.685
And there we were specifically interested whether there are

4
00:00:09.795 --> 00:00:14.725
also people who specifically are anticon conform.

5
00:00:15.185 --> 00:00:20.005
And that means in our context, conformity means

6
00:00:20.235 --> 00:00:24.815
that people do something in the direction

7
00:00:25.075 --> 00:00:29.615
of others and deviating from their

8
00:00:29.875 --> 00:00:31.375
own taste, so to say.

9
00:00:31.435 --> 00:00:34.535
And anticon conformity means that they

10
00:00:35.165 --> 00:00:38.695
move away from the others, so do something different,

11
00:00:38.995 --> 00:00:41.055
but also against their own taste.

12
00:00:41.155 --> 00:00:45.135
So that's essentially the, the, these two behaviors.

13
00:00:45.155 --> 00:00:47.965
And anticon conformity

14
00:00:49.665 --> 00:00:51.045
is rarely studied.

15
00:00:51.265 --> 00:00:56.045
And, actually we find it's also rare in,

16
00:00:56.145 --> 00:00:58.765
in our study, and I will talk about that later.

17
00:00:58.945 --> 00:01:02.245
The second question, relates how this interacts

18
00:01:02.245 --> 00:01:03.445
with reward and punishment.

19
00:01:03.505 --> 00:01:07.965
And the idea here is that there are environments

20
00:01:08.505 --> 00:01:10.285
in which reward is dominant,

21
00:01:10.425 --> 00:01:13.925
and there are environments in which punishment is dominant.

22
00:01:14.065 --> 00:01:18.485
And that means, for instance, when you

23
00:01:19.095 --> 00:01:21.045
think about art

24
00:01:21.305 --> 00:01:25.605
or science, their reward is much more relevant

25
00:01:25.605 --> 00:01:29.845
because, it's though you are rewarded for good stuff.

26
00:01:30.305 --> 00:01:33.405
And if you have once in a while a bad paper,

27
00:01:33.405 --> 00:01:35.005
nobody punishes you for that.

28
00:01:35.705 --> 00:01:37.685
On the other hand, if you are a driver

29
00:01:37.865 --> 00:01:42.125
or a pilot, then nobody cares about your best performance.

30
00:01:42.585 --> 00:01:45.125
But people really care about the worst performance,

31
00:01:45.145 --> 00:01:48.365
and so you are more likely to be punished for,

32
00:01:48.905 --> 00:01:50.005
bad performance.

33
00:01:50.005 --> 00:01:52.485
And, and in that sense, punishment is more,

34
00:01:52.595 --> 00:01:54.285
much more relevant in this case.

35
00:01:54.665 --> 00:01:56.885
And generally that's also the case.

36
00:01:56.915 --> 00:01:58.365
Generally when there is a hiring

37
00:01:58.625 --> 00:02:01.365
that's more a rewarding environment.

38
00:02:01.505 --> 00:02:03.525
In the case of a threat of layoffs,

39
00:02:03.635 --> 00:02:05.725
this is more a punishment situation.

40
00:02:05.825 --> 00:02:10.765
And the idea is that when there is this,

41
00:02:11.875 --> 00:02:14.845
this risk of being punished, people try

42
00:02:14.845 --> 00:02:18.925
to hide while when there is a reward possible, people tried

43
00:02:19.385 --> 00:02:20.565
to stand out.

44
00:02:21.845 --> 00:02:26.705
And, this is was kind of our starting point

45
00:02:26.705 --> 00:02:28.385
with respect to reward and punishment.

46
00:02:28.885 --> 00:02:33.465
Uh, the idea was that this evaluators, so those

47
00:02:33.765 --> 00:02:37.785
who exert reward or punishment, we call them evaluators,

48
00:02:38.445 --> 00:02:40.425
and, they,

49
00:02:41.135 --> 00:02:43.585
they could target the special ones,

50
00:02:43.925 --> 00:02:45.585
but this is not the only possibility.

51
00:02:45.615 --> 00:02:47.745
It's also possible that they care more about

52
00:02:47.745 --> 00:02:51.785
that they do something similar as what they would do.

53
00:02:51.965 --> 00:02:53.705
We analyze this theoretically

54
00:02:53.805 --> 00:02:55.985
and interestingly the same occurs.

55
00:02:56.205 --> 00:02:58.985
So in the case of punishment,

56
00:02:59.135 --> 00:03:03.675
There is a shift towards,

57
00:03:04.185 --> 00:03:05.435
towards conformity.

58
00:03:05.535 --> 00:03:09.555
If there is a reward, there can be a shift even

59
00:03:09.655 --> 00:03:11.915
to anticon conformity and less conformity.

60
00:03:12.145 --> 00:03:16.715
This rule that evaluate this care about what,

61
00:03:16.765 --> 00:03:19.275
where the people do what they would do

62
00:03:19.655 --> 00:03:22.635
or decide, decide what they would do.

63
00:03:24.355 --> 00:03:29.045
This rule has has the implication

64
00:03:29.115 --> 00:03:31.965
that people try to get the information from,

65
00:03:32.435 --> 00:03:33.565
from the social feedback.

66
00:03:33.585 --> 00:03:35.325
So they see what others do.

67
00:03:35.825 --> 00:03:39.165
And one impact is they learn what others do.

68
00:03:39.785 --> 00:03:41.965
And when they learn what others do, they,

69
00:03:42.355 --> 00:03:46.165
they get information about how the, how the taste

70
00:03:46.185 --> 00:03:48.005
of the evaluator could look like,

71
00:03:48.115 --> 00:03:50.725
what the evaluator would actually like.

72
00:03:51.305 --> 00:03:55.005
And this leads to conformity.

73
00:03:55.425 --> 00:03:57.165
But there is also another effect,

74
00:03:57.305 --> 00:03:59.245
and that's the selection process

75
00:04:00.625 --> 00:04:03.475
when you are quite uncertain about the taste

76
00:04:03.535 --> 00:04:04.555
of the evaluator.

77
00:04:05.145 --> 00:04:09.115
It's good to be in the majority if punishment applies,

78
00:04:09.185 --> 00:04:10.795
because then the risk

79
00:04:10.815 --> 00:04:13.075
to be punishment is shared among more people.

80
00:04:13.575 --> 00:04:16.835
And when reward is relevant, the, it's good

81
00:04:16.835 --> 00:04:17.915
to be in a minority

82
00:04:17.915 --> 00:04:20.635
because then you have a get better chance

83
00:04:20.695 --> 00:04:22.395
to be selected individually.

84
00:04:22.775 --> 00:04:25.395
And these two rules interact,

85
00:04:25.855 --> 00:04:27.195
and in the case of punishment,

86
00:04:27.195 --> 00:04:30.235
both go in the same direction towards conformity.

87
00:04:30.535 --> 00:04:33.555
In the case of reward, there is a conflict between the two.

88
00:04:33.975 --> 00:04:37.525
And it depends on how,

89
00:04:37.625 --> 00:04:40.285
how informative the social feedback is.

90
00:04:40.305 --> 00:04:43.845
And when the social feedback is not too informative, it,

91
00:04:44.065 --> 00:04:47.085
it makes sense even, to be, to,

92
00:04:47.105 --> 00:04:48.525
be anti anticon conform.

93
00:04:48.705 --> 00:04:51.805
And one studies conformity.

94
00:04:52.505 --> 00:04:56.605
So the, the, the first studies on conformity were done

95
00:04:56.605 --> 00:05:01.205
by Solomon Ash, and he used confederates who did, who took

96
00:05:01.955 --> 00:05:05.165
strange or very uncommon, unlikely decisions.

97
00:05:05.585 --> 00:05:08.765
And then when people copy these unlikely decisions,

98
00:05:09.345 --> 00:05:12.045
he concludes that this must have something to do

99
00:05:12.045 --> 00:05:14.045
with conformity, that they adjust to the others.

100
00:05:15.025 --> 00:05:18.245
But that means that those who do are not conform.

101
00:05:18.395 --> 00:05:20.645
It's not clear whether they intentionally want

102
00:05:20.645 --> 00:05:22.885
to be different or whether they just,

103
00:05:23.785 --> 00:05:25.685
are independent.

104
00:05:26.265 --> 00:05:27.485
And what we wanted,

105
00:05:27.505 --> 00:05:30.845
we wanted the design in which we can distinguish conformity

106
00:05:30.845 --> 00:05:33.525
and anticon conformity from independence.

107
00:05:34.815 --> 00:05:38.795
And as I said, we define conformity as a shift

108
00:05:39.135 --> 00:05:41.875
of these individual decisions towards the decision

109
00:05:41.935 --> 00:05:43.035
of a majority and

110
00:05:43.035 --> 00:05:45.875
otherwise in anticon conformity away from the majority.

111
00:05:46.615 --> 00:05:49.195
And this is about behavior.

112
00:05:49.615 --> 00:05:51.675
And ideally, we would like

113
00:05:51.675 --> 00:05:56.595
to observe the behavior without social feedback and with

114
00:05:56.655 --> 00:06:00.925
Social feedback, this triggers a lot of consistency.

115
00:06:01.465 --> 00:06:05.365
And for that reason, we use a slight twist.

116
00:06:05.785 --> 00:06:07.365
We use triplet of options.

117
00:06:07.585 --> 00:06:10.485
So we have three options, let's say A, B, and C.

118
00:06:10.785 --> 00:06:14.165
And we ask people about their choice between A and B,

119
00:06:15.545 --> 00:06:19.045
and we ask them about P and C without social feedback.

120
00:06:19.265 --> 00:06:21.485
And then we ask them about A

121
00:06:21.545 --> 00:06:26.405
and C with social feedback from the decisions between A

122
00:06:26.405 --> 00:06:29.205
and P and P and C, we derive

123
00:06:29.515 --> 00:06:32.445
what they would decide in A and C,

124
00:06:32.825 --> 00:06:35.525
and then we can identify conformity

125
00:06:35.525 --> 00:06:36.765
and anticon conformity

126
00:06:36.865 --> 00:06:39.925
as a deviation from this predicted choice.

127
00:06:41.375 --> 00:06:45.795
So this is our main technique, and we do it in a way.

128
00:06:46.175 --> 00:06:48.755
So we have all, three people in a group.

129
00:06:48.935 --> 00:06:51.475
So we can actually quite efficiently,

130
00:06:51.505 --> 00:06:52.795
conduct this experiment.

131
00:06:52.855 --> 00:06:56.275
We have different domains in which we,

132
00:06:56.345 --> 00:06:57.515
conduct these experiments.

133
00:06:57.515 --> 00:07:01.915
We have knowledge questions, and we have art tastes.

134
00:07:02.095 --> 00:07:03.395
So that's a different domain.

135
00:07:03.815 --> 00:07:06.475
Uh, there, in, in the knowledge question,

136
00:07:06.475 --> 00:07:09.075
there was no specific reward for being correct

137
00:07:09.215 --> 00:07:13.595
or so in the art taste, they got a postcard of,

138
00:07:14.015 --> 00:07:17.075
this art piece so that they had an incentive

139
00:07:17.255 --> 00:07:21.315
to choose something according, to their taste roughly.

140
00:07:23.125 --> 00:07:26.935
Because we didn't observe that much unconformity,

141
00:07:27.035 --> 00:07:29.455
we thought it might be necessary

142
00:07:29.595 --> 00:07:32.935
to have also an environment in which,

143
00:07:33.115 --> 00:07:34.775
the choice is not so restricted.

144
00:07:34.875 --> 00:07:37.655
So where kind of creativity is relevant.

145
00:07:37.675 --> 00:07:40.255
And we designed and color

146
00:07:40.895 --> 00:07:44.775
creation task in which people could construct colors out of,

147
00:07:44.955 --> 00:07:46.495
of the RGB cubes.

148
00:07:46.495 --> 00:07:49.495
So, so they could, mix colors

149
00:07:49.715 --> 00:07:51.215
and so generate new colors.

150
00:07:51.275 --> 00:07:54.295
And essentially the same procedures, occurs.

151
00:07:54.515 --> 00:07:55.895
People,

152
00:07:56.875 --> 00:07:59.615
decide on a color without social feedback,

153
00:07:59.615 --> 00:08:03.135
and then they can correct their color with social feedback.

154
00:08:03.235 --> 00:08:05.695
So that's kind of our main treatment.

155
00:08:05.915 --> 00:08:09.215
The theoretical model is based on the binary decision.

156
00:08:10.075 --> 00:08:14.495
And in this model there are a group of three people,

157
00:08:15.035 --> 00:08:19.555
who, who from which two,

158
00:08:19.665 --> 00:08:22.715
take the decision unconditionally, this binary decision.

159
00:08:22.735 --> 00:08:25.435
And one person is informed about the decision of the others,

160
00:08:25.495 --> 00:08:28.595
and con can condition on what the others do.

161
00:08:30.335 --> 00:08:32.395
And then an evaluator rewards

162
00:08:32.395 --> 00:08:35.395
or punished one in the group without knowing

163
00:08:36.215 --> 00:08:37.915
who took the conditional decision.

164
00:08:38.295 --> 00:08:40.915
So just selects one of the three people.

165
00:08:41.055 --> 00:08:43.635
So of course when all three take the same choice,

166
00:08:43.665 --> 00:08:45.715
it's not a particularly interesting decision.

167
00:08:45.855 --> 00:08:49.155
So essentially it's interesting whether chooses the minority

168
00:08:49.175 --> 00:08:51.435
or the majority in this case, because there are three.

169
00:08:51.975 --> 00:08:56.475
And then we analyze, the situation

170
00:08:58.175 --> 00:09:01.235
for different preferences of the evaluator.

171
00:09:01.415 --> 00:09:05.715
So we look at when the evaluator decides based on salience,

172
00:09:05.775 --> 00:09:09.115
and we investigate when the evaluator

173
00:09:09.905 --> 00:09:13.555
decides based on the own preference on the own taste.

174
00:09:14.695 --> 00:09:19.035
And, what we find is then,

175
00:09:19.095 --> 00:09:21.715
so we have four sit, four situations

176
00:09:21.715 --> 00:09:22.835
that we have to analyze.

177
00:09:22.895 --> 00:09:27.595
So there is, so the, it depends on punishment

178
00:09:27.595 --> 00:09:30.795
and reward and depends on the preference of the evaluator.

179
00:09:31.465 --> 00:09:33.875
When the evaluator,

180
00:09:34.135 --> 00:09:36.875
so let's first look at the conditional decision

181
00:09:37.575 --> 00:09:39.595
in the salient treatment is totally simple.

182
00:09:40.015 --> 00:09:43.555
In the salient treatment, when there is punishment, you,

183
00:09:43.695 --> 00:09:46.235
you go with the majority, if this is possible.

184
00:09:46.375 --> 00:09:49.355
If it's reward, you try to single out.

185
00:09:49.575 --> 00:09:51.755
So that's simple on that side.

186
00:09:52.335 --> 00:09:55.715
In the homophily tree, in the homophily situation.

187
00:09:55.715 --> 00:09:58.435
So if the evaluator decides,

188
00:09:58.525 --> 00:10:01.795
based on the own preference, how to say,

189
00:10:01.945 --> 00:10:06.675
then on the punishment situation, it's also, it goes,

190
00:10:06.905 --> 00:10:11.285
towards, go going with the majority,

191
00:10:11.655 --> 00:10:12.685
going with the majority.

192
00:10:13.445 --> 00:10:14.685
Actually there is a slight twist.

193
00:10:14.825 --> 00:10:17.725
So we also have a preference for the,

194
00:10:17.865 --> 00:10:19.765
for sticking to the own taste.

195
00:10:20.425 --> 00:10:23.165
And so you need, if you too,

196
00:10:23.665 --> 00:10:26.605
if you too much care about your own taste, then

197
00:10:26.605 --> 00:10:28.765
of course you do not switch,

198
00:10:29.505 --> 00:10:32.125
in the reward treatment there.

199
00:10:33.485 --> 00:10:37.425
It depends on a parameter that describes

200
00:10:38.005 --> 00:10:39.865
how strongly the preferences

201
00:10:39.925 --> 00:10:42.305
of the different people are correlated,

202
00:10:42.955 --> 00:10:47.585
which means it depends on how informative the,

203
00:10:48.525 --> 00:10:50.665
the decision of the others are about the

204
00:10:50.665 --> 00:10:51.945
preference of the evaluator.

205
00:10:52.365 --> 00:10:55.145
And if this information is strong,

206
00:10:55.855 --> 00:10:58.505
then conformity is possible.

207
00:10:58.535 --> 00:11:01.265
Otherwise anticon conformity is possible.

208
00:11:01.405 --> 00:11:03.585
That's roughly what we find there,

209
00:11:03.585 --> 00:11:07.105
and I think that's the most interesting, case.

210
00:11:07.925 --> 00:11:11.545
We have also analyzed the behavior on the,

211
00:11:12.915 --> 00:11:14.735
for the unconditional.

212
00:11:14.835 --> 00:11:18.015
So for the decisions without social feedback,

213
00:11:18.765 --> 00:11:21.175
they're in all situations with the exception

214
00:11:21.175 --> 00:11:22.335
of salience rewards,

215
00:11:22.335 --> 00:11:24.935
people just choose according to that taste.

216
00:11:25.235 --> 00:11:28.455
In the salience, in the situation

217
00:11:28.455 --> 00:11:30.855
where salience matters in the reward treatment,

218
00:11:30.855 --> 00:11:35.495
there are two, two results, two domains.

219
00:11:35.495 --> 00:11:38.815
So first we find that for the evaluator,

220
00:11:39.515 --> 00:11:41.735
the own taste was more important than salience.

221
00:11:41.995 --> 00:11:45.095
So this is, one result.

222
00:11:45.875 --> 00:11:48.445
Uh, but nevertheless,

223
00:11:50.035 --> 00:11:53.105
their behavior incentivized

224
00:11:53.385 --> 00:11:55.975
Conformity in the punishment situation

225
00:11:55.975 --> 00:12:00.375
and anticon conformity in the, in the reward situation.

226
00:12:00.555 --> 00:12:04.495
And that's, and it's interesting that we do not observe

227
00:12:04.525 --> 00:12:06.975
that much anticon conformity in the, in the,

228
00:12:07.075 --> 00:12:08.455
in the reward situation.

229
00:12:08.455 --> 00:12:09.935
So that's kind of interesting.

230
00:12:09.935 --> 00:12:14.405
So there is a tendency, towards, conformity.

231
00:12:14.405 --> 00:12:16.205
There is also, a difference

232
00:12:16.205 --> 00:12:17.565
between the different domains.

233
00:12:17.565 --> 00:12:22.435
So most conformity in, in the domain

234
00:12:22.465 --> 00:12:26.675
with the questions, less in the situation with taste,

235
00:12:26.675 --> 00:12:27.715
with the art taste,

236
00:12:27.855 --> 00:12:31.835
and least in the situation with colors.

237
00:12:32.655 --> 00:12:37.425
Now, I think what, what, what's interesting,

238
00:12:38.755 --> 00:12:42.305
let's say as, as a general message for is

239
00:12:42.935 --> 00:12:44.585
when we relate it to the beginning.

240
00:12:44.805 --> 00:12:48.495
So what are environments in which

241
00:12:49.395 --> 00:12:50.735
reward is used and

242
00:12:50.805 --> 00:12:54.335
what are environments in which punishment is used

243
00:12:54.675 --> 00:12:59.655
that's actually related to, in, in a reasonable way

244
00:12:59.925 --> 00:13:02.055
with conformity and non-conformity.

245
00:13:02.355 --> 00:13:07.295
So I do not want a pilot who is particularly creative

246
00:13:08.075 --> 00:13:09.135
in what he does,

247
00:13:09.675 --> 00:13:12.895
but I want to have an artist who is particularly creative.

248
00:13:13.115 --> 00:13:17.935
And in that sense, the, it, it bends well back

249
00:13:17.935 --> 00:13:21.945
to, so it, it, so it really, the, the environment

250
00:13:22.815 --> 00:13:27.665
fits usually the the need for conformity

251
00:13:27.685 --> 00:13:28.825
or anticon conformity.

252
00:13:28.885 --> 00:13:31.945
And I think that's, the general message we find.

Exploring Anticonformity and the Influence of Reward vs. Punishment

00:00:00 - We were interested in two questions. One, refer refers to heterogeneity in conformity. And there we were specifically interested whether there are also people who specifically are anticon conform. And that means in our context, conformity means that people do something in the direction of others and deviating from their own taste, so to say.

00:00:35 - move away from the others, so do something different, but also against their own taste. So that's essentially the, the, these two behaviors. And anticon conformity is rarely studied. And, actually we find it's also rare in, in our study, and I will talk about that later. The second question, relates how this interacts

00:01:03 - And the idea here is that there are environments in which reward is dominant, and there are environments in which punishment is dominant. And that means, for instance, when you think about art or science, their reward is much more relevant because, it's though you are rewarded for good stuff. And if you have once in a while a bad paper,

Behavioral Responses to Reward and Punishment Dynamics

00:01:35 - On the other hand, if you are a driver or a pilot, then nobody cares about your best performance. But people really care about the worst performance, and so you are more likely to be punished for, bad performance. And, and in that sense, punishment is more, much more relevant in this case. And generally that's also the case. Generally when there is a hiring that's more a rewarding environment. In the case of a threat of layoffs,

00:02:05 - And the idea is that when there is this, this risk of being punished, people try to hide while when there is a reward possible, people tried to stand out. And, this is was kind of our starting point with respect to reward and punishment. the idea was that this evaluators, so those

00:02:38 - and, they, they could target the special ones, but this is not the only possibility. It's also possible that they care more about that they do something similar as what they would do. We analyze this theoretically and interestingly the same occurs. So in the case of punishment, There is a shift towards,

Conformity Dynamics: The Role of Reward and Punishment

00:03:05 - If there is a reward, there can be a shift even to anticon conformity and less conformity. This rule that evaluate this care about what, where the people do what they would do or decide, decide what they would do. This rule has has the implication that people try to get the information from, from the social feedback. So they see what others do.

00:03:39 - And when they learn what others do, they, they get information about how the, how the taste of the evaluator could look like, what the evaluator would actually like. And this leads to conformity. But there is also another effect, and that's the selection process when you are quite uncertain about the taste of the evaluator. It's good to be in the majority if punishment applies,

00:04:10 - to be punishment is shared among more people. And when reward is relevant, the, it's good to be in a minority because then you have a get better chance to be selected individually. And these two rules interact, and in the case of punishment, both go in the same direction towards conformity. In the case of reward, there is a conflict between the two. And it depends on how, how informative the social feedback is.

Understanding Conformity and Anticonformity in Decision-Making

00:04:44 - it makes sense even, to be, to, be anti anticon conform. And one studies conformity. So the, the, the first studies on conformity were done by Solomon Ash, and he used confederates who did, who took strange or very uncommon, unlikely decisions. And then when people copy these unlikely decisions, he concludes that this must have something to do

00:05:15 - But that means that those who do are not conform. It's not clear whether they intentionally want to be different or whether they just, are independent. And what we wanted, we wanted the design in which we can distinguish conformity and anticon conformity from independence. And as I said, we define conformity as a shift of these individual decisions towards the decision of a majority and

00:05:46 - And this is about behavior. And ideally, we would like to observe the behavior without social feedback and with Social feedback, this triggers a lot of consistency. And for that reason, we use a slight twist. We use triplet of options. So we have three options, let's say A, B, and C. And we ask people about their choice between A and B,

Exploring Conformity and Anticonformity through Experimentation

00:06:19 - And then we ask them about A and C with social feedback from the decisions between A and P and P and C, we derive what they would decide in A and C, and then we can identify conformity and anticon conformity as a deviation from this predicted choice. So this is our main technique, and we do it in a way.

00:06:48 - So we can actually quite efficiently, conduct this experiment. We have different domains in which we, conduct these experiments. We have knowledge questions, and we have art tastes. So that's a different domain. there, in, in the knowledge question, there was no specific reward for being correct or so in the art taste, they got a postcard of, this art piece so that they had an incentive

00:07:23 - Because we didn't observe that much unconformity, we thought it might be necessary to have also an environment in which, the choice is not so restricted. So where kind of creativity is relevant. And we designed and color creation task in which people could construct colors out of, of the RGB cubes. So, so they could, mix colors

Decision-Making Dynamics: The Role of Social Feedback and Evaluator Preferences

00:07:51 - And essentially the same procedures, occurs. People, decide on a color without social feedback, and then they can correct their color with social feedback. So that's kind of our main treatment. The theoretical model is based on the binary decision. And in this model there are a group of three people, who, who from which two, take the decision unconditionally, this binary decision.

00:08:25 - and con can condition on what the others do. And then an evaluator rewards or punished one in the group without knowing who took the conditional decision. So just selects one of the three people. So of course when all three take the same choice, it's not a particularly interesting decision. So essentially it's interesting whether chooses the minority or the majority in this case, because there are three. And then we analyze, the situation

00:09:01 - So we look at when the evaluator decides based on salience, and we investigate when the evaluator decides based on the own preference on the own taste. And, what we find is then, so we have four sit, four situations that we have to analyze. So there is, so the, it depends on punishment and reward and depends on the preference of the evaluator.

Influence of Rewards and Punishments on Decision Making

00:09:34 - so let's first look at the conditional decision in the salient treatment is totally simple. In the salient treatment, when there is punishment, you, you go with the majority, if this is possible. If it's reward, you try to single out. So that's simple on that side. In the homophily tree, in the homophily situation. So if the evaluator decides, based on the own preference, how to say,

00:10:06 - towards, go going with the majority, going with the majority. Actually there is a slight twist. So we also have a preference for the, for sticking to the own taste. And so you need, if you too, if you too much care about your own taste, then of course you do not switch, in the reward treatment there.

00:10:38 - how strongly the preferences of the different people are correlated, which means it depends on how informative the, the decision of the others are about the preference of the evaluator. And if this information is strong, then conformity is possible. Otherwise anticon conformity is possible. That's roughly what we find there,

Influences of Salience and Taste on Conformity and Anticonformity

00:11:07 - We have also analyzed the behavior on the, for the unconditional. So for the decisions without social feedback, they're in all situations with the exception of salience rewards, people just choose according to that taste. In the salience, in the situation where salience matters in the reward treatment, there are two, two results, two domains.

00:11:39 - the own taste was more important than salience. So this is, one result. but nevertheless, their behavior incentivized Conformity in the punishment situation and anticon conformity in the, in the reward situation. And that's, and it's interesting that we do not observe that much anticon conformity in the, in the,

00:12:08 - So that's kind of interesting. So there is a tendency, towards, conformity. There is also, a difference between the different domains. So most conformity in, in the domain with the questions, less in the situation with taste, with the art taste, and least in the situation with colors. Now, I think what, what, what's interesting,

Conformity and Anticonformity in Reward and Punishment Environments

00:12:42 - when we relate it to the beginning. So what are environments in which reward is used and what are environments in which punishment is used that's actually related to, in, in a reasonable way with conformity and non-conformity. So I do not want a pilot who is particularly creative in what he does,

00:13:28 - And in that sense, the, it, it bends well back to, so it, it, so it really, the, the environment fits usually the the need for conformity or anticon conformity. And I think that's, the general message we find.

Faculti AI can make mistakes

Research

Policy

Analysis