Background
In a 2 (informed vs. uninformed) X 2 (rational vs. irrational)
between-subjects design, participants read about one of three taboo
transactions (cancerous cell-phone tower; storing hazardous chemicals;
testing beauty products for side effects) and were asked to indicate to
what extent the seller benefited from each of them, to what extent the
buyer benefited from each of them, and the power balance in each of
them.
Attention check
What are the roles of Person A and Person B in the transaction
that took place?
The correct answer is: Person A paid money and Person B received
money
## `summarise()` has grouped output by 'informed', 'rational'. You can override
## using the `.groups` argument.
|
informed
|
rational
|
failcheck
|
passcheck
|
|
0
|
0
|
7
|
47
|
|
0
|
1
|
0
|
47
|
|
1
|
0
|
5
|
42
|
|
1
|
1
|
1
|
51
|
Ok, not bad. Pretty evenly distributed eligible participants across
conditions. And we’re left 187.
Demographics
Race
|
race
|
N
|
Perc
|
|
asian
|
14
|
7.49
|
|
black
|
14
|
7.49
|
|
hispanic
|
11
|
5.88
|
|
multiracial
|
5
|
2.67
|
|
white
|
142
|
75.94
|
|
NA
|
1
|
0.53
|
Gender
|
gender
|
N
|
Perc
|
|
man
|
105
|
56.15
|
|
woman
|
80
|
42.78
|
|
NA
|
2
|
1.07
|
Age
|
age_mean
|
age_sd
|
|
40.26738
|
11.58527
|
Education
|
edu
|
N
|
Perc
|
|
GED
|
48
|
25.67
|
|
2yearColl
|
22
|
11.76
|
|
4yearColl
|
73
|
39.04
|
|
MA
|
33
|
17.65
|
|
PHD
|
10
|
5.35
|
|
NA
|
1
|
0.53
|
Income
Analysis
Manipiulation check
Participants answered the following questions:
To what extent do you agree or disagree with the following statement
about Person B?
1. They were fully informed of the consequences of making this
transaction
2. They were fully capable of weighing the costs and the benefits of
this transaction when making their decision
|
informed
|
rational
|
checkinformed_M
|
checkinformed_SD
|
checkrational_M
|
checkrational_SD
|
|
0
|
0
|
2.11
|
1.70
|
2.23
|
1.56
|
|
0
|
1
|
2.00
|
1.78
|
3.28
|
2.39
|
|
1
|
0
|
5.26
|
1.94
|
3.43
|
2.30
|
|
1
|
1
|
6.22
|
0.76
|
6.16
|
0.99
|
CheckRational: Two-way ANOVA
|
Effect
|
DFn
|
DFd
|
F
|
p
|
p<.05
|
ges
|
|
informed
|
1
|
183
|
55.151
|
0.000
|
|
0.232
|
|
rational
|
1
|
183
|
47.230
|
0.000
|
|
0.205
|
|
informed:rational
|
1
|
183
|
9.439
|
0.002
|
|
0.049
|
Cool. Main effect for the “rational” manipulation as well.
Binarized manipulation checks
These items were measured on a 1 to 7 scale. Let’s see how many
people said over 4 when it was necessary.
Rational
|
rational
|
rational_over4
|
n
|
|
0
|
0
|
89
|
|
1
|
0
|
32
|
|
1
|
1
|
66
|
32 people didn’t “pass” the rational manipulation check.
Condition -> Benefit A
To avoid a three-way interaction, I’ll look at the effect of
condition on each party’s benefit separately. And only then I’ll show
everything in one plot.
Descriptives
|
informed
|
rational
|
benefit_A_M
|
benefit_A_SD
|
|
0
|
0
|
2.13
|
0.99
|
|
0
|
1
|
1.81
|
1.50
|
|
1
|
0
|
2.00
|
1.21
|
|
1
|
1
|
1.69
|
1.12
|
Benefit A: Two-way ANOVA
|
Effect
|
DFn
|
DFd
|
F
|
p
|
p<.05
|
ges
|
|
informed
|
1
|
183
|
0.490000
|
0.485
|
|
3.0e-03
|
|
rational
|
1
|
183
|
3.140000
|
0.078
|
|
1.7e-02
|
|
informed:rational
|
1
|
183
|
0.000231
|
0.988
|
|
1.3e-06
|
Cool. We’re not seeing an effect of condition on the buyer’s
benefit.
Condition -> Benefit B
Let’s take a look at the seller.
Descriptives
|
informed
|
rational
|
benefit_B_M
|
benefit_B_SD
|
|
0
|
0
|
-1.30
|
1.72
|
|
0
|
1
|
-1.21
|
1.64
|
|
1
|
0
|
-0.81
|
2.04
|
|
1
|
1
|
-0.29
|
1.85
|
Benefit B: Two-way ANOVA
|
Effect
|
DFn
|
DFd
|
F
|
p
|
p<.05
|
ges
|
|
informed
|
1
|
183
|
7.014
|
0.009
|
|
0.037
|
|
rational
|
1
|
183
|
1.278
|
0.260
|
|
0.007
|
|
informed:rational
|
1
|
183
|
0.656
|
0.419
|
|
0.004
|
Ok, so there’s main effect of the informed manipulation, but not the
rational manipulation. And no interaction.
Plot: Condition -> Buyer and Seller Benefit
Ok, this is pretty informative. Looks like we’re actually not too far
from an interaction. And if, in fact, there’s an interaction, the
message would be something like this: If a seller is not informed, no
amount of rationality with help them: They would be harmed as much as of
they were irrational. But if the seller is informed, rationality can
really help. It brings them almost up to 0 (neither harmed nor
benefited). We might have to up the power to see this intraction, but if
we do see it, I think it’s a pretty cool take-away.
Condition -> Power
Let’s take a look at the effect on power. Power was rated from -3
(Buyer has much more power) to 3 (Seller has much more power).
Descriptives
|
informed
|
rational
|
power_M
|
power_SD
|
|
0
|
0
|
-1.96
|
1.46
|
|
0
|
1
|
-1.77
|
1.37
|
|
1
|
0
|
-0.98
|
1.70
|
|
1
|
1
|
0.00
|
1.44
|
Power: Two-way ANOVA
|
Effect
|
DFn
|
DFd
|
F
|
p
|
p<.05
|
ges
|
|
informed
|
1
|
183
|
39.442
|
0.000
|
|
0.177
|
|
rational
|
1
|
183
|
7.126
|
0.008
|
|
0.037
|
|
informed:rational
|
1
|
183
|
3.218
|
0.074
|
|
0.017
|
Two main effects, but no interaction. Alright, that’s actually pretty
cool. It also looks like we’re not too far from an interaction. Let’s
visualize this.
Plot: Condition -> Power
Alright. Cool. Again, information is playing a very big role here.
Basically, if the seller is uninformed, the buyer clearly has
more power. That makes a lot of sense. Seller rationality doesn’t really
matter much here. When the seller is informed and
rational - the power is actually perfectly balanced. But when
they’re informed and irrational, the power gravitates back to
the buyer.
Supplementary analysis
Let’s break these down by transaction (stats aren’t really necessary
here because we’d be underpowered anyway)
Cancerous cell-phone tower
Condition -> Benefit
|
party
|
informed
|
rational
|
benefit_M
|
benefit_SD
|
|
buyer
|
0
|
0
|
2.83
|
0.39
|
|
buyer
|
0
|
1
|
1.50
|
2.26
|
|
buyer
|
1
|
0
|
2.27
|
1.03
|
|
buyer
|
1
|
1
|
2.22
|
0.94
|
|
seller
|
0
|
0
|
-2.83
|
0.39
|
|
seller
|
0
|
1
|
-1.67
|
2.34
|
|
seller
|
1
|
0
|
-1.45
|
1.84
|
|
seller
|
1
|
1
|
-2.11
|
1.13
|
Condition -> Power balance
|
informed
|
rational
|
power_M
|
power_SD
|
|
0
|
0
|
-1.42
|
2.07
|
|
0
|
1
|
-2.50
|
0.84
|
|
1
|
0
|
-0.68
|
1.96
|
|
1
|
1
|
0.17
|
1.34
|
Storing hazardous chemicals
Condition -> Benefit
|
party
|
informed
|
rational
|
benefit_M
|
benefit_SD
|
|
buyer
|
0
|
0
|
2.06
|
1.00
|
|
buyer
|
0
|
1
|
1.96
|
1.22
|
|
buyer
|
1
|
0
|
1.80
|
1.40
|
|
buyer
|
1
|
1
|
2.00
|
1.05
|
|
seller
|
0
|
0
|
-1.00
|
1.75
|
|
seller
|
0
|
1
|
-1.35
|
1.23
|
|
seller
|
1
|
0
|
-0.40
|
2.12
|
|
seller
|
1
|
1
|
0.50
|
1.51
|
Condition -> Power balance
|
informed
|
rational
|
power_M
|
power_SD
|
|
0
|
0
|
-2.12
|
1.31
|
|
0
|
1
|
-1.50
|
1.36
|
|
1
|
0
|
-0.80
|
1.40
|
|
1
|
1
|
-0.40
|
1.51
|
Testing beauty products for dangerous side effects
Condition -> Benefit
|
party
|
informed
|
rational
|
benefit_M
|
benefit_SD
|
|
buyer
|
0
|
0
|
1.74
|
1.05
|
|
buyer
|
0
|
1
|
1.67
|
1.68
|
|
buyer
|
1
|
0
|
1.60
|
1.35
|
|
buyer
|
1
|
1
|
1.13
|
1.06
|
|
seller
|
0
|
0
|
-0.58
|
1.64
|
|
seller
|
0
|
1
|
-0.80
|
1.97
|
|
seller
|
1
|
0
|
0.20
|
2.04
|
|
seller
|
1
|
1
|
0.78
|
1.28
|
Condition -> Power balance
|
informed
|
rational
|
power_M
|
power_SD
|
|
0
|
0
|
-2.16
|
1.07
|
|
0
|
1
|
-1.93
|
1.49
|
|
1
|
0
|
-1.80
|
1.14
|
|
1
|
1
|
0.04
|
1.52
|
Main analyses excluding those who didn’t pass manipulation
checks
|
informed
|
rational
|
n
|
benefit_B_M
|
benefit_B_SD
|
|
0
|
0
|
47
|
-1.30
|
1.72
|
|
0
|
1
|
17
|
-0.29
|
1.86
|
|
1
|
0
|
32
|
-0.28
|
2.05
|
|
1
|
1
|
48
|
-0.23
|
1.86
|
|
Effect
|
DFn
|
DFd
|
F
|
p
|
p<.05
|
ges
|
|
informed
|
1
|
140
|
2.558
|
0.112
|
|
0.018
|
|
rational
|
1
|
140
|
2.438
|
0.121
|
|
0.017
|
|
informed:rational
|
1
|
140
|
1.981
|
0.162
|
|
0.014
|
underpowered, but it looks like the effect, for those who passed the
manipulation checks, only holds for the difference between the
“irrational/uninformed condition” vs. the rest of the sample.
