Content uploaded by David Kravitz
Author content
All content in this area was uploaded by David Kravitz on May 27, 2014
Content may be subject to copyright.
Journal
of
Personality
and
Social
Psychology
1986, Vol.
50, No. 5,
936-941
Copyright 1986
by the
American Psychological
Association,
Inc
0022-3514/86/S00.75
Ringelmann Rediscovered:
The
Original Article
David
A.
Kravitz
University
of
Kentucky
Barbara
Martin
Lexington,
Kentucky
A
current
focus
of
research
on
individual versus group performance
is
social
loafing,
the
decrease
in
individual
effort
that
occurs when
the
individual works within
a
cooperative group rather than alone.
Theory
and
research
on
this issue have been strongly influenced
by
results reported
in
Moede
(1927)
and
attributed
to
Ringelmann. Despite
the
importance
and
frequent
citation
of
Ringelmann's
study,
the
location
of his
original report
has
been
a
mystery.
In
this article Ringelmann's original article
is
discussed
and
described
in
detail. Ringelmann
was a
French agricultural engineer
who
gathered
his
data
in the
1880s.
He
(Ringelmann, 1913b) reported
the
performance
of
human workers
as a
function
of
the
method that
the
workers used
to
push
or
pull
a
load horizontally. Comparison
of
individual
and
group performance
was a
secondary interest
in
this experiment. Ringelmann interpreted
the
obtained decrement
in
group performance
in
terms
of
coordination loss, although
he was
also aware
of
motivational factors. Ringelmann's results
are
briefly
related
to
contemporary theory
and
research.
For
decades
social psychologists have been studying
the
effects
of
group membership
on
individual performance (e.g., Allport,
1924;
Dashiell, 1930; Triplett, 1898; Zajonc, 1965)
and the
dif-
ference
between individual
and
group performance (e.g., Hill,
1982;
Laughlin,
1980;
Lorge,
Fox, Davitz,
&
Brenner, 1958;
Shaw,
1932;
Thomas
&
Fink,
1963).
One
phenomenon within this area
that
has
recently received considerable attention
is
social
loafing,
where
social
loafing
refers
to the
"decrease
in
individual
effort
when
performing
in
groups
as
compared
to
when they perform
alone"
(Latane, Williams,
&
Harkins,
1979,
p.
822).
In
line with
Steiner's
(1972)
discussion, this
effect
has
generally been
ex-
plained
as due to
either coordination
or
motivation losses,
or to
both.
Contemporary work
on
social
loafing
was
initiated
by
Ingham,
Levinger,
Graves,
and
Peckham
(1974).
They
in
turn were stim-
ulated
by
data attributed
to
Ringelmann
in
Moede
(1927).
The
importance
of
Ringelmann's work
for the
group performance
area
in
general
and for
social
loafing
in
particular
is
well illustrated
by
the
title
of the
Ingham
et
al.
(1974)
article, "The Ringelmann
Effect:
Studies
of
Group
Size
and
Group Performance." Ingham
et
al.
(1974)
bemoaned
the
fact
that
"Apart
from
Moede's sum-
mary description,
we
have been unable
to find any
additional
information about
the
Ringelmann study
despite
its
widespread
citation
in
American literature
on
group performance"
(p.
372).
Indeed this study
has
been widely cited during
the
past
few de-
cades.
The
chapter that apparently introduced Ringelmann's
The
research
for
this article
was
completed while
the first
author
was
a
Fulbright
senior professor
at the
University
of
Freiburg
in the
Federal
Republic
of
Germany.
He
would like
to
thank
the
Fulbright Commission
for
granting
the
year
in
Germany,
and the
University
of
Freiburg
for
providing
the
necessary
facilities
and
resources.
All
quotations
from
French
and
German sources
are
given
in
English
in the
text—translated
by the
present
authors—and
in the
original lan-
guage
in
Footnotes
1 and
3-10.
Communication concerning this article should
be
addressed
to
David
A.
Kravitz, Department
of
Psychology,
115
Kastle Hall, University
of
Kentucky,
Lexington,
Kentucky 40506-0044.
work
to
English-speaking researchers
was
written
by
Dashiell
(1935).
Subsequently, Steiner (1972) devoted several pages
to
il
in
his
excellent
and
influential
monograph. Largely
due to the
important works
of
Dashiell
and
Steiner,
and to
Moede
(1927)
from
which they drew their information, Ringelmann's work
has
been regularly covered
in
monographs
and
texts
on
small group
behavior
(e.g., Davis, 1969;
Forsyth,
1983; Hare, 1976; Zajonc,
1966).
In
the
secondary source, Moede
(1927),
Ringelmann's results
were
given
in a
small
figure
with
the
name Ringelmann
in pa-
rentheses
at the end of the figure
caption. There
is no
citation
and
no
mention
of
Ringelmann
in the
text. Furthermore, Moede
consistently
used
the
word
we
when discussing
the
results. There
seems
to
have been
a
general assumption
that
Ringelmann
was
a
German psychologist (e.g., Zajonc, 1966,
p.
102), possibly
Moede's student (e.g., Ingham
et
al.,
1974,
p.
371),
who
gathered
his
data
in the
1920s (e.g., Steiner,
1972,
p.
32).
It has
also been
stated
that Ringelmann never published
his
study (Latane
et
al.,
1979,
p.
822). These inferences were quite reasonable given
the
manner
in
which Moede
(1927)
presented Ringelmann's results.
However,
Kunze
and
Schulhof (1925,
p.
171)
wrote,
"In
[ref-
erence] Ringelmann
has
already reported
on
experiments
thai
fall
in
this area.
The
research results that
are
reported there,
and
that have already been widely taken over
in the
German literature
. . .
."'
This suggests
that
Moede
failed
to
cite Ringelmann
be-
cause
no
citation
was
needed:
Moede's
readers were already
well
acquainted with
the
study.
Max
Ringelmann
was a
French agricultural engineer,
not a
German psychologist.
His
data were
not
collected
in the
1920s.
but
rather between
1882
and
1887.
As
Moede
was not
born
until
1888
(Marbe,
1938),
it is
clear that Ringelmann
was not
his
student. Furthermore, Ringelmann's
1913
research
report
con-
tains much information
that
Moede
(1927)
did not
mention.
1
"In
[reference] berichtet Ringelmann bereits
u'ber
Versuche,
die in
dieses Gebiet fallen.
Die
dort beschriebenen
und
vielfach
auch
schon
in
die
deutsche Literatur
iibernommenen
Ergebnisse dieser
Untersuchun-
gen.
. .
."
936
RINGELMANN (1913)
937
The
remainder
of
this article
will
be
devoted
to a
discussion
of
Ringelmann's work, particularly Ringelmann
(1913b).
We
describe this study
in
detail because
we
think
it is
important
and
we
suspect
that
it
would
be
difficult
for
most
interested
readers
to
obtain and/or read
the
original
report.2
Ringelmann
and His
Research
At
the
time
of his
1913
article, Ringelmann
was a
professor
of
agricultural engineering
at the
French National Institute
of
Agronomy
and
director
of the
Machine Testing Station.
He was
also
a
member
of the
exclusive French National Society
of
Agri-
culture.
Ringelmann's interest
lay in
determining
the
relative
efficiency
of
work furnished
by
horses,
oxen, men,
and
machines
in
various
agricultural applications.
His
work
on
horses
and
oxen began
in
1881
when
he was
still
a
student. Ringelmann (1907) reported
some studies
on
oxen
in an
article that also contains
a
long dis-
cussion
of
general factors relevant
to the
efficient
performance
of
animals. Ringelmann
(1913a)
is an
example
of his
tests
of
agricultural machinery.
In
each case,
his
concern seems
to
have
been
the
following: Which machine, method,
or
animal type
is
most
efficient
in
actual use?
He did not
draw sharp distinctions
between animate
and
inanimate sources
of
power,
and
within
animate sources hardly distinguished
men
from
animals. This
focus
on
efficient
performance predated
the
scientific
manage-
ment movement developed
by
Gilbreth
(1914),
Taylor
(1911),
Moede (1920),
and
others;
and
Ringelmann's work obviously
influenced
Moede.
Ringelmann
(1913b)
is a
report
of
research
on
human workers
carried
out
primarily between
1882
and
1887
at the
agricultural
school
of
Grand-Jouan. Thus this
may be
considered
the first
social psychological experiment, although
it was
published
well
after
Triplett
(1898).
Ironically,
his
subjects were enthusiastic
male student volunteers. Ringelmann
(1913b)
focused
on
max-
imum
performance
as a
function
of the
method
that
the
worker
used
to
push
or
pull
a
load horizontally; that
is, it was
human
factors
research. Comparison
of
individual
and
group
perfor-
mance
was
only
a
secondary interest
in the
report.
The
report consists
of a
discussion
of 8
preliminary
and 26
primary series
of
investigation, where each series
is
equivalent
to
what
we
would
now
call
an
experimental condition.
Before
discussing
the
preliminary series, Ringelmann pre-
sented normative data
on
human dimensions
and on
certain
aspects
of
human performance that
had
previously been gathered
by
Quetelet
and
Galton.
He
then gave comparable information
about
a
subset
of his
subjects
so
that
the
reader could compare
them
to the
normative
data.
The
preliminary tests consisted
of
eight series
in
which
the
subjects pulled horizontally
on a
rope.
The
rope
was 5 m
long.
(This
is
explicitly stated only
in the
discussion
of
Series
A, but
we
assume
that
it was
true
in all
eight series.)
The
track
was a
garden walk
of
well-beaten earth covered lightly with small
frag-
ments
of
schist. Here
and
later
it is
unclear whether Ringelmann
encouraged, discouraged,
or
simply ignored
the
possibility
of
competition among
the
subjects.
For a few
tasks Ringelmann
asked
the
subjects
to
maintain
a
maximum pull
for 4 to 5 s, but
his
primary variable, which
is the
measure
that
concerns
us, was
the
momentary maximum force
exerted.
This
force
was
measured
with
a
recording dynamometer.
The
manner
in
which
the dy-
namometer
was
anchored
and in
which
it was
attached
to the
rope being pulled
by the
subject
was not
specified.
Before
the
actual
experimental
tests,
Ringelmann carried
out
unspecified
preparatory tests
to
make sure that
his
subjects were
in the
proper
frame
of
mind.
Series
A
related maximum momentary
effort
to
subject weight
and to
maximum sustained
effort.
As we are
concentrating
on
the
social psychological rather than
on the
human factors aspects
of
this research, this series does
not
concern
us. The
position
used
by the
subjects
in
Series
B
through
H was
probably
that
shown
in
Figure
la.
(This
figure was
actually presented
in the
discussion
of the
primary series,
but the
description
of the
task
it
represents
is
exactly
the
same
in
primary
and
preliminary
series.)
In
Series
B,
each
of the 14
male subjects
pulled
alone.
In
Series
C, 7 of
these
14
subjects pulled together
as a
group,
and
in
Series
D the
remaining
7
subjects pulled
as a
second
group. Series
C and D
were
run
after
completion
of
Series
B,
permitting
a
fatigue
effect.
But
Ringelmann allowed
lh
hr
after
B
before
beginning
C,
probably precisely
to
reduce
fatigue.
(Here
and
later Ringelmann
was
careful
to
control errors
in his
con-
clusions that could arise through
differences
among subjects
or
differences
among days,
and as a
result
was
forced
to
accept
the
confound
of
fatigue.) Series
E, F, and G
were replications
of
Series
B, C, and D,
with
a
different
set of 14
male subjects.
Finally,
after
another rest this second
set of
subjects pulled
to-
gether
as a
14-man
group (Series
H). In all of the
group series,
the
subjects attempted
to
pull simultaneously
on
command,
an-
other example
of
Ringelmann's careful
attention
to
control.
Ringelmann
presented
all of the
individual
and
group
data,
some
of
which
are
repeated
in
Table
1.
If we
consider mean
individual
performance—rather
than summed
performance—
over
all 28
subjects,
the
mean
force
per
individual
was
85.3
kg
when
they pulled alone, 65.0
kg
when they pulled
in
7-man
groups,
and
61.4
kg
when they pulled
in the
14-man group.
A t
test indicates that
the
difference
between
the
individual
and 7-
man-group
conditions
was not
significant,
t(3)
-
3.015,
p =
.0570,
although
we
assume that this
was due to the
lack
of
power. Rin-
gelmann
(1913b)
presented
no
statistical tests,
but he did
verbally
compare
the
mean performances
in the
different
conditions. (The
first
article
on
Student's
t
appeared
in
1908 [Kirk, 1982,
p.
55];
there
is no
reason
for the
test
to
have been
in
common usage
in
1913.)
Ringelmann
(1913b)
explained
the
decrease
in
performance
with
increasing group size
as
being
due to
coordination loss: "the
lack
of
simultaneity
of
their
efforts"
(p.
9).3
Related
to
this issue,
he
presented
a
table "summarizing
a
great number
of findings"
(p.
9).4
This
table
is
repeated
in
Table
2 and
includes
the
data
that have been discussed
by
Moede (1927)
and
various other
authors during
the
past
7
decades. Note that total performance
asymptotes
at
size
7.
Ringelmann pointed
out
that
the
efficiencies
of
his
7-man groups ranged
from
.63 to
.83, whereas
the figure
given
in
Table
2 is
.56.
He
explained this
difference
by
stating
2
We
will
be
happy
to
send
a
photocopy
of
Ringelmann
(1913b)
at
cost
to
interested
readers.
3
"manque
de
simultaneite
de
leurs
efforts."
4
"resumant
un
grand
nombre
de
constatations."
938
DAVID
A.
KRAVITZ
AND
BARBARA
MARTIN
Figure
1.
Positions used
by
subjects
in
Ringelmann
(1913b).
that
"it is
necessary
to
take account
of the
fact
that,
in our
tests,
the
students
put all
their attention into acting simultaneously
on
command,
a
condition which
is not
encountered
in
practice
in
Table
1
Results
of
Individual
and
Group
Conditions
in
Preliminary Series
Individuals
01-07
08-14
15-21
22-28
15-28
Individual
efforts
(sum)
764.0
516.0
533.7
575.5
1
109.2
Group
effort
480
432
435.4
471.2
858.9
Ratio
of
group/individual
.628
.837
.815
.818
.774
Table
2
Relative
Performance
as a
Function
of
Group
Size
Work
usable
in
practice
(relative figures)
No. of
workers
1
2
3
4
5
6
7
8
Furnished
per
worker
1.00
0.93
0.85
0.77
0.70
0.63
0.56
0.49
Total
1.00
1.86
2.55
3.08
3.50
3.78
3.92
3.92
Note.
This
table
is a
translation
and
copy
of the
table
given
on
page
9 of
Ringelmann
(1913b).
work"
(p.
9).5
This implies that
the
data given
in
Table
2 are
drawn
from
a
number
of
uncontrolled
field
studies
or
observa-
tions. Note that Ringelmann
did not
specify
the
tasks
on
which
the
data
in
Table
2 are
based.
This
is
inconsistent
with
the
state-
ment
by
Moede
(1927)—and
thus
all
subsequent
writers—that
these
data
were
based
on
rope pulling. Also note that Moede
and
subsequent
authors have given only
the
data
from
group sizes
1,
2,
3, and 8.
Unfortunately Ringelmann
(1913b)
provided less
information
about these
data
than about
any
other data discussed
in
his
article.
It is
ironic that
it has
been precisely
these
data
that
have
had
such
a
profound impact
on
social psychology.
Ringelmann
explicitly stated that similar performance dec-
rements
had
been observed
in
draft
animals, although
he
gave
no
references.
He
then
gave
other examples
in
which
the
effect
had
been
observed,
as
with
men
turning cranks
to
operate
a
machine
or in
animals harnessed
to a
roundabout (mill).
He
also
mentioned that
men
have attempted
to
reduce
the
effect
in
some
tasks
by
singing
to
coordinate
their
efforts.
Ringelmann
did
discuss
one
example
of
reduced productivity
explained
by
motivation
loss—one
in
which prisoners provided
the
motive power
for a flour
mill.
He
reported that "the result
was
mediocre because
after
only
a
little while, each man, trusting
in
his
neighbor
to
furnish
the
desired
effort,
contented himself
by
merely
following
the
movement
of the
crank,
and
sometimes
even
let
himself
be
carried along
by it" (p.
10).6
But
Ringelmann's
common
interest
in
men, animals,
and
machines
led him to
focus
on
coordination loss rather than motivation loss
as an ex-
planation
for the
performance decrement, especially
as he
knew
that similar drops
in
performance occurred with inanimate
sources
of
power.
(He
cited supporting
data
drawn
from
obser-
vations
of
multicylinder
combustion engines,
in
which larger
en-
gines
produced less power
per
cylinder.)
Note.
These
are the
data
as
presented
by
Ringelmann
(1913b).
Effort
exerted
is
given
in
kilograms. Note that
the
number
of
significant
digits
given
varies,
and in two
cases
the
group/individual ratio
is
off
by
.001.
5
"il
faut
tenir compte que, dans
nos
essais,
les
Eleves
ont mis
toute
leur
attention
a
agir
simultanement
au
commandement,
condition
qui
ne
se
rencontre
jamais
pratiquement dans
les
travaux."
6
"le
resultat
a
etc
mediocre
parce
qu'au bout
de peu de
temps, chaque
homme,
se fiant sur son
voisin pour
fournir
Feffort
voulu,
se
contentait
de
suivre seulement
le
mouvement
de la
manivelle,
et
quelquefois
meme
de
se
laisser
entrainer
par
elle."
RINGELMANN
(1913)
939
Ringelmann
(1913b)
then turned
to the
primary
focus
of his
article.
These
26
series were
run
sequentially
on a
single
day in
1883,
and the
subjects were
20
male students
who
"very much
wanted
to
take part
in
these
tests"
(p.
11).7
Ringelmann explicitly
stated
that
he
used
the
same subjects
on the
same
day as a
form
of
experimental
control.8
The
object
of
these series
was
"finding
out
what
efforts
a man can
supply under
different
conditions
of
work"
(p.
11).9
The 26
series involved pulling
a
rope with
and
without
a
crosspiece,
pulling with
a
harness,
pushing
or
pulling
on
the
crossbar
or
shafts
of a
small hand cart (sometimes also
with
a
harness), pulling
and
pushing
a
wheelbarrow,
and
pushing
a
low
platform cart that resembled
a
mining cart.
In no
case
did
Ringelmann
explicitly state
how the
dynamometer
was
anchored
or how it was
attached
to the
rope
or
vehicle
that
the
subjects
were
attempting
to
move. Most
of
these
series
can
best
be
con-
sidered human factors research
and do not
concern
us
here.
In
each case, Ringelmann rank ordered
the
methods
of
accom-
plishing
a
task based
on
their
efficiency
in his
study.
The
four
series that dealt with individual-group
differences
are
discussed
shortly.
Before
presenting
his
results, Ringelmann
gave
the
heights
and
weights
of his
subjects
for
comparison with
the
normative
data previously given.
In
Series
8 and
13
the
subjects pushed
at a
crossbar that con-
nected
the
shafts
of a
two-wheeled
cart (see Figure
Ib).
A
weight
of
8.5
kg was
attached
to the
crossbar
to
represent
a
slightly
off-
balance load
on the
cart.
In
Series
8 the
subjects pushed alone,
and in
Series
13
they pushed
in
dyads. Ringelmann
did not ex-
plicitly
state
how the 2
subjects were placed,
but as the
cart (one
of
the fire
engines
of the
school)
was
normally moved
by two
men, they were probably side
by
side.
Ringelmann presented
all
of
the
data
of the
individuals
and
dyads,
and
these data
are re-
peated
in
Table
3. A t
test indicates that
the
difference
between
the two
series
is
significant,
t(9)
=
2.637,
p =
.0271.
The
mean
force
of the
dyads
(143.2
kg) was
less
than
the
mean
sum of the
dyad
members pushing alone
(160.8
kg).
In
Series
11
and 14, the
position shown
in
Figure
Ic
was
used.
Note that
the
only
difference
between Positions
Ib
and
Ic
is the
addition
of a
simple harness
in the
latter case.
The
same two-
man
cart
from
Series
8 and
13
was
used.
The
loading
of the
crossbar
was
18.5
kg
when
the
subjects pushed alone,
but
only
8.5
kg
when they pushed
in
pairs. Thus
it is not
strictly appro-
priate
to
compare these
two
conditions. Note, however, that
the
expected
effect
of
this confounding
is
directly opposed
to the
obtained superiority
of the
individual
condition.
In
addition,
crossbar loading seems
to
have made little
difference.
Ringelmann
used three
different
loadings with
the
position shown
in
Figure
Ib
(Series
8, 9, and
10),
and
loading
had no
significant
effect:
Loading
= 8.5 kg,
force
=
80.4
kg;
loading
=
12.5
kg,
force
=
79.7
kg;
loading
=
18.5
kg,
force
=
83.5
kg; all
differences
ns.
Although
Ringelmann discussed
the
difference
between Series
11
and
14
in the
text,
he did not
present
the
data together
in the
appendix
as he did for
Series
8 and
13.
Thus
he may
well
have
been
cognizant
of the
confounding
caused
by the
different
load-
ing.
In any
case,
the
results
of
Series
11
and
14
are
given
in
Table
4.
Once again
a t
test indicates
a
statistically
significant
difference,
t(9)
=
2.686,
p =
.0297.
The
mean
force
of the
dyads
(154.1
kg)
was
less than
the
mean
sum of the
dyad members pushing
as
individuals
(170.8 kg).
Table
3
Individual
Versus
Dyadic
Performance
Using
the
Position
of
Figure
Ib
Individual
Subject
1
60.0
85.2
97.2
72.0
84.0
54.0
78.0
78.0
78.0
72.0
performance
Subject
2
114.0
79.2
78.0
81.6
78.0
72.0
88.8
102.0
86.4
69.6
Sum
174.0
164.4
175.2
153.6
162.0
126.0
166.8
180.0
164.0"
141.6
Dyadic
performance
180.0
120.0
174.0
156.0
132.0
140.4
144.0
152.4
122.4
110.4
Note.
These data
are
adapted
from
Ringelmann
(1913b,
pp. 34 &
36).
Effort
exerted
is
given
in
kilograms.
*
Note that
the sum of the
individual performances
is
incorrect here.
Ringelmann
(1913b)
clearly
explained
these
results
in
terms
of
coordination
losses:
In
comparing
the
series
XIII
and
VIII,
as
also
the
series
XIV
and
XI, one
verifies
what
we
have shown above: when several sources
of
motive
force
work
simultaneously
on the
same thing,
the
utilizable
force
of
each
is
less, with
the
same
fatigue,
than
if the
sources
of
motive
power
function
separately.
We
have seen that this
is due to
the
lack
of
simultaneity
of the
muscular
contractions
of the
individ-
uals,
(p.
19)'°
Note
the
point about "with
the
same fatigue." Again
it is
clear
that Ringelmann
was
aware
of the
importance
of
experimental
control,
and he
tried
to
have
his
subjects attain
the
same level
of
fatigue
in
each series.
In
summary,
in
1913
Ringelmann reported
on
research
dealing
with
individual versus group performance that
he had
completed
3
decades previously,
and
explained
the
results
in
terms
of co-
ordination loss.
He was
aware
of the
possibility
of
motivation
loss
but did not
consider
it
likely
in his
research.
We
found
Rin-
gelmann's understanding
of the
importance
of
experimental
control
to be
remarkably
sophisticated
in
today's
terms.
It is
unfair
to
judge Ringelmann's research
by
standards that
developed
slowly
during
the
century
after
it was
completed.
We
have
pointed
out
places where
his
procedures
do not
meet current
standards, simply because
the
modern reader
will
find
this
in-
formation
relevant.
We
also pointed
out
that Ringelmann
left
unspecified
important information
about
his
procedures.
We
7
"Ont
bien
voulu
prendre part
a ces
experiences.
. .
."
8
"Pour
obtenir
des
chiffres
comparatifs,
il
nous
fallait
experimenter
dans
la
memejournee
sur un
grand nombre
de
memes
moteurs."
9
"de
chercher
les
efforts
que
1'homme
peut
fournir
dans
differentes
conditions
de
travail."
10
"En
comparant
les
series XIII
et
VIII,
ainsi
que les
series
XIV et
XI,
on
verifie
ce
que
nous avons expose plus
haul:
lorsque
plusieurs
moteurs travaillent
simultanement
sur la
meme piece,
1'effort
utilisable
de
chacun
est
plus
faible,
avec
la
meme
fatigue,
que si les
moteurs
fonc-
tionnaient
separement.
Nous avons
vu que
cela
est
du
au
manque
de
simultaneite
des
contractions
musculaires
des
individus."
940
DAVID
A.
KRAVITZ
AND
BARBARA MARTIN
Table
4
Individual
Versus
Dyadic
Performance
Using
the
Position
of
Figure
Ic
Individual
Subject
1
73.2
98.4
98.4
84.0
73.2
78.0
97.2
96.0
76.8
60.0
performance
Subject
2
121.2
81.6
79.2
75.6
96.0
90.0
73.2
94.8
81.6
79.2
Sum
194.4
180.0
177.6
159.6
169.2
168.0
170.4
190.8
158.4
139.2
Dyadic
performance
174.0
138.0
180.0
162.0
158.4
168.0
175.2
145.2
120.0
120.0
Note.
These
data
are
adapted
from
Ringelmann(
1913b,
pp.
35-36).
Effort
exerted
is
given
in
kilograms.
have
the
impression that
he did so
either because
he
assumed
that
his
readers would know
the
information without
its
being
given,
or
because
he
assumed
his
readers would trust
him to
have
used
the
proper procedures. Ringelmann's methodological
so-
phistication
is
consistent with
the
importance
of
agricultural
re-
search
in the
original development
of
research techniques (Kirk,
1982,
p. 9).
Note
that
when Ringelmann collected
his
data, Karl
Pearson
was in his mid 20s and R. A.
Fisher
had not yet
been
born.
Ringelmann's work
is
still relevant
to
contemporary theory
and
research.
The
connection
to
Steiner's
(1972)
conceptualiza-
tion
is
obvious,
and not
accidental.
As
Steiner pointed out, Rin-
gelmann's
(1913b)
research dealt with additive tasks
and
illus-
trates
the
principle that actual group productivity equals potential
productivity
minus motivation
and
coordination losses.
Ringelmann's research
is
most closely linked
to the
contem-
porary
research
on
social
loafing,
which
it
inspired.
An
interesting
difference
between some
of
Ringelmann's data
and
current social
loafing
results
is the
shape
of the
function
relating mean indi-
vidual
performance
to
group size. Current research
has
obtained
a
curvilinear relation (e.g.,
Ingham
et
al.,
1974; Latane
et
al,
1979),
as did
Ringelmann
in his
preliminary series (see Table
1).
But the
data
from
Ringelmann
that
have been cited over
the
years
exhibit
a
linear
relation
(see Table
2),
consistent
with
other
early
research
on
group performance (Kohler,
1927),
but
incon-
sistent with current
data
and
theories about group
effects
(Latane,
1981;
Mullen,
1983).
Unfortunately
we
know
so
little about
the
conditions under which
the
data
in
Table
2
were gathered that
it is
impossible
to use
them
to
evaluate contemporary theories.
For
example,
we
don't
know
how
many other people
(nonworkers)
may
have been present.
In
addition, competition
or any
other
factor
that
could
eliminate
social
loafing might well lead
to a
linear rather than
a
curvilinear
function.
Rather than
use
Rin-
gelmann's data
to
evaluate contemporary research
and
theories
in
detail,
we
prefer
to let
them stand
on
their own. They have
had a
long-term impact
on
social psychology,
and
that
suffices.
Ringelmann
(1913b)
provided
all of the
data
of all 26
primary
series
in an
appendix.
If we
recall
the
length
of
time this article
has
been lost
to
sight
and
that
the
only data
from
it
that have
been
cited
are
also those about which
he
provided
the
least
in-
formation,
there
is a
terrible irony
in his
expressed reason
for
giving
the raw
data:
We
thought
briefly
of
suppressing
all
these
tables
giving
the
individual
figures
ascertained
for
each
[subject];
we
abandoned
the