Académique Documents
Professionnel Documents
Culture Documents
Forthcoming
in
Journal
of
Economic
Behavior
and
Organiza6on
(2015)
CompeJJveness
Measures
Performance
in
tournaments
vs
piece-rate
schemes
Self-selecJon
into
tournaments
vs
piece-rate
schemes
What
do
we
know
about
adults?
Women
are
(if
anything)
Less
compeJJve
Math
related
tasks
Less
risk
taking
More
altruisJc
CompeJJveness
in
children/adolescents
o If
gender
gap,
boys
more
compeJJve
Israel,
Sweden,
Austria,
Colombia:
Running,
skipping
rope,
dancing,
children
and
adolescents
Austria
and
UK:
Math
and
mazes,
children
and
adolescents
o ExcepJon
Sweden:
Skipping
rope,
math
Performance
change
in
both
Gneezy
&
RusJchini
2004,
Booth
&
Nolen
2009,
Dreber
et
al.
2011,
Suber
&
Rutzler
2010,
Crdenas
et
al.
2011
Running
Dierence
between
GR04
and
DvER11
Colombia
vs
Sweden
Unexpected
ndings:
No
gender
gaps
in
Colombia
in
all
tasks,
but
some
gender
gaps
in
Sweden
Gneezy
&
RusJchini
2004,
Booth
&
Nolen
2011,
Dreber
et
al.
2011,
Crdenas
et
al.
2011,
Dreber
et
al.
2011,
Grosse
&
Riener
2010,
Gunther
et
al.
2009,
Wozniak
et
al.
2010
Harbaugh et al. 2002; Booth & Nolen 2011, Borghans et al. 2009, Crdenas et al. 2011,
Impact
of
tasks
Some
inuence
ShooJng
baskets
vs
solving
anagrams
Running
vs
skipping
rope
and
dancing
Solving
mazes/math
vs
word
search
Gneezy
&
RusJchini
2004b,
Dreber
et
al.
2011,
Grosse
&
Riener
2010,
Gunther
et
al.
2009,
Wozniak
et
al.
2010,
Crdenas
et
al.
2011
o Measure
of
compeJJveness
speed
in
stage
2
speed
in
stage
1
number
of
jumps
in
stage
2
number
of
jumps
in
stage
1
could
also
use
a
relaJve
measure
Age Group
Running task
Gender
Round 1
(Ind.)
Round 2
(Comp.)
SR testa
(pvalue)
Round 1
(Ind.)
Round 2
(Comp.)
SR test
(pvalue)
Nb
Grades 2 to 6
Boys
Girls
17.39
18.95
16.98
18.32
< 0.001
< 0.001
5.92
8.38
6.39
7.12
0.340
0.692
154/153
147/145
Grades 7 to 9
Boys
Girls
15.13
16.94
14.92
16.27
< 0.001
< 0.001
67.10
70.92
72.27
79.45
< 0.001
< 0.001
100/96
91/88
All Grades
Boys
Girls
16.50
18.19
16.17
17.54
< 0.001
< 0.001
29.66
32.50
31.79
34.44
0.002
0.016
254/249
238/233
a
b
Wilcoxon matched pairs signed-rank test, testing whether (Ind.)=(Comp.), henceforth SR test in the tables
5 boys and 5 girls took part in the running task, but did not complete the skipping rope task.
Running
Seconds
-.8
-.6
-.4
p=0.001
p=0.023
-0.66
-0.63
-0.41
-.2
-0.21
0
Grades 2 to 6
Grades 7 to 9
Skipping Rope
p=0.527
10
Jumps
8
6
8.93
p=0.800
5.51
4
2
0
0.47
-1.26
Grades 2 to 6
Grades 7 to 9
Boys
Girls
Sample of
Boys
Girls
pvalue
Boys
Girls
pvalue
Boys
Girls
-0.25
-0.46
-0.49
-0.76
0.008
0.006
162/91
94/144
2.12
2.57
2.74
2.60
0.558
0.996
129/118
118/118
a
b
However, in the skipping rope task we do not find any gender differences in competitiveness (fo
the whole sample: p = 0.917).13
All the numbers and mathematical operations were randomly generated to insure th
he level of diculty of the math task was the same throughout all the stages of th
Math task
Years 4 to 5
Years 6 to 7
Years 8 to 10
1 + 12 = . . .
3 + 5 = ...
11 + 4 = . . .
17 + 18 = . . .
10 + 23 = . . .
82 + 18 = . . .
48 + 10 = . . .
47 + 14 = . . .
39 + 6 = . . .
68 + 16 = . . .
93 + 67 = . . .
63 38 = . . .
2 38 = . . .
71 + 52 = . . .
89 47 = . . .
96 + 93 + 3 = . . .
33 9 85 = . . .
83 + 97 + 14 = . . .
31 39 + 28 = . . .
9 41 75 = . . .
Figure 20 shows an example of the word search task that was used in the experimen
he children had to find and circle words in any direction on a straight line. Since the
ord search puzzles were not generated by a computer there might have been slig
&
'
&
&
25
25
<
25
>
&
&
&
25
&
<
25
25
&
>
25
25
<
&
25
&
CompeJJveness
i
n
M
ath
a
nd
V
erbal
t
asks
Table 3 Individual and competitive performance in the mathematical
and verbal tasks
Math taska
Age Group
Gender
Verbal taskb
Stage 1
(Ind.)
Stage 2
(Comp.)
SR test
(pvalue)
Stage 1
(Ind.)
Stage 2
(Comp.)
SR test
(pvalue)
Gr. 2 to 6
Boys
Girls
17.44
16.36
20.83
20.12
< 0.001
< 0.001
5.31
5.80
5.56
6.09
0.071
0.137
224c
204
Gr. 7 to 10
Boys
Girls
8.12
7.96
10.32
8.29
< 0.001
< 0.001
9.20
10.61
9.74
11.61
0.112
0.003
164
173
All Grades
Boys
Girls
13.50
12.51
16.18
15.62
< 0.001
< 0.001
6.95
8.01
7.33
8.62
0.018
0.001
388
377
for all pair-wise comparisons). However, while boys solve slightly more exercises in the math
p=0.322
5
4
p=0.327
3.76
3.39
3
2
2.36
1.71
1
0
Grades 2 to 6
Grades 7 to 10
Verbal Task
p=0.195
Number of Words
Number of Exercises
Math Task
.5
0.99
p=0.920
0.54
0.25
0.29
Grades 2 to 6
Boys
Grades 7 to 10
Girls
Compete
.8
.6
p=0.855
0.61
p=0.924
0.60
.4
0.43
0.43
.2
0
Grades 2 to 6
Grades 7 to 10
Verbal Task
Compete
.8
p=0.942
p=0.519
.6
.4
0.56
0.60
0.59
0.53
.2
0
Grades 2 to 6
Grades 7 to 10
Boys
Girls
Risk
Preferences
o Choose
between
a
sure
amount
and
a
gamble,
10
points
if
heads,
0
points
if
tails
o Sure
amount
increasing
from
2
points
to
6
in
increments
of
1,
and
then
7.5
points
o Our
main
measure
of
risk
aversion
relies
on
the
unique
switching
point
(certainty
equivalent
taken
as
the
midpoint
of
the
switching
interval)
o 18%
of
subjects
are
inconsistent,
leaving
us
with
a
sample
of
624
subjects
o AlternaJve
measure
of
risk
taking:
number
of
risky
choices
Results
are
qualitaJvely
the
same
Risk Preferences
Girls
4
6
7.5
Sure amount in the switching row
10
.1
Fraction
.2
.3
Boys
.1
Fraction
.2
.3
.4
Grades 2 to 6
2
3
4
Number of risky choices
Fraction
.2
.3
.1
0
.1
Fraction
.2
.3
.4
Grades 7 to 10
.4
Grades 7 to 10
2
3
4
Number of risky choices
10
Fraction
.1
.2
0
Fraction
.1
.2
.3
All Grades
.3
All Grades
4
6
7.5
Sure amount in the switching row
2
3
4
Number of risky choices
4
6
7.5
Sure amount in the switching row
10
Risk
Preferences
Boys
are
more
risk
taking
than
girls
Risk Preferences
Average number of risky choices
p=0.116
3.24
3
2.86
2.52
2.48
Grades 2 to 6
Grades 7 to 10
Boys
Girls
Altruism from DG
Girls
donate
9%
more
on
average
then
boys
do,
but
doesnt
vary
with
age
Altruistic Preferences
Preferences
ByAltruistic
Age Group
and Gender
70
60
50
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
70
p=0.014
p=0.007
p=0.007
p=0.014
60
61.66
58.88
50
50.58
61.66
58.88
52.70
52.70
40
50.58
30
40
Grades 2 to 6
Grades 7 to 10
Boys
Girls
30
Grades 2 to 6
Grades 7 to 10
These results on risk preferences are in line with those of Cardenas et al. (2012), who find
that Colombian boys on average take 40% more
risk than girls, with
the corresponding number
Boys
Girls
in Sweden being 15%.
In this section we look at gender differences in altruism as measured via donations in a dictator
findpreferences
that girls give significantly
more with
than boys
(p < 0.001),
and this is the
These resultsgame.
on We
risk
are in line
those
of Cardenas
et case
al. for(2012), who find
both age groups, see Figure 5. Girls donate on average 60.16 points whereas boys donate 51.48
Summary
of
Results
o Girls
are
signicantly
more
compeJJve
than
boys
in
a
running
task
o In
all
other
tasks
and
measures,
girls
and
boys
are
equally
compeJJve
o Boys
and
girls
choose
to
compete
at
equal
rates
in
either
task
o Boys
are
more
risk
taking
than
girls
and
the
gender
gap
appears
around
the
age
of
puberty
o Boys
are
less
altruisJc
than
girls
o Culture
and
context
play
even
a
bigger
role
than
iniJally
expected,
and
not
always
in
predictable
ways
Thank You!