Motivation and Volition in the Course of Action

Abstract

Characteristics of the Action Perspective For Kurt Lewin (cf. Lewin, Dembo, Festinger,&Sears, 1944), there was never any doubt that motivational phenomena can only be properly understood and analyzed from an action perspective. Indeed, as he pointed out in support of this claim, processes of goal setting and goal striving are governed by distinct psychological principles. These insights went unheeded for several decades, however, probably for the simple reason that goal setting research based on the expectancy-value paradigm proved so successful (Festinger, 1942; Atkinson, 1957) and captured the full attention of motivation psychologists. It was not until the emergence of the psychology of goals (starting with Klinger's current concerns, 1977, and Wicklund's and Gollwitzer's self-definitional goals, 1982) and the psychology of action control (based on Kuhl's analysis of state vs. action orientation, 1983; see Chapter 12) that the processes and potential strategies of goal striving began to receive the attention that Kurt Lewin had already felt they deserved back in the 1940s (Oettingen&Gollwitzer, 2001). In contrast to the behaviorist approach, an action perspective on human behavior means extending the scope of analysis beyond simple stimulus-response bonds and the execution of learned habits. The concept of action is seen in opposition to such learned habits and automatic responses; it is restricted to those human behaviors that have what Max Weber (1921) termed “Sinn” (“meaning” or “sense”). In Weber's conceptualization, “action” is all human behavior that the actor deems to have “meaning.”

Full text

Motivation
and Volition
in the Course
of Action
A. Achtziger
and P
M,
Gollwitzer
r:,1:
:itt.l
:'!i::
:il
'::'
,'i
:]
11.1 Characteristics
of the Action Persoective
For Kr:rt Lewin (cf. Lewin, Dembo, Festilger, & Sears, Ig44),
therewas never any doubtthat motivationalphenomena can
onlybe properlyunderstood and analyzedftom an action pei-
spective. Indeed, as he pointed out in support of this clairn,
processes
of goal setting and goal strivin g are
governed by dis
- .
tinct psychoiogical principles. These insights went unheeded -
for several decades, however, probably for ttle simple rea--"--
son that goal setting research
based on the eq)ectancy-value
paradigm proved so successfir.L
(Festinger, 1942; Atkinson,
1957) and captured the firll attention of motivation psychol-
ogists. It was not until the emergence of the psychology of -
goals (starting with Klinger's cunent concerns, L977, and
Wicklund's and Gollwitzer's self-definitional goa1s,
1982)
and
the psychology of action control (based on Kuilt analysis of
state vs. action orientation,.1983; see Chapter 12) that rhe
processes and potential strategies of goal striving began to
receive the attention that Kurt Lewin had already felt they
deserved back in the I 940s
(Oettingen & Gollwitz er,.200
1).
in
contrast to the behaviorist approach, an action perspectivg
on human behavior meals ertending the scope of anaiysis
beyond simpie stimulus-response bonds and the execution
of leamed habits. The concept of action is seen in opposition
to suchlearnedhabits andautomatic responses;
itisrestricttid-
to those human behaviors that have what Max Weber Q92L) ,
termed "Sinn" ("mealing" or "sense"). InWeber's conceptu:-'
alization, "action" is allhumanbehavior that the actor deems
to have "meaning." Likewise, externa-l observers apply the
criterion of "meaaing" to determine whether or not another
person's behavior constitutes "action": are there discernibie'
"reasons" for that behar,ior?
DEFINITION
From
this perspective,
actions can be defineci
as all activities
directed
ioward
an
"intended
Eoal."
772
First publ. in: Motivation and action / ed. Jutta Heckhausen. Cambridge: Univ. Pr., 2008, pp. 272-295
Konstanzer Online-Publikations-System (KOPS)
URL: http://www.ub.uni-konstanz.de/kops/volltexte/2008/6393/
URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-63939

Motivation
anci
Voiiiion in the Course
of Action 273
The motivation psychoiogy of action focuses on ques- oithose goais (Lervin, I926b) - and, atthe same time, to incor-
dons of action control. These issues are imporrant because
- poratebothwithin asingle, unilvingftamework(Heckhausen,
asactionpsychoiogyresearchhasshournrepeatedly-asirong 1987a, 1989;
Hecltrausen & Gollwitzer, 1987). In a marrner
motivation to achieve a certain outcome or engage in a of speaking, the model examines the transition from wish-
cprtajn behavior does not normally suffi.ce
for that behavior ing to weighing in goal selection arrd from weighing to will-
+^ lrc imnlemcnteri illd t}1s
goal to be realized (Gollwitzer & ing in actual goal pursuit (Heckhausen, f987b). knportantl,v,
' Bargh, 1996;
Heckhausen, 1989; Kuhl, 1.983). In fact, success- it highlights the disitnctions between goal setring and goal
r various acdon control strategies (e.g.,
formulating "if-then" It was precisely thal kind of indiscriminate approach that
- pluns, resuntgggi_egypt:d actiols, slepppg up e ed confusion ür the history of motivation psychol-
:..''tfrefaceofdif0culties;cf.GoIlwitzer&na1phenomenabeingnegiecteci
' tions
11.5-11.7).
tl.2 The
Rubicon Model of Action Phases
tt--'= 'Th" fo"us of this section is on the course of action, which the
.,r-tir---
Rubicon model of action phases
understands to be a tempo-
, .rat,
horizontalpath startingwith aperson's desires
andending
. with the evaluation of the action outcomes achieved (Goll-
witzer,1990; Heckhausen, l987a; 1989; Heckhausen & Goll-
., t, witzer, 1987).
The Rubicon model seeks to provide answers to
the following questions:
'::, s Howdbpeopleselecttheirgoals?
i,, -: s Howdotheyplantheexecutionofthosegoals?
s How do they enact these plans?
i i ciflc goal?
,:'' -"" 0 The major
innovation of the Rubicon model was
to define clear
1 I boundaries between
motivational and volitional action
phases.
', i These boundaries
mark functional shifts between mindsets con-
ment. The
three most important boundaries
are
atthe transition from
,
' , ' the motivational
phase
before a decision
is made to the subsequent
. in itiation of action,
and finally atthe
üansition
from
the
action
phase
: back to the
motivational
(postactional)
evaluation
phase.
t!.2.7 Action Phases
Heckhausen's Rubicon modei of action phases
was inspired
bythe necessity
to distinguish two major issues in motivation
psychology- the selection ofa;tion goals and the realization
Figure
11.1 The Rubicon model of action
phases.
(Heckhausen
& Gollwiuer, 1987) Nlotivation
predecisional
for decades
Gleckhausen, 1987c, 1989; Kuhl, 1983, Goliwitzer,
f 990, 1991). Given that the processes
of goal setting and goal
striving serve a common function, however, ii was impoi-
tant that they should not be seen as isolated, independent
phenomena either. The Rubicon model gets around this dif-
ficulty by tracking the ernergence of a motivational tendency
over time - ftom fhe awakening of wishes to goal selection
and commitment, arrd finally goal deactivation. It seeks to
describe the emergence, maturation, and fading of motiva-
tion, dividing a course of action into four natural, consecutive
phases separated by ciear boundaries or transition points.
These
four action phases differ in terms of the tasks that have
tobe addressedbefore the individual canmove onto thenext
phase. The distinctions the model draws beilveen consecu-
tive action phases are thus both structural and functional in
nature.
According to the Rubicon model, a course of action
involves a phase of deliberating the positive and negative
potential consequences of various nonbinding wishes and
action alternatives (predecisional phase), a phase of plan-
ning concrete strategies for achieving the goal selected at the
end of the predecisional phase (preactional/postdecisional
phase), a phase of enacting these strategies (actional phase),
and finally a
phase of evaluating the action outcome (postac-
tional phase; Fig. 11.1; see also Fig. 1.3 in Chapter 1).
(B IIe four
phases
of the Rubicon model differ
in
terms
of the tasks
that have to be addressed
before the individual can
move 0n t0
' the next
phase.
Motivational
episodes are thus
broken down
into
"natural"
and seemingly
independent
phases.
Cntically, the
Rubic0n
model seeks to explain
both
goal
setting and
goal
striving.
Iniention Intention
Realization Deactivation
htention
Fomration Intention
Initiation
I
9l
El
JJ I
I
V
Volition
preactional Volition
actional iVlotivation
postactional

The first phase (predecisional phase) is characterized by ciuding) tendency to facilitate predictions of when the
deliberation.Anindividualfusthastodecider,vhichofhisor motivationaL task of deliberation wiil be completed. The
her manywishes to pursue. A person's motives are assurned more thoroughly arr individual has weighed the positive
to produce certail wishes. For example, a person rnith a and negative short- and Long-term consequences of engag-
strongachievementmotive (Chapter6; andaweakaffrliation ing or not engaging in a particular behavior, the closer he
motive (Chapter 7) is expected to experience more wishes or she comes to the belief of having exhausted all possi-
related to achievement than to affiliation. Yet because peo- ble routes of action. The chances of gaining new insights
p1e'sneedsandmotivesproducemorewishestharrcanpossi- into potential consequences
decrease,
and the facit ten-
blybeenacted,theyareforcedtochooseamongthem,com- dency, i.e., the tendency to decide on a certain wish or
mitting ttremselves
to certain selected goals.
To this end, they potential goal, increases apace. However, a decision is only
weigh the desirability ald feasibility of their many wishes. made when a previousiy stipulated levei of clarification
The objective of the predecisional phase is thus to decide - has been attained. Thjs level of clariication is positively
based on the criteria of feasibility (i.e., the e4pectanc-v that correlated with the personal importänce of the decision and
the actionwillsucceed) ald desirability (i.e.,
thevalue of the negativeiy correlated with the costs incurred in acquiring
expected action outcome) - which of their wishes they really information on potential consequences arld thinking that
want to pursue. hedividuals contemplating the feasibility of information through. As shown by Gollwitzea Heckhausen,.
a potential goal wiJl ask themselves questions such as the andRatajczak(1990),however,theprocessofdeliberationca11
following: be shortened by thinking in depth and detail about how one of
s Carr I obtain the desired outcomes by my ol.rm activity the alternatives under consideration mightbe transiatedinto
(action- outcomq exp ectancyf action. In an experimental design, these authors found that
E Is the situational context facilitating or iahibiting participants who anticipated a decision and pianned their
274
The Predecisional
Phase
(action-by-situation expectanry)?
The following questions are also crucial:
the desüed outcome? t
* Might favorable opportunities to pursue it arise?
the questions have no hard and fast arrswers
(e.g.,
it is rlifficult
to gauge outcome-consequence expectancies when the con-
sequerrces
irr question involve exiernal er,'aluation
orprogress
toward a superordinate goal), and irl most cases, there is not
even enough time to address a_li of the questions that might
be answered-
A. Achtziger
and P M.
Gollwitzer
The Rubicon model thus postulates the facit (i.e., con_
subsequent actions were quicker to make a decision.
However, even a wish with a hish resultant motivational
ity) do
es not necessarily gain access
to the executive. Rather, it _
first has to be traasformed into a concrete.goai.
This transfor-
Preactional Phase
It may not be possible for nevrr$ formed goa1
intentions to
be impiemented immediately. The individual maJ' first have
to compLete other activities, or wait for suitable oppofu-
nities to arise. Moreovet many goal intendons specify goal
states
(e.9.,
spending rnore time rvith one's famitSi graduating
il
I'
irr
a Do I have the necessary time arrd resources to pusue tendency (i.e.,
high expected value and hence high desirabil-
The desirabilityof apotential goa-l
or desired outcome is deter- mation is often described as crossing the Rubicon in ailusion------
mined by reflecting on questions such as the following: to Julius Caesar's crossing of the stream that once marked
* \ly'hat
are the short- and long-term consequences
of pw- the boundary between Italy and Cisatpine Gaui. By leading
suing this goal.? his army across the Rubicon and rnarching on Rome, Caesar
n Howpositive or negative might these consequences
be committed himself irrevocably to civil war. The transforma-
for me? tion of a wish into a goal invoives a shift from a fluid state of
I Howprobable is it that these consequences will occur? deliberating the value of a potential goai to a firm sense of
In addressing these questions, the individual weighs the cornmitmenttoitsenactment,i.e.,totheformationofa"goal
expectedvalueof awishorpotentialgoal;reflectsonitspos- intention' (see
Section 11.5 for a deflnition of "goal inten-
itive and negative, short- and long-term consequences; and tion'). Phenomenoiogically, it results in a feeling of determi-
assesses
the probability *rat achieving the desired outcome nation arrd certainty of taking the necessary action (Mchotte .
or potential goal will bring about these consequences. It is & Pnim, L9I0). The goal specified in the wish thus becomes ..'. --..
assumed that people do not contemplate their wishes and an end state to which the individual feels comrnitted.
potential goals
in isoladon, but see them in relation to other
wishes
and potentiai goals.
Awish associated
with a number e In the
predecisional
phase,
individuals
contemplate
the feasibility
of artractive consequences may thus sud.denly appear less :l:-tlt: wishes
as well
as the
desirability
of potential
action
out-
desirable
in the light of a superordinace wish. conversely, u tottt:.This process
of deliberation
culminates in commitment
to
wish may appear more feasible when contempl"t"t; ä; a.lecific goal
(goal
intention)
- in
crossing
the
"Rubicon"
between
cÖntexi-ofother-w{shesi-irarrwhen.seeninisoiu.io,,#csa!S.IetE-LSI9rye!is!-.ol9w!-s!intoabinding8oaI.-
tion of the deliberation process varies ftom case
to case. It is :j:t:]li::*'"t results
in a firm
sense of commitment
to translate
rare
for answers
to be found to all questions.
in fact. marrv of trroL
suqr
"
'to
action'

Moiivaiion
and
Voiülcn in ihe Course
of Action 275
ftom coliege,
etc.) thai cannot be achieve<i instarrtl,v.
Conse- phase.
They
choose sirategies
and formulate
plans (e.9.,
implemen-
quently,
peoplemaybe forcedto waitfcrfavorable opportu- iation
intentions; see also Section 11.5) that seem conducive
to
niries to arise before progressing
toward rhe iniended goal atiaining the aspired
goal
state.
state. Äccording to the Rubicon model, individuals in tb-is
waiting stage are in the second phase of a course of action - Actiona.lPhase
the volitional preactional (or postdecisional) phase. The The initiation of action designed to further the plals for-
term "volition' indicates that the motivational d.eliberation mulated in the pieactional phase signa-ls the transition to
of potential action goals has been terminated by crossing the actional phase. ha this phase, the individua-l's efforts are
the Rubicon, and that the individual is now cornmitted to focusedonpursuinggoal-directedactions andbringingthem
achieving a specific goal state. The task facing individuals in to a successful conclusion. These efforts are best facilitated
--trüspoStdecisionai(bütpreacilonältpFis,whichimp1iessteppingupeffort
how best to go about attaining the chosen goal. Thus, it is no in the face of difflcr:lties, and resuming goal-directed actions
'- longeraquestionofselectingdesirabieandfeasiblegoals,but aftereveryinterruption.Whetherornotanactionisexecuted
of deterrnining how to facilitate the achievement of the goals is deterrdned by the volitional strength of the goal intention.
chosen; e.9., by means of routine behaviors that are more Thelevelofvolitionalstrengthactsasakindofthresholdvalue
1..- .
or less automatic or newly acquired behaviors that require for effort exertion. Although this threshold is primarily deter-
conscious thought. Ideally, peopie in the preactionai phase mined by the strength of the motivational tendency, it may
-- should also develop plans specifying when, where, and how bespontaneouslyshiftedupwardwhensituationalditfrculties
goal-directed behavior is to be performed (Goliwitzer, 1999). are encountered. The primary source of increased volition is
, These plans are calied implementation intentions (Section the extra effort mobilized in response to situational difficul-
11.5).AccordingtotheRubiconmodelandthetheoryofinten- ties. Lr this phase, action implementation is guided by the
tional abtion control (Gollwitzer, 1993,
1999), implementation mental representation of the goal to which the individual has
intentions concerningthe initiation, execution, and termina- committed, which may well be outside his or her conscious
i
I
'l
lr
ti
ri
ti
il
tion of actions help people to overcome the dilfrculties that
cal be anticipated as they progress toward their goals.
Peo-
ple often find it particularly ditfrcult to get started, instead
engaging in extended procrastination and overlooking viable
opportudties to initiate goal-facilitating behavior. These are
''"- the problems to be overcome in the second phase of action.
HoW then, is action initiated when a rnore or less favor-
able opportunity arises? The co'ncept of the fiattendencywas
introduced to answer th-is
question. By crossing the Rubicon,
people commit themselves to enacting their chosen goals.
The stength of this commitrnent, which ttre Rubicon model
labels volitional strength, is a positive linear function of the
strength of the corresponding motivational tendency (i.e.,
the
desirability andfeasibility of the intended goal). The srength
of a goal intention's fiat tendency is the product of its voli-
tional strength (i.e., the commitment to püsuing the goal
state) and of the suitabüity of the situation for its initiation.
The suitability of a situation is not determined in isolation,
but relative to other opportunities that might occur in the
future (longitudinal competition). The fiat tendencies of an
individual's other goal intentions also have to be considered.
Itwould be mrong to assume that peopie always take action
to promote a goal with a high fiat tendency. Many situations
are
conducive to awhole range of intentions, not all ofwhich
can be implemented at once (cross-sectional competition).
Ia this case,
the goal intention with the highest fiat tendency
gains access to the executive, and actions seeking to accom-
plish it are irritiated.
E ln
the preactional phase,
individuals contemplate
how bestto
pur
sue the
goal
to
whioh
they committed atthe
end ofthe
predecisional
awaleness.
O In the actional
phase,
individuals seek to enact the
plans
made in
the
preactional phase
with
the aim
of enactingthe
goal
formulated at
the end of the
predecisional phase.
These
efforts
are best facilitated
by steadfast
pursuit
ofthe goal
and by stepping up the effort
exeded
in the
face of difficulties.
Postactional Phase
Thetralsitionto the fourth and final action phase,
the postac-
tional phase occurs once the goal-oriented actions have
been completed. The task to be addressed at this stage is
again a motivational one. Specifically, individuals measure
the results of their actions against the goai set at the end
of the predecisional phase, asking questions such as the foi-
lowing:
s Howwell have I succeeded in achieving my goal?
s Did the action result in the positive consequences
anticipated?
* Can I now consider my action intention completed?
w If the goal was not attained, do I need to keep working
toward it, perhaps by other means?
Individuals in the postactional phase thus look back at the
action outcome attained arld, at the same time, cast their
thoughts forward to future action. If the action outcome cor-
responds with the aspired goal state, the underlying goal is
deactivated. In many cases, shortcomings in the predeci-
si.onal
deliberation of an action's positive andnegative, short-
and long-term consequences may become apparent at this
point. It may, for example, emerge that the desirabilitv of ttre
goal was overrated because certain outcome erpectancies

il
tl
j
I
zto
were overestimated or overiooked. Of course,.not a.ll
comparisons between intended and achieved outcomes
result in the deactivation of the goal: the action outcome
may deviate from the intention in quaütative or quantita-
tive terms. The goal may then be adjusted to the outcome
by lowering ttre level of aspiration. Alternativeli4 individuals
may choose to retain the original goal despite the rinsatisfac-
tory outcome, and renew their attempts to achieve it. Deac-
tivation of a goal that has not been achieved seems to be
facilitated by the prospect of a new goai taking its place. For
example, Beckmann (1994) showed that pa-rticipants could
only detach mentaliy from a poor score on an intelligence
test if they expected a new test to be admilistered in the next
round. Participants who didnothave this prospectkept ttrink-
ing about the poor test result, i.e., engaged in self-evaluative
rumination.
O In the
postactional phase,
individuals evaluate
the action outcome
achieved.
lf they are satisfied with
the outcome, they deactivate
the goal
set at the end of the
predecisional
phase.
lf they are not
satisfied with
the outcome, they either lower
the level of aspiration
and
deactivate the
goal,
or retain
the original level of aspiration and
increase their
efforts t0 achieve the desired
goal.
LL.2.2 Motivational vs. Volitional Action Phases
Kurt Lewin (f926b) and Narziss Ach (1935)
understood voli-
tion to be the form of motivation involved in goal striving,
and goal striving tö encompass all processes of motivational
regulation that serve the pursuit of existing goals.
Thus, voli-
tion concerns the translation of existing goals into action and,
specificaliy, the regulation of these processes. Motivation,
in contrast, concerns the motivational processes
involved in
goal setting. The focus here is onwhich goals apersonwishes
to pursue. People who have to decide between different goals
are assumed to weigh the erpected value and attainability of
the available options very carefully (Gollwitzer, 1990).
Classic
motivation theories rely on this narrow defiaition of moti-
vation, assuming the motivation to act to be determined by
both the desirability and perceived feasibility of the aspüed
goal. If someone does not believe him. or herself capable of
doing what is needed to attajn a goal, or does not consider a
goal particularly desirable, he or she will not be motivated to
do all she carl to pursue it.
In the early 1980s, KuhI reestablished the distinction
between motivation and volition, and drew a clear line
between modern volition research and the more philo-
sophical debate on "foee
*ill" Kr.rru, 1983; see also Chap-
ter 121. KuhI was the first modern motivation researcher to
drar,rr attention to the contrasting functions and character-
istics of "choice motivatiorl' ald "controi motivation," and
süongi-y adiT6cated that a cüstincüon be made betr,treen rnoti-
vational and volitional issues in research (Kuhl, 1983, 1984,
1987).
.A.
Achtziger
and
P. M. Gollwitzer
SUMMARY
Motivation concerns the processes
and phenomena involved
in goal setring, i.e., the selection of goa-ls
on the basis of their
desirability a.lrd
feasibility. Nlotivational processes dominate
in the predecisional and postactional phases ofthe Rubicon
model. Volitional processes and phenomena, on the other
hand, concern the tanslation of these goals into action, Voli.
tional processes dominate in the preactional and actional
phase.
L1.3 Action Phases and Mindsets:
How Can
Psychological Processes
Be lncorporated
in
an ldealized.
Structural Model?
The Rubicon model of action phases implies that goal-
directedbehavior canbe broken downinto aseries of consec-
utive phases.
The prernise for this kind of research approachis
that the phases
identifi ed descnbe qualitativeiy different psy-
chological phenomena that correspond to the different func-
rtrons of each action phase. The Rubicon model is thus both
structural and fu nctional in nature ft{eckhausen, 1 9B7a). The
main fi:nctions of the four action phases
identifi.ed are listed
in the following ovewiew.
Each of these functions is assumed to be associated with a
different mindset; i.e., a form of information processing that
is appropriate to the action phase at hand. Based on the ter-
minology of the wtirzburg school (Chapter 2), the concept of
mindset refers to the states of mind that are associated
with
the assumption ald execution of specific tasks fMarbe, 1915;
Heckhausen, 1989).
DEFINITION
The
term
"mindsef'
describes
a certain kind of
cognitive
orientation
that
facilitates
performance
of the task to be
addressed
in
eacn
action ohase. i
Mindset research is based on the idea that distinct tasks
have l
to b e solved in Cagb lh4Cg glthe-Bub:tlln ms-de-l {Goii}4!Zg-",j
1990).
In theü comprehensive research program, Gollwitzer and
colleagues (see
the.overview il Gollwitzer, 1991) have found
evidence for qualitative differences betrveen action phases,
and they have shown that task-congruent mindsets deter-
mine the content arrd form of information processing in
each acrion phase. Within the research paradigm founded by
Gollwitzer, the characteristic task demands of the deli.bera-
tion, implementation, action, and evaluation phases are
fust
,Lil
,iill

-
' 1y1e1ir:ation
and
Volition in the Course
of Action
-
277
aneJ4ed,,
allowing hypotheses about phase-specific differ- prevents iis postponement. To
ttris end, ttrere
is cognitive tun-
I .rr"", in informatiön processing to tien be derived and s-vs- ing tor,t'ard
iäformation relevant to where, when, and hou' to
' Th"r" h14,otfreses,
which are
outlinedbelow, concern the cog- in the sense
that people shouid concentrate on inJormation
nitive orientations that are functional for addressing phase- relevant to task performance, a:ed
ignore incidental, less rel-
j., ' ,f ".in. turtr. tt is assumed that each
phase is associatetiwith evarrt information. Thus, attentioil is focused on a specified
, acertainmindset (i.e.,
withthe activation of speciflc cognitive opporhrnityto act, andthe individualis shieldedfromthe dis-
r, procedures) that facilitates performance of the task at hand. tracti.ons
of competing goals, etc. This shielding ftmction also
-.: aDpliestoi::-formationaboutthedesirabilityandfeasibilityof
=**Deliberative-Miqdset --- - ---------.j;oaiselectedarthqend-o-f-the predecisionalphase, rvhich
. The deliberative mindset is associated
with the predecisional is irälevant to the initiation of goal-directed behavior and is,
.,.. . phaseandthuswiththetaskof goalsetting.\Mhatkindof cog- in fact. disrractins.
nitirre orientadon characterizes this mindset? How do peo-
, ple in tiris mi:rcisei altend to and process information? i:::.di- S Individuals
in the implennental
mindset
are
partlcularly
receptive
to
,ria,ra. in the predecisional phase are faced with the task of information
relating
to the initiation of goal-directed
behavior.
At
" "'äeciding which of their wishes to iranslate into action; they tie same
time,
there
is closed-mindedness
in the sense
that only
.=_-._..have
to weigh ttre relative desirability and feasibilitv of theü information
that
wiil help
to promote
the chosen
goal
js processed.
... wishes in order to select comparatively atftactive and attain- Actional Mindset
': able action goa-ls.
Solving this task requires individuals in the if* u.aio"uf mindset is associated with the actional phase,
. , deliberativemindsettobe primarilyconcernedwith informa_ *re task of which can be described as acting toward the goal
;y,:Tää::f:,x'J:-Ji,::ilffi1T:ffiH#:lT,T"1
;:ir"T,ä3.":ll*;",ffi*t;il:ffi.ä:;l;ff,.J:H:
tially as
possible; it is important that negative consequences äence characteristics of what Csikszentmihalyi (1925)
called
should not be overlooked. Likewise, feasibili-ty assessments ,,flowerperience,,andWicklund (f986) labeled "dynamic ori-
. should be as accurate as possible, i.e., neither overiy opti- entation.,, Specifically,
ind.ividuals in this mindset no longer
. .misticnorunnecessarilypessimistic.Onlyifexpectanciesald -^d^^- ^- ,r.'- ^,,"r'r." nf the onal tn hc achievea- or on tt
incenrives
are assessed
in an objective
.'u;;;1ä3# äffi::Til, :,:fJi:';.#::"#iiJ""i:fr"#H'":T
ä'i:
can the predecisional task of selecting a comparatively destr sider alternative.strategies, neither dp they form implemen-
able and attainable goal be accomplished successfullv. tation intentions or action plans specifyingwhen, where, and
:'*"" "'*":^'." "^
*:*"*''*-
ImplementalMindset how to act. Rather, they are totally absorbed in the actions
The implemental mindset is associated
wifh the preactional being executed.Accordingiy, theyonlyattendto those aspects
phase; its task is to prepare for goal strivi:i;; e.g., by under- of the seif and the environment that sustain the course of
iaking efforts to initiate appropriate actions. The concrete action, and ignore any potentially disruptive aspects (e'g''
approachtaken depends on the type of goal
set.If, upon cross- self-reflective thoughts, competing goals, or distracting envi-
ingthe nubicon, the goalwas furnishedwithimplementation ronmentalstimuli).The actionalmindsetisthereforehYpoth-
intentions (Sections 11.5-f 1.7) specifiTingwhen, where, and esüed to be one of cioSed-mürdedness to any in-formation
how actions are to be irritiated, all that remains to be done that might trigger reevaluation of the goal selected
at the end
is to wait for an appropriate opportunity to arise (i.e., the of the predecisional phase, reevaluation of the implementa-
"when'r and "where" specified.
in the implementation inten- tion süategy chosen, or anyform of self-evaluation (e.g',
"Can
. tion). As soon as a potentially viable oppcrtr:nity arises,
the I be proud of myperformance thus far?",
"Do Ihave the neces-
individual compares itwith the oppoftunity defined as favor- saryskills to achieve the goal?").
Rather, the actional mindset
able in the implementation intention. If a match is ascer- should evidence cognitive tudng toward internal and erter-
tained, goal-directed behavior is initiated immediately. The nal cues that guide the course of action toward goal attain-
same holds for goals that do not require implementation ment. This information shouid be as accurate as possible;
intentions because they are habitually initiated in a specific its evaluation should not be positively biased. The actional
wav F{era too. the individual simplv has to wait for a suitable mindset should emerge whenever people move effectively
opportunityto arise, and then initiate goat-directed behavior. toward goal attainment-
If neither implementalion intentions nor habits that might
f^^il:+^+^ -^^l ^^Li,
rauurdLs
6u@
a.,,,evement are
in place, corresponding action Evaluative Mindset
plans first have to be formr:lated. Solving these
tasks requires The evaluati.ve
mindset is associated with the postactional
individuals to. be receptive to alld process information thal phase, r,v-hen
t]le task is to evaluate the action outcome and
facilitates the initiation of goal-oriented behavior, and that its consequences in order to establish whether goal pursuit
,i
l,i.
l,i,,
ri
iii
,iilr
itl
ifl
ii.,ll
'rlil:ri'
,iijlii,i
1; 1;,r,
llili,
ilir,,ilil
,llll,i
il,,lili.;
li' lltl
1r,|l

i
278
haS led to the intended outcome and desired consequences.
Säiving this task requires individuals to be primarily con-
cernedwjth the quality of the action outöome ald the actual
desirability of its conseo;rences. In other words, individu-
als in the evaluative action phase compare what has been
achieved (outcomes) and obtained (consequences)
withwhat
was originally expected or intended. Accurate assessments of
the quality of the outcome and objective, impartial views of
the desüability ofits consequences are thus required. Accord-
ingly, the evaluative mindset should evidence ttre following
characterisrics:
r cognitive tuning toward information relevant to assess-
ing the quality of the achieved outcome and the desirabil-
ity of its consequences,
{ accurate and impartial processing of that information,
and
r a comparative orientation: the intended outcome and
its expected consequences are compared with the actual
outcome and its conseouences.
SUMMARY
The action phases of the Rubicon model are characterized
by four distinct goal-oriented behaviors: deliberating, plan-
ning, acting, and evaluating. Because each phase involves a
distinct task, each is associated with a mindset condücive to
perforrning that task. The cognitive characteristics of each
mindset can be specified by critically analyzing the demands
of the tasks addressed in each action phase. For example, the
deliberative rnindset is characterized by open-mindedness,
and by the objective processing of all available information
on the potential consequences of an action outcome (desir-
ability) and the viabilityof the individual's wishes (feasibili$.
The implemental mindset is characterized by cognitive tun-
ing toward information that facilitates the injtiation of goal-
oriented behavior, and that prevents its postponement. The
actionalmindsetfocuses attention on those aspects of the self
and the environment that sustain the course of action; arry
potentially disruptive aspects (e.9., self-reflective thoughts,
competing goals, or distracting environmental srimuli) are
ignored. Finally, in the evaluative mindset, there is cognitive
tuning toward information t}lat heips to assess
the quality of
the achieved outcome as objectively and accurately as pos-
sible. To this end, the individual compares what has actually
been achieved (action outcome) and obtained (consequences
of that outcome) withthe intended or expected
outcomes and
consequences.
11.4 Contrasting Effects of the Deliberative
and lmDlemental Mindsets
Having discussed the theoreücal backgror:nd to the four
mindsets inSection I l.3,we nowpresent empiricalfindings in
A. Achtziger
and
P. M. Gollwitzer
support of the hypotheses formu-lated about the deliberative
ald implemental mindsets. We focus on these trr"ro mindsets
simply because research has yet to examine the actional and
evaluative mindsets, or to test the hl4lotl-ieses derived about
information processing and cognitive orientations in these
läst two phases of the Rubicon model. We begin by describ-
ing how the deliberative and implemental mindsets can be
induced ero erimenta.llv.
I
I
I
I
I
J
,l
-t
Experimental Design Comparing Deliberative
and lmplemental Mindsets
x induciion
of the
Deliberiitive
Mindsei:
Participants are asked
to
identify
a
personal
concern
(problem)
that they
are
currently deliberating, without
yet
having decided
whether
to make a change
(i.e.,
to act) or to let things take
their course
(i.e.,
to remain
passive).
For
example, they may oe
contem
plating
whether it makes
more sense
to switch
majors
or
to
stick
with their current
one. ParticiDants
are then asked to
list
the potential
short-term and long-term,
pdsitive
and negative
consequences of making or failing
to make a change
decision,
and to estimate the
probability
ofthose consequences actually
occurring
(cf.
Gollwiüer & K-,rney,
1989, Study 2; Gollwiüer &
Bayeri 1999).
x Induction of the lmplemental
Mindset:
Panicipants are asked
to identify a goal (project)
that they
intend to accomplish within
the nextthree monihs; e.g., apply-
ing for a grant
to study abroad. They
then list five
steps that
have
to be taken to accomplish that
goal,
and finally
write down
concrete
plans
on when, where,
and how
to take each
step.
They
thus specify the exact time,
place,
and manner in
which each
step toward
realizing
the goal is to be taken (cf. Gollwiter
&
Kinney,
1989, Study 2; Gollwitzer
& Bayer,
1999).
r Alternative Means
of lnduction:
Puca
(2001)
and Puca
and'Schmalt
(2001)
induced the
delib-
erative mindset
by
interrupting
the decision-making
processes
of
participants
who
were
poised
to make
a decision, such
that
they continued to deliberate
on the alternatives
available.
They
induced the implemental mindset
by allowing
participants to
make a decision
(between
alternatives).
Participants
were then
administered
tasks that had nothing
to do with the decision
task, but served
to investigate
the effects of the respective
mindset on different
cognitive
processes.
Gollwitzer and
Kinney
{a989tStudyl) had
alieady-taken a simiiar appi'oach,
induc-
ing an implemental
or a deliberative mindset
by presenting
participants
with
a decision task. Specifically,
the implemental
mindset was
induced by asking
participants
to decide
on a
certain
sequence of trialb before
the dependent variables
were
assessed. The deliberative
mindset was induced
by intenupt-
ing
panicipants
shortly before they made
a final decision
on
a
sequence 0f trials.

-
Motivation
and Voliiicn in ihe
Couise of
Action
Lj-.4.1-
Cc
gn
itive Tu n i
n
g Toward
Task-Con
gruent
lnformation
The implemental mindset is assumed
to promote goal attain-
ment by helping people to overcome the classic probiems
of goa-l
suiving; e.g., doubting the attractiveness and hence
the desüability of the goai being pursued, the practicability
of goal-directed strategies, or the feasibiJity of the aspired
project. Empirical data support these assumptions, show-
-ing that the implemental s3 nqsg!_elrckeq Jggiltiyg h4u
' toward information related to goal attainment. Parricipants
....=-
in aD implemental mindset repoft more thoughts relating
,. to the execution of an aspired project (i.e., "implemental"
thoughts of the tlpe "I'11
start with X and then move on to
: Y') than participants in a deliberative mindset (who tend
":-t-leport "deliberative" thoughts of the type "If I do this,
:_it will have positive/negative consequences, if I dont, then
X Y or Z is likely to happen"; cf. Heckhausen & Gollwitzer,
1987;
Taylor & Gollwitzer, 1995, Study 3; Puca & Schmatt,
200r1.
. ' (1990)
induced either an implemental or a deliberative mind-
. set
using the pro cedure described in Section I 1.4.
Participants
lvere
then presented with three fairy tales that were cut short
- at
a certain point in the plo t. Inwhat was ostensibly a creativity
' t test,
theywere asked
to continue the story.
Participants in the
, . imilemental mindset were more likely to have the protago-
-,-==nists
of tleir stories plan howto carry out a chosen goal than
': ' were participants in the deliberative mindset. In a second
;"::=study,
participantsin an implemental or adeliberative mind-
' : setwere shown a series
of slides, eachpresentingaa image of a
. person alongwith sentences reporting that person's
thoughts
: ort the pros and cons of a specific course of action and plans
to put it into practice. After viewing the slides and working
on a short distacter task, parricipants were administered a
' cued-recall test of the information presented. Implemental
, participants were better able to recall information relating to
.' the when, w-here, arrd how of goal achievement tharr infor-
mation relating to the pros and cons of a change decision.
. The recall performance of deliberative participants showed
. the reverse pattern.
SUMMARY
The thoughts of individuals in the deliberative mindset are
more attuned to action alternatives than to strategies of
goal achievement; Likewise, individuals in the deliberative
rnindset recall information associated with the delibera-
tion of alternatives betier thal information pertaining to
the accomplishment of goal-directed actions. Individuals in
the implemental mindset devote more thought to ptanning
goal-directed behavior than to contemplating action alter-
natives, and find it easier to recall bJormation relating to
the piarrning of actions than to the contemplation of action
al.ternatives.
t1.4.2 Processing of Relevant and lrelevant
Information
Gollwitzer and Bayer (1999) report that the implemental
mindset leads to "closed-mindedness," to the extent that
individuals in this mindset do not allow themselves to be
djstacted by irrelevant inforrnation, but focus exclusively
to the accomplishment of thöir
goal. This find:ng is substantiated by the empirical data
of Heckhausen arrd Goilwitzer (1987, Study 2), who found
that impiemental participanrs have shorter noun spans (a
good indicator ofreduced cognitive processing speed;
Demp-
ster, 1985) than do deliberative participants. A set of stud-
ies using a modified Müller-Lyer task confirmed that imFle-
mental participants' attention is more centrally focused
thari that of deliberative participants, and that people in a
d.eliberative mindset are more fikely to attend to incidenta.L
informadon than people in an implemental mindset (Goll-
witzer & Bayer, 1999). In a decision experiment that requted
respondents to choose between the Rubicon model and Fes-
tinger's dissonance theory Beckmann and GoUwitzer (1987)
showed that information relevant to the ongoing action is
processed preferentially in the impiementa-l mindset, even
when it is not in line with the decisions that have been
qrade.
SUMMARY
Empirical research has shoum that people in the deliberative
mjndset are more likely to be distracted by information that
is irrelevant to goal attainment. This finding is in line with
the observation that individuals in the deliberative mindset
attend to incidental information. The reverse holds for the
implemental mindset. Here, processing is attuned to infor-
mation of direct relevance to goal attainment, and attenJion
is centrallyfocused.
i
1-L.4.3 Biased Processing of Information Relating
to Goal Feasibility and Desirability
Mindset research assumes that the implemental mindset fos-
ters apositive evaluationof the chosengoal (i..e.,
itshighdesir-
ability) and, at the same time, promotes a highly optimistic
assessment
of its practicabilityand attainability. The deliber-
ative mindset, by contrast, is assumed to generate objective
assessments of the positive and negative consequences of
goal attainment, ald a more careful evaluation of the prob-
abilitv of achieving the goal. Various studies (cf. Gollwitzea
1990) have been conducted to test these hlpotheses; one of
the classic studies is described on the next page.

280
Classic Study on "lllusionary Optimism" in the
lmplerhental Mindset
ln what is known
as the "contingency
learning
task" (Alloy
&
Abramson,
1979),
participants
perform
a series
of
trials 0n
a single-
stimulus
apparatus.
Their task
is to determine
to what
degree
they
can influence
the onset
of a target
Iight
(i.e.,
the intended
outcome)
by
choosing
to press
or not press
a button.
In other
words,
par-
ticipants
are told
that alternative
actions
(pressing
the button/not
pressing
the button)
can lead
to
the outcome
"target
light
onset.,
What
they do not know
is that
target light
onset
is in
fact entirely
independent
ofwhether
or notthey
press
the button;
it
is determined
by a random
generator.
The
experimenter
can vary
the apparent
degree
of control by
changing
the
setting of
the random
generator,
thus manipulating
the frequency
oftarget Iight
onset
associated
with
each
ofthe action
alternatives
(i.e., pressing
or not
pressing
the response
button). An
extensive
body
of research.
using
this
contingency
learning
task
(cf.
Alloy
& Abramson,
1988) has
shown
that nondepressed
panici-
$ants
believe
themselvesto
have
control
overtarget
light
onset
when
this desired
outcome
occurs
frequently (e.g.,
when
the
target light
comes
on in 75%
ofpressingandT5o/o
ofnonpressing
responses)
as
compared
to infrequently
(e.g.,
when
the target light
comes
on
in
25%
of
pressin
gand
25Vo of nonpressing
responses).
Given that
tArget light
onset is in fact
noncontingent
to participants,
actions,
but
governed
by a random
generator,
these findings
of inaccdrate,
optimistic
judgments
of control
are remarkable.
Gollwiüer
and Kinney
(1989) assumed
that this unrealistic
illusion
of control
over
target light onset
would
be less
pro-
nounced
in deliberative
mindset
participants
than in the imple-
mental
mindset
participants.
The authors
assumed
that peo-
ple in the implemental
mindset
tend to see themselves
and
their abilities in a much more positive
light than do peo-
ple in the deliberative
mindset (Section
11.4). They
there-
fore modified
the contingency
learning
task by adding
a sec-
ond apparatus
and asking
participants
to work
on 5 sets
of
20 trials.
A single
trial consisted
of the choice
to press
or not
press
the response
button
followed
by
task light
onset
or non-onset.
A
deliberative
mindset
was induced
by teiling
participants
that their
objective
in the first
part
of the experiment
was
to decide
which
of the
two available
apparatuses
to work
on during
the second
part
of
the experiment.
Deliberative
participants
were
encouraged
t0 try out both
apparatuses
before
the experiment
proper
began
to ensure
an informed
decision.
The
implemental
mindset
was
induced
by asking
participants
to specify
which
apparatus
üey
wouidüse--in-each
iriai beiore
startinE
ing
this decision,
they were
instructed
to try
to produce
as many
.light onsets
as
possible,
whether
by
pressing
or not
pressing
the
response
button.
The
participants
were
thus instructed
to ,,find
out" for
themselves
whether
pressing.or
not
pressing
the button
gave
them
more
"control"
over target
light
onset.
0f course, the
experimenter
knew
that target
light
onsei was
in fact
governeo
0y a
A. Achtziger
and P M.
Golhviüer
random
generator,
and entirely
independent
of pafticipants,
actions.
Besides
the
two
mindsets,
a "iarget
light
onset'condi-
tion was
implemented:
* eitherthe
"high
frequency
oftarget
light
onset,,condition,
in
which
the
target light
comes
on in 75o/o
of
pressing
and
75%
of
nonpresstng
resp0nses
s or the "low
frequency
of target light
onset,
condition,
in
which
the target light
comes
on in 25%
of
pressing
and 25%
of
nonpressing
responses.
Accordingly,
both
apparatuses
presented
either
noncontingent
fre-
quent
or noncontingent
infrequent
onset
of the
target liSht.
When
target
light
onset was frequent
and
thus seemed
t0 be
',contingent,,
on
pafticipants'
actions
(pressing/not
pressing
the response
bui_
ton), implemental
mindset
participants
reported
inaccurately
high
judgments
of the degree
of control
they exerted
over target
light
onset
(illusionary
optimism),
whereas
deliberative
mindset
rated
their
level
of control
to be much
lower.
The
deliberaüve
mindset
participants
evidently
recognized
that
high frequency
of an event
was not necessarily
a valid
indicator
of their
own influence
over
it. The
deliberative
mindset
thus seems
to prevent
people
from
adopting unrealistically
optimistic
beliefs
about how
much
influ-
ence they
have over
uncontrollable
events.
When,
on the other
hand, target
light
onset was
infrequent
and
thus seemingly
n0n-
contingent,
both mindset
goups
showed
rather
modest
control
judgments.
lhis finding
indicates
that people
in an implemental
mindset
can adapt
to external
constralnts
if necessary.
lf environ-
mentalfeedbacktellsthem
otherwise
(e.g.ia,righ
rate.of
"non-hits,,
in the button-press
task),
they
do not
cling
blindly
to a belief of
being in
control
over
target
outcomes,
but
abandon
this illusion
of
control.
On the subject of "illusionaryoptimi.srn'in the implemen-
tal mindset, Gagnd and Lydon (2001a) report that individu-
als in an impiemental mindset see the future of their cur-
rent romantic relationship in a more optimistic light than do
individua-ls in a deliberative mindset. Likewise, puca (2001,
Studies 1 and 2) estabüshed
that the implemenral mindset is
associated
with an optimistic approach to the choice of test
materials of varying difficulty (Study f) and the predictiön
of future task performance (Study 2). Relative to delibera-
tive partici.pants, implemental participants opted for more
difficult tasks and were more optimistic about their chances
of success.
Finaliy, Harmon-Jones ald Harmon-Jones (2002,
Study 2) discerned differences between the deliberative ald
implemental mindsets in terms of how information on the
. and nonahosen
ätemätlves is pro-
cessed.
Dissonance research discovered that, once a choice
has been made, the chosen option is seen in a much more
positive light than the nonchosen option. Harmon-Jones and
Ha;mon-Iones observed that inducijon of an implemental
mindset increases
this effect,
whereas induction of a deliber-
ative mindset reduces it.
iinii
;;,llii.
iirii
i,ilti
irliif'
.ir
L'
l

Motivation
and Voliticn
in
the Ccurse cf Action
SUMMARY
Relative
to the deliberative mindset, the implementai mirld-
set is associated with i.ncreased cptimism about ihe degree
of personal contol over iniended action outcomes and with
a preference for rljfficult tasks. MoreoveS the impiemental
mindset is associated with higher estimations of the proba-
biLity of success
tharr the deliberative mindset.
11.4.4 Mindsets and Self-Evaluation
__, _-Pdi!_q1qgye__q+4=1!qp_131g9ntal
mr$elq_!ryS also been
. -- shovm to affect the way people see themselves. Experimen-
tal findings show ttrat people in a deliberative rniadset score
much lower on the Rosenberg Self-Fsteem
Sca-le
(Rosenberg,
. 1965)
than do people in an implemental mindset. Likewise,
, students judge themselves to be more creative, intelJigent,
:=' "-
-"1opu-lar, etc., when an implemental mindset is induced than
when a deliberative mindsetis induced (Tayior & Gollwitzer,
1995).
leduction of an implemental mindset evidentlyboosts
peopie's
belief in themseives and their abilities. \A4rere
seH-
ratings of susceptibilityto variousrisks are concemed, more-
over,
findings show that people in an implemental mindset
considerthemselves less likelyto fallvictim to various strokes
of fate (e.g.,
being involved in a plane crash or developing
diabetes)
than do people in a deliberative mindset. Tabie 11.1
presents
the resr:Its ofthis study.
1-1.4.5 Moderator Effects in the Deliberative
and lmplemental Mindsets
Mindset research has now a-lso established that the effects
of deliberative and implemental mindsets are moderated by
bothindividual differences (see the foliowing overview) and
context variables (cf. Goilwitzer, 2003).
Bayer and Gollwitzer (2005) discovered that students with a
high self-view of intellectual capability look for both posi-
tive and negative information that is trighly diagnostic with
respect
to their achievementpotentialwhen in a deliberative
mindsei, but focus only on positive in-formation, whether its
diagnosticity is high or low, when in an implemental mindset.
In contrast, individuals with a negative self-view of inteilec-
tual capability focus on positive information (irrespective of
its diagnostlcity) when in a deliberative mindset and look for
highiy diagnostic information, whettrer positive or negative,
when in an implemental mindset.
The situational context has aiso been sholvn to moder-
ate the effects of deliberative and implemental mindsets. To
date, research on this aspect has focused on predictions on
the stability of participants' römantic relationships (Gagn€
& Lydon, 2001a; Gagn6, Lydon, a'eartz, 2003). For exampie,
Gagn6 and Lydon (2001a) found that deliberating on deci-
sions
that have already b een made can initiate defensive pro -
cessing of relationship-related information. Participants who
were involved in a romantic relationship were asked to con-
sider the positive and negative consequences of a goal deci-
sion that was either associated with the reiationship or had
nothing to do with relationstrips in general, and the proba-
bility that those consequences
wouid occur (see
Section 1I.4
for detaiis of mindset induction). Gagn6 and Lydon found
that participants gave their partner much higher ratings if the
goal decision theyhad consideredwas related to the reiation-
ship than if it was not. Interestingly, the partner ratings given
by participants in a delilerative mindset were more positive
than those given by participants in an implemental mind-
set, Gagnd
and Lydon (2001a)
concluded that deliberation on
one's relationship may be perceived as ttueatening, ard that
participants evaluated thelr partrrer in more positive terms
in order to ward off this threat. In a further study, Gagn6 &
Lydon (2001b) assessed
the commitrnent pa-rticipants felt to
their relationslrip using a questionnaire measure. It emerged
that onty high-commitment participants boosted their rat-
ings of their partner to defend their relationship against the
threat posed by deliberating on a relationship problem; iow-
commitment participants did not. Thus, commitment to the
relationship is another important moderator of the effects of
the deliberative and implemental mindset in the context of
romantic relationstrips.
SUMMARY
Self-concept and the context of romantic relationships have
beensholvrrto moderaie the effects of deliberative andimpie-
mental mindsets. Self-concept moderates mindset effects on
the processing of high or low diagnostic in-formation about
281
il,]l;
i,iir
irl
l
r'
Control Deliberative Implemental
Mood 10.05 -252 11.30
Risk 6.05 6.00 9.77
Self-esteem 4L.77 37.55 41.08
Optimism 30.55 27.36 29.03
scales: mood:
Multiole Affect
Adiectivl
Checklist
(MAACL;
Zuckerman &-
Lubin, 1965);
risk: Measure of Reiative
Perceived Risk
(Perloff
& Feuer, 1986); seif-esteem:
Rosenberg
Self-Esteem
Scaie
(Rosenberg,
1965);
opiimism: Life orientation
Test
(L0T;
Scheier &
Carver.
1985).

282 A. Achüiger
and P. M.
Gollwiüer
Effects
on
self-concept
Effects
on information
processing
Effects
on optjmism/
h ^^^:-:^-
Pssrilr>Il
Effects
on motivation
Low
assessment
of
self-esteem
Respondents
rate
themselves
somewhat
higher on
positive
cha racteristics (e.g.,
intelligence,
creativity)
than compared
to others
High
ratjngs
of
own vulnerability
to controllable
and
uncontrollable
risks
0pen-mindedness
to information
of all kinds
Thoughts
tend
to focus
on
"deliberative"
behavior
Good
recall
of
others' deliberative
behavior
0pen-mindedness
to
incidental
information
Low feeling
of control
over
uncontrollable
events
Realistic
view
of
one\ future
performance
Comparatively
negative
mting
of one's
relationship/partner
Lower
peßistence
in
Butting
intentions
inio
practice
High
assessment
of self-esteem
Respondents
rate
themselves
much higher
on
positive
cha racteristics
(e. g.,
intelligence,
creativity)
than
compared
to others
Low ratings
of own vulnerability
to
controllable
and
uncontrollable
risks
Preference
for information
conducive
to the
enactment
of
an
intention
Thoughts
tend
to focus
on
"implemental"
behavior
Good recall
of
others'
implemental
behavior
Attention is
centrally focused
Illusionary
feeling
of control
over
uncontrollable
events
optimistic
view
of one's future
pedormance
Comparatively
positive
rating
of
one's
relationship/partner
Higher
peßistence
in-putting
intentions
into
practice
Deliberative
mindset lmplemental
mindset
iliili
iit,i
ltlr,l
i]][i
li,li
:i.
personal strengths or weaknesses.
The context and commit-
ment to a relationship moderate mlntdset effects on people,s
evaluations of their parfners.
LL.4.6 Mindsets and Goal Achievement
Studies on the effects of deliberative and implemental mind-
sets on goal achievement supported the hypothesis that the
implemental mindset is more conducive to goal attainment
than the deüberative mindset, because
both information pro-
cessing and self-evaluation are focused on the task at hand
(Section
11.4
).
Agoodpredictor of goal attainmentin everydaylife isper-
sistence of goal-directed behavior, i.e., the tenacity people
show in their endeavors to overcome öfficulties and mas-
ter challenges. Accordingly, some authors have investigated
the effects of the deiiberative and implemental mindsets
on persistence of goal striving. Findilgs presented by pösl
(1994)
and Braadstätrer and Frank (2002)
suggest
that people
in the implemental mindset show greater persistence when
faced with difficult tasks. For example, Braadstätter arrd Frark
(2002, Study 1) found that participaats in the implemental
mindset persisted longer at a dfficr:lt puzzle than did partic-
ipants in the deiiberative mindset.
directed behavior and the perceived desirability of the goal
were either high or low, the persistence of goal striving was
not influenced by the mindset induced. However, when per-
ceivedfeasibiüty and desirabilitrrwere in opposition (i.e.,
one
was high and the other low), parricipants in the implemental
mindset showed greater persistence in goal-düected behav-
ior tha:r öd participants in the deliberative mindset. lmpor-
raltly', moreover, the persistence oi goal-directed behavior
associated with the implementai mindset is not rigid and
inflexible. Brandstätter arrd Frank (2002,
Study 2) observed
that as soon as a task is perceived to be impossibie, or per-
sistence in what was assumed to be goal-directed behavior
proves to be aversive, individuals in the implemental mind-
set are quicker to disengage from goal pursuit than are indi-
viduals in deliberative mindset. Thus, the persistence insti-
gated by the implemental mindset serms to be flexible and
adaptive.
With respect to the effectiveness of goal striving in the
implemental and deüberative mindsets, experimental find-
ings rep
orted byArmor and Taylor (2003)
indicate that imple-
mental mindsets are associated
with better task performance
than deliberative mindsets, and that this effect.is mediated
by the cognitive orientation of the implemental mindset, e.
g.,
enhanced self-efficary, optimistic outcome expectations, etc.
(Section
II.4.4).
lO The
implemental
mindset
is more
conducive
to goal
striving than
the deliberative
mindset.
All effects of deliberative and implemental mindsets identi-
fied to date are documented in Table 11.2.
The fildings presented above
raise questions about ttre self-
regulation of goal striving. Can people intentionally induce a
certain mjndset in order to increase
their prospects ofreach-
ing a certain goal, or to facilitate disengagement from a goa1,
should it prove unrealistic or undesirable? The implemen-
tal mindset has proved ptu-tic'Jarly effective for promoting
goal strir'ing (Section 11.4.6).
In the studybyArmor and Tay-
lor (20031
mentloned above, ihe optimistic assessments of
T!9 E$"g. presented
by Pösl
(1994)
paint
a differenti- L7-4'7 ConcludinS
Discussion:
Mindsets
atedpicture.ll/henboththeperceivedfeasibitity-öTaegoa-
-nti€eif-Reguiaiion'of Goai-striving-
-'-

E:
a.
ä,,
F:
Motivation
and
Vcliiicn
in the
Course of Action
goal success
associ.ateci
with the implemenial mindsei led to
1nore effectiye self-reguJ.ation
of goal striving and to better
outcornes on an achie,.,ement-related
iask rhan the iess
opti-
mistic expeciations associated with the deliberative mind-
set. Iikewise, PösI (f994) and Brandstätter ald Fralk (2002,
Studies 1 and 2) showed that induction of an implemen-
ta1 mindset increased the likelihood of goai attainment; this
effect seems to be primarüy attributable to the greater per-
sistence in goal striving associated with the implemental
.-'=,'
'-'- forget that the positive effects of this mindset apply primar-
' ily to tasks conducted immedlately after it has been induced.
The more time elapses
between the induction of the imple-
, - ... =:.,mental mindset and task performalce, the less
pronounced
: its positive effects on goal attainment, as Gagn6 and Lydon
. ",,,','''-'
(2001u)
and Puca (2001)
have shown.
SUMMARY
Critically, the induction of a mindset does not have a
pefinzrnent influence on information processing and seH-
evaluation; the effects of the deliberative and impiemental
mindsets only applyfor a certain time.
zöö
ol self-reguiation. Goür,r'iüer
concludeci thai goais can often
only be attained when goal pursuit is supported by ttre seH-
regul.aio4i strategy of plaaning. Planning is understood to
be the menta-l anticipation of goal achievement- Against this
background, iwo tq)es ofintention are distinguished:
x goal intenrions and
* implementation intentions.
The concept of "goal intentions" has much in common with
Lewin's (1
92 6b) conceptualüation of intentions.
sense.
Examples of goal intentions are: "1 intend to be a good psy-
chologist" or "I intend to be ftiendly to a certain person."
O Implementation
intentions
are subordinated
to goal
intentions;
they
are
plans
that
promote
the attainment
of
goal
intentions.
In
forming
implementation
intentions,
individuals
specify
the anticipated
situ-
ations
or conditions
that
will trigger
a ceftain
goal-directed
response
(see
the
example
below).
lmplementation
intentions
have
the struc-
ture
"When (if)
situation
X arises,
(then)
| will
perform
response
I'
and are
often
called
if-then
Plans.
i mnaset'
L: any discussioä of tl:e,reladonshipbetweer*J:e-inp,le--4-Goal-islP-r$-iorls-qpeciry desired
end
stat-es
that hav.e-lot-yetieen
i -";;J.-a."r and goai reatjzation, it is impoftart not to attained.
Hence,
goal intentions
are "goals"
in the conventional
11.5 Different
Kinds
of Intentions:
Goal
lntentions
and
lmplementation
lntentions
Both scientif.c psychology and naive everydaytheories often
advocate goal setting as a good strategy for enacting wishes
and meeting demands. Yet numerous studies have shown
, that goal setting alone does not guarantee the accomplish-
ment of those goals - even highly motivated people often
flnd it difflcult to transiate their goals into action. Sometimes
they are simply hesitant to äctually take action to achieve
their goa1s, and do not initiate goal-directed behavior for this
reason. Sometimes they strive for too man-v, often compet-
ing, goats
at the same
time, i.ncluding long-term projects that
cali for repeated efforts over extended periods. Sometimes
the situational conditions are not conducive to goa-l
attaia-
ment. For example, someone whose attention is focused
on intensive emotional experiences will be distracted and
may thus fail to register an oppoftunity to act on his or her
goals.
(E Contrary
to the
widespread
notion that
goal
setting
is a sufficient
condition f0rthe
accomplishment
of
personal
goals
and
projects,
an
extensive
body of
research shows
that many
goals
are
never actually
Put
into
Practice.
Drawiag on the work of Narziss Ach (f 905, 1910, f 935) and
KurtLeurin (1926b),
Golb,vitzer
(1993,
1999)
addressedüre
clif-
f,culties of translating goals into action ftom the perspective
How then, do implementation intentions differ from habits?
In both cases, behavior associatedwith a certain situation or
stimulus is initiated automatically as soon as that situation or
stimuius is encountered.
lD lmplementation
intentions
differ
from habits
to the
extent
that they
originate
from a single
act of
will: the
conscious
pairing
of
a desired
goal-directed
behavior
with a critical
situation
or
stimulus.
By con-
trast,
habits are
formed
by the repeated
and consistent
selection
of
a certain
course
of action
in a specific
situation
(cf.
Fitts & Posner,
1967;
Newell
& Rosenbloom,
1981).
!1.5.1 How Do lmplementation lntentions Work?
Numerous studies have investigated the psychological pro-
cesses
underlying the effects of implementation intentions-
The focus of research has been on the chronic aclivation of
the situation speci.fied
in the implementation intention and
on the automatic initiation of the action specified.
lit
l!
i!i
i,

284
Ttre Situation Specified: Chronic Activation
Because forming an implementation intention impües the
conscious selection of a critical situatbn or stimu_ius
as the
if-part of the impiementation intention, the mental rep_
resentation of this situation is assumed to be highly acti_
vated andthus easily accessible (Gollwitzer, 1999;
Gollwitzer,
Ba_ver,
& McCulloch, 2003).
This heightened cognitive acces_
sibility makes it easier for people to detect arld attend to
the critical situation in the surrounding environment, even
when they are busy with other things. At the same time, it
facilitates iecall of 'J:e
critical situationinterms ofhorr,,
where,
and when the goal-düected behavior is to be enacted..
Fur-
thermore, speed of perception differs: criticai situa[ons are
perceived more quicklythan situations no.t
specifi ed.in
imple_
menta-Lion
intentions. Ä ciassic cognitive accessibüitv sturiv
is described below.
Clässic Study on the Cognitive Accessibility of
Situations Specified in lmplementation Intentions
Findings
from a dichotic
listening
experiment
show
that words
describing
the anticipated
critical
situation
are
highly
disruptive
to focused
attention.
Mertin (1994) presented
participants
with
words
to both
eaß simultaneously
via headphones.
participants
were
instructed
to "shadow"
the words
presented
on
one channel,
i.e.,
to repeat
these
words
as soon
as
they
heard
them,
and
to
ignore
the
words
presented
on the other
channel.
Attention
was
thus
focused
on
one
channel.
lt emerged
that
participants'shad_
owing
performance
was
much
slower
when
words
retating
to the
critical
situation
were
presented
to
the nonattended
channet
than
when
unrelated
words
were
presented.
In
otherwords,
critical
words
attracted
attention,
even
when
efforts
were
made
to direct
aüention
to the
shadowing
task. The
same
effect
was
not
observed
either
in
a goup of participants
who
had
only
formulated
a goal
intention
without
furnishing
it with implementation
intentions,
or in
a group
who
had
not
formulated
any intentions
at all
on
how
to approach
the task
at hand.
This
finding
indicates
that the critical
situatjons
specified
in
implementation
intentions
are
unlikely
to escape
people,s
atten_
tion,
even
when
they are
busy
with
other
things.
:nG
:+
A. Achüiger
and
P. M.
GollwiEer
48 hours iater, particrpants who had specified their choices :
in an implementation intention recalled these options much
more effectively than participants who had formujated goal
intentions only (Goliwitzer et al., 2002).
Finally,
Aarts,
Dijksterhuis, and Midden (
1 99
9), using a
iex-
ical decision task, provided further support for the assurnp_
tion that implementation intentions lead to heightened acti_
vation of specified situational cues. participants who had
specified critica-l
cues in implementation intentions showed
faster iexical decision responses than did participants who .
had onlyformed goä_l
intentions. ;
:
O The
chronic
activation
of
the situation
specified
in
the
implementa- _..
tion
intention
is
thus reftectqd
in its
heightened
cognitive
accessi_
bility,
which
in
turn facilitates
effectively
detecting,
readily
attending .
to,
and
successfully
remembering
critical
situational
cues. i
,:
,,.::
Implementation Intentions andAction kritiation .
As mentioned above, action initiation becomes automatic
once an implementation intention has been formulated
tf_uough a single act of wili. In forming implementation , . I,
iritentions, individuals can stategically switch between the : - .',1
conscious and effortful control of goal,directed behaviors .:.
and the automatic control of these behaviors in response ,
to selected situational cues. Goilwitzer et al. (2002; e.g., j
Gollwitzer & Schaal, l99B; Gollwitzer, Fujita, & Oettingen, ,
2004)
call this type of automatic action control strategic auto
- - ',,,''':
maticity. The goal-directed behavior specified in the imple-
mentation intention is assumed to be triggered immediately, , :
efflcientiy, and without conscious intent whenever the criti- , ;, :
i
ca]situationisencounteIed.Thus,someonewhohasformed
an implementation intention does ' rt have to invest cogni-
tive resources in conscious and effordul control of the goal-
directed behaviors specified in an implementation intention; ,
rt I,
rather, their performance is placed under the d.irect control
ofsituationalcues. ,1.t..
Implementation intentions are thus more effective than
goa-l
intentions alone in various respects. For example, it has , ' , -,
been shown thatparticipants who have formed implementa-
tion intentions respond to the critical situation immediately,-. , _-J''
even at high levels of distraction. The flndings of dual-task r, .il
rhe
nndings
or
a
study.using
the
Embedded
Figures
rest :ffi'ffi"ffi,ä:ä:&'jf.'jrä?r"j#:",ä1[ä:in#; "''.
(Gottschaldt, lg26) providefurtherevidencefortl-leenhaaced 2001; Achtziger, Michalski, a couwltzer] i"rarr""*gl.-p*- '-',-=.
cognitive accessibility of the critica-l situation. The objective ticipants in these experiments have to perform two tasks at - .,.:.
-.
ofihis.iesi.isiodetecismajier..4.ffgWeS'"that;l';ru;""";".i;;r.r*=
within larger "b-figures." Participants who had specifled the interpreted as indicating that the other task taxes cognitive
"a-figure" in the if-part of an implementation intention were resources. A series of studies using this arra-turt p*äaigot
better able to detect these hidden figures than participants have showa that cognitive ."ro*"".. *. not requted to ini- l
who had onlyformulated a goal intention (steller, 1992). tiatö the responses
Lciuced by implementation intentions.
In a cued recalJ'
experiment, participants had to decide For example, two experiments by Brandstätter et aI.
(200L,
when, where, and how to play certain games by choosing Studies 3 arrd
4) showed that students working on a task that
between a
number of set options offered bvthe er:peiimenter. required them io press the response burton as soon as a par-
krasurprisememorytestadministeredbothim-inediaiel-vand iicular stimuii appeared on the computer screen r"rpood"d
*-- -./

F
t::
t'.
&.:
,i-
.eo
IC
a
E
a
3
o
E
Motivation
and
Voliiioß in
the
Course
of Acticn
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
285
firmed, wj.th frontal Iobe patienis shorting significantly fasrer
reacti.on
times than the conftol group.
@ This
finding indicates
that
-rhe
execuiive
frinctJons
governed
by
the
frontal lobe
are not required in implementation
intentions, thus
suggestingthat
implementation
intention
effects
are
primarily
based
on
automatlc
processes.
Further experimental support for tJlis finding has been pro-
vided by Achtziger et aI. (forthcoming) and Gart'rilow and
executive was biocked, and therefore judged the target per-
son in a stereo
typical manner. Blocking the central executive
puts aheavyload onthe ftontallobes (Baddeiey,
1996)'
mean-
ing that automatic processes take precedence. The finding
that implementation intentions take effect even when the
central executive of working memory is blocked conflrms
that implementation intention effects do not tax cognitive
resources.
Gawrilow and Goliwitzer (2004)
demonstated the effects
of implementation intentions in a group of children diag-
nosed with attention deficit hyperactivity disorder (ADHD).
Children withADHD are knovrn to have important deficits in
executive functioning and hence in processes that tax cog-
nitive resources. They consequently flnd it very ditf,cult to
respond quickly and reliably to stop signals. Before being
administered avariation of the stop task (cf.
Logan, Schachar,
&Tannock, 1997),
childrenwithADHD were asked
to formu-
late an irnplementation intention specifying that theywould
stop what they were doing as soon 3s they encountered a
certain stimulus. Findings showed that, having formulated
this implementation intention, ADHD children managed to
inhibit the behavior in question just as well as a control group
of healthy children- Thus, the studyprovided further evidence
that impiementation intention effects are
primarily based on
automatic processes, and not on processes that involve cen-
tral executive functions, ald hence tax cognitive resources.
Gollwitzer and Brandstätter (f 997, Study 3) demonstrated
tJleimmediacyof actioninitiation as soon as
the critica-l situa-
tion is encountered. One group of participants formed imple-
mentation intentions that specified viable opportunities for
presenting counterarguments to a series of racist remarks
made by a confederate of the erperimenter; another group
formulatedgoalintentions to the same effect.As expected,
the
implementation intention particip ants initiated their coun-
terarzuments to the raclst comments more quickly.than
lmplementation
lntention
sruDrEs lwrrr cFrNrcAr sAMPLEs. In further stuöes,
' Brandstätter et al. (2001)
showed that even
patients who have
severe
problems with action control from chronic cognitive
load can benefit from implementaticn intentions. For exam-
' ple, drug addicts under withdrawal benefited ftom forming
. implementation intentions specifying when and where to
, perform actions that would.facilitate thet return to "nor-
mat" life. Most implementation intention patients succeeded
. in vwiting a curriculum vitae to be used in job applica-
:-
" 'tions before a set deadline, whereas goa-l
intention partic-
ipants missed the deadline. In other words, the chronic
' cognitive load associated with withdrawal did not inhibit
goal-directed behavior if an implementation intention had
been formed.
: Lengfelder and Gollwitzer (200I) tested the hypothesis
'.hat irnplementation intentions automate action initiation
in studies with frontal lobe patients. Individuals with frontal
lobe injury typically have problems with the conscious con-
trol of automated actions or habits. \^/henever they see
a pair
of scissors,
for example, they will reach for the scissors and
begin cutting, and are not able to consciously and deliber-
ately inteüupt.tfiat action, no matter how hard they try. In
other words, a stimulus associated with the execution of a
particular action will involuntarily and inevitably trigger that
action in these patients. Against this background, Lengfelder
and Goliwitzer (2001)
administered a go/no-go task to ftontal
lobe
patients. ln this type of task,
participants have to respond
to selected stimuli (e.g.,
to press a button when two of five
visual patterns appear on a computer screen),
but not to oth-
ers (i.e., seiective attention). If implementation intentions
are indeed based on automatic processes, as assumed by
Lengfelder and Gollwitzer (2001), the padent group should
show.faster reaction times to the situational cues specified
in an implementation intention in the go/no-go task tban a
control group.
of healthvindividuals. This predictionwas con-
Conlrol Il
Condition
Fiswe
11.2 Reacrion
times in a duar-task experiment
with and
wi-thout
::xff:fl:1tyjä:::::äiff.fi::t-ä5:,5.iä:i
f lgults rr'z nudLuull ullltr5 lll d uudr-toö^ o Pcrrrrrcrrt
wrtrr orru wrurvur -vanrr+irra ^f ,"rarleino.momnnr fcf Parldolo.r l qoAl-Aahtziqer
impiemenmtrori
inteniions._1Branoslätt-eaIen$erd-ea
&Göllwiuer,
2l0f)--KT-".i-"=:=-.! --:':::1 ::'l::--:::',:^:;^:*:"::
lrlrPrv'rrwrruuv.r
of participaa* who had formed an impiementation inten-
substantially faster if they had formed an impiementation tion to support the processing oi stereotype-inconsistent
intention, even when a dual task had to be performed at the information about a target person did not differ depend-
same
time. Students who had onlyformed a goal intention to ing on whether or not the functions of the central execu-
respond as quickiy as possible did not show enhanced reac- tive had been blocked. However, participants who had not
tion times under the duai-task condition. The results of ttr-is formed an implementation intention proved unable to pro-
study are presented il Fig. 11.2. cess stereotype-inconsistent inJormation when the cenüal

,;i]
]
ZöO
did dre goal intention only participants. The study pre-
sented.below provided empirical evidence that implemen-
talion intentions lead to action initiation even in the absence
of conscious intenr.
Study on Action lnitiation in the Absence
of Conscious lntent
Bayer, Achuiger, Malzacher, Moskowiu, and Gollwiüer
(forthcom-
ing)
conducted
two experiments to test
whether
implementation
intentions lead
t0 action
initiation without conscious intent once
the critical situation
is
encountered.
ln these experiments, the crit-
ical situation
was
presented
subliminally
(i.e.,
below the threshold
for
perception).
In
Study
1, Bayer and colleagues
investigated whether
partic-
ipants were
able to achieve their
goal
of asserting themselves
against a rude experimenter
by formulating an implementation
intention. Half ofthe
participants
were encouraged to set the
goal
of reprimanding the experimenter
by drawing attention to her
rude
behavior
(goal
intention condition);
the other half were addition-
ally instructed to plan
to take this action
as soon as they set eyes
on her
(implementation
intention condition). Afterwards, faces of
either the exDerimenter who had showed
the rude behavior or a
neutral, unknown
person
were
presented
subliminally
(as
primes)
to all
participants
by means of a tachistoscope
(presentation
times
of
less
than
10 ms). Primes are stimuli that serve to activate
asso-
ciated cognitive
contents. These cognitive contents are
presented
subsequentto the
primes
and their effects are measured, usually
in
terms of
reaction
times.
lmmediately after each
prime, participants
were
presented
with ceftain words, some of which were associated
with rudeness
(e.9.,
offensive, aggressive, arrogant). Participants
were
asked to
repeat all ofthe words as
quickly
as
possible,
and the
latencies
of their
responses were measured by the computer.
After
the subliminal
presentation
of the critical
primes, participants
who
had formed an implementation
intention to reprimand the exper-
imenter
as soon as they
set eyes on her showed faster response
times to words related to rudeness than did
panicipants
who had
only
formed
goal
intentions.
This finding
provides
fufther confirmation that the
goal-directed
behavior
specified in implementation intentions
is initiated auto-
matically
- i.e., triggered
immediately, efficiently, and without con-
scious intent
- as soon
as
the
critical situation is encountered.
TroN EFFEcrs. Mght the effects of implementation inten-
tions be atüibutable in part or even wholly to an associ-
ated trcrease in goal commitment? If furnishing goals with
implementation intentions indeed produces an increase in
the level of commi.tment to superordinate goal intentions, the
assumption that implementation intentions automatize the
initiation of goal-directed behavior and other cogrritive pro-
cesses would b e immaterial. However,
this hlp othesis has noi
received arilr empirical support. For example, Brandstätter
er a-1.
(200I,
Srudy
11 found thatrhe posiuve
effect
of an imple-
A. Achüiger and
P. M. Gollwitzer
mentation intention to submit a curricr:Iurn vitae before a
specified deadline was independent of the patients' general
commitnent to writing a curricr:Ium vitae. Patients in the
implei'nentation intention group were no more committed to
the goal than were patients in the goal
intention group. Anal-
ogous results have been reported in numerous studies ftom
domains such as disease
prevention (e.9.,
Orbell, Hodgkins,
& Sheeran,
t997), social impression formation (Seifert,
2001,
Studies I and 2; Achtziger, 2003, Studies t and 2), and ten-
nis competitions (Achtziger, Gollwitzer, & Sheeran,
in press,
Study 2).
AII mechanisms knormn to underlie the effects of imple-
mentation intentions are listed in the foilowing overview.
vacation. Students were given the task of writing a report
about Christrnas Eve no later than 48
hours after the event.
As
expected, students who had formed a corresponding imple-
mentation intention were significantly more likely to write a
report within the allotted time thart students who had onl1r
formed a goa-l intention.
Orbell, Hodginks, arrd Sheeran (1997)
found that women
who had sei themselves flre goal of performing regular
breast self-examilations gready benefited fron forming
implementaiion intentions. Similar patterns of results harze
t1-.5.2 lmplementatiqn lntentions and the Initiation
of Wanted Behavior
Because'
implementation intentions facilitate attending to,
detecting, and remembering situations conducive to goal-
directed behavior and, in addition, help to automatize action
initiation, people who form implementation intentions can
be expected
to showhigher goal-
attainment rates
ttranpeople
who do not fumish their goal intentions with implementation
intentions. Thö resuits of a host ofstudies in very different
domains provide empirica-l support for this hypothesis.
Effects of Implementation Intentions on Achievement-
and Health-Related Behavior
Research on implementation intentions tends to examine
goal intentions that are difficult to attain for reasons already
mentioned; e.g., because of external or internal distractions
or because the action required is unpleasalt or painfui. For
exampie, Gollwitzet and Brandstätter (1997)
analyzed a goal
TIIE ROLE OF COMMITMENT IN IMPLEMENTATION INTEN- intention that had to be performed during the Christmas

Motivatlon
and Voiiiion !n
the Course
of Action 287
emergeci for parricipation in -'roluntary caacer screenjag implementation intention effects than those in the former
(sheeran & Orbe1l, 2000), resumption of functional activiry $oup.
afrer
hip ieplacement surgery
(Orbell
& Sheeran, 2000), and Sheeran
et a-1.
(2005,
Srudy
2) fo'.:nd
t&at implen:entaCon
engagenenr in physical exercise Milne, Orbell, & Sheeran, intention effects only occur when the respective superordi-
2002). Furthermore, impiementation intentions have been nate goal intention is activated. The implementation inten-
found to facilitate the attainment of goal intentions that tionto move ontothenertiteminanintel.ligencetestimme-
are otherwise easilyforgotten;e.g., regularintakeof vitamin diately after finishing the previous one enhanced speed of
tablets (Sheeran & Orbell, 1999) or sigrr.ing
each page of an task processing only when the goal intention of working as :
intelligence test (Chasteen,
Park,
& Schwarz, 200\). quickly as
possible was activated. Likewise, in a! eq)eriment
usiag the Rogers.and ivlonsell (1995) task-switch para.ligm,
--Si3=lficentModeretsrsofIn:plencntati'oGoüw-itzer(2002)ioirn<ithatimpiemen-
IntentionEffects tation intention effects are deoendent on the superordinate
The strengttr of impiementationintention effects depends on goal being activated.
the presence or absence of various moderators. Some stud- Fina-lly, it carr be assumed that the strength of the men-
ies (e.g.,
Gollwitzer & Brandstätter, 1997, Study 1)'show that tal link beilveen the if- and then-parts of an implementa-
_the more diffrcult it is to initiate a goal-directedbehavior, the tion intention moderates its effects. For example, if a person
more pronounced implementation intention effects become. irrvests a lot of time aIrd concentration in encoding an imple-
..- The findings of the studywith frontal lobe pati6nts described mentation intention in long-term memory and/or menta-lly
above[Lengfelder&Gollwitzer,2001,Study2;Section. 11.5.1) rehearsing ttrat intention, stonger menta] links should be
are relevant here as we1l. Patients with a ftontal Lobe injury {orged between the two parts, which should in turn produce
tlpicaltyhave problems with the consci.ous
control of behav- stonger implementation intention effects. This assumption
ior because
their access
to executive
functions and cognitive has not yet been subjected to experimental testing, however.
, resources
is limited. Findings show that patients who formed
an implementation intention in preparation for a reaction
. time task outperformed a sample of college students who
had formed the same
implementation intention. Because the
reaction time task can be assumed
to be more diffi.cult for the
patients than for the healthy students, this finding confrms
--" .'
that forming implementation intentions is particularly ben-
eficial to people faced with difficult tasks.
' r Q6s1151i111ent
to the soal intention also seems
to moderate
the effects of implemenlation intentions. Orbell et al. (1997)
report that implementation intentions only enhanced com-
pliance in performing breast self-examilations in women
lvho strongiy intended to examine their breasts, i.e., who
were comrnitted to the superordinate goal intention. Simi-
larly, Gollwitzer et al. (2002, Study 3) found that beneflcial
effects of implementation intentions on participants' recall
of critical siruadons were only observed. when the goal inten-
tion had yet to be translated into reality. Ifit had already been
accomplished, no implementation intention effect on mem-
oryperformancewas detected. F.urthermore, Sheeraa,
Webb,
and Gollwitzer (2005, Study 1) showed that the beneficial
effects of implementation intentions conceming the goa-l
of
To date, research has focused a-lmost exclusively on how
implementation intentions cal help to translate goals into
action by facilitatingwanted, goal-directed behavior, andpar-
ticularly the initiation of goal-directed behavior. Yet merely
initiating goal pusuit rarely suffices to achieve a goal. Once
initiated, a process of goal striving ha- to be maintajned. Peo-
ple need to shield their goals ftom distractions or conflicting
bad habits. Ways in which implementation intentions can be
used to control these "urrwanted" effects are outlined below
Unwanted responses that hamper tfie successful pursuit
of goals can be controlled by different types of implemen-
tation intentions. For example, someone who wants to avoid
being unfriendlyto afriendwho is knorrrn to mäke outrageous
requests can protect herself from showing the unwanted
response by forming the goal intention "I intend to stay
ftiendly'l and furnishing it with one of the following three
suppression- oriented implementation intentions:
I Ist suppression-oriented impiementation lntention:
"A-nd
if my friend makes an outrageous request, then I will
not resp ond in an unfriendly marrner.
" The strate
gy here is
to control and suppress unwanted behavior by specifring
preparing for an upcoming exam increased as a fi:lction
of the amount of studying required. Irr addition to strength
of commitment to the goal intention, commitment to the
specif.c implementation irltention is required. In the mem-
ory study by Gollwitzer et al (2002), the strength of the
commitment to the implementation intentionwas varied by
telling partici.pants (after administering a battery of personal-
ity tests)
that theyrvere the qpe of person who wouldbenefit
either fromstrictly adhering to theirplans (high commitrnent
condition) or from staying flexible (low commitment condi-
tion). Participaats irr the iatter group shor,ved
notably weaker
SUMMARY
The diffl culty of initiating goal-directed behavior, the strength
of commitnent to goal
intentions and implementation inten-
tions, and the activation of the goal intention have proved
to be significant moderators of implementation intention
effects.
11.6 lmolementation lntentions and the Control
of Unwanted
Behavior
ri
i,,lli
,r r,. i
iili;l
li,.l.l
r'iit
i'rllit
i,ri!iii,
i.i:ilijl.
ii',.lli.i
lrr,iiilr
iilil,
i:,iil,i
iiilll
lilll,l'
.rl,..'
tl.,iii
t'lii,j
.liil,l,
r':-
I
;lil
I

288
the critical situation in the if-part of the implementation
intendon, and ruJjag out ttre un',vanted response in the
then-piut. Alternativeiy the focus may be on facilitating
the initiation of a r,rrarrted
response:
x 2nd suppression-oriented impiementation intention:
'And if my friend makes an outrageous request, ttren
I will respond in a fuiendly malner." In thjs case, the
critical situation is again specified in the if-part, and
the wanted response that is threatened by disruptive
unwanted responses
is endorsed in the then-part.
r 3rd suppression-oriented impiementation intention:
"t\nd if my friend makes an outf,ageous request, then I
will ignore it." In this variant, the critical situation is again
specified.in the if-part of the implementation intention,
and the then-part focuses
the person awayfr omthe critical
situation.
Goilwitzer and colleagues
have conducted a serj.es
of studies
using these three types of suppression-oriented. impiemen-
tation intentions. Most of these studies investigated the con-
nol ofunwanted spontaaeous responses to distractions or of
automatic activation of stereotypes and prejudice.
AL6.t Suppression-Oriented lmplementation
Intentions
lVhen goal pursuit is threatened by distracting stimuli, imple-
mentation intentions should be formed to inhibit those dis_
tractions, as illustrated by the study described beiow.
lmplementation
Intentions
and
Resistance
to Distractions
In
a eomputer-based
experiment
(Gollwi?er
& Schaal,
199g)
col-
lege
students
peformed
a series
of
arithmetic
problems
while
dis-
tracting
clips
of
popular
commercials
were
shown
at random
inter_
vals on
a W screen mounted
above
the
computer
monitor.
Findings
showed
that
goal
intentions
("1
will not
let
myself
get
distracted,,)
were less
effective
in
protecting
participants
flom
the distractions
of
the
commercials
than were
implementation
intentions.
Moreover,
implementation
intentions
phrased
as distraction-inhibitin g (,And
if a distraction
arises,
then
I will ignore
it,')
produced
better results
than
those
phrased
as
task-facilitating ("And
if
a distraction
arises,
then Iwill focus
my attention
on the arithmetic
tasks,,).
Specifi_
cally,
distraction-inhibiting
implementation
intentions
helped
par_
ticiDa[ts
to__Ward
off
the distractions
of
the cornne:.cials
re{erC]
0f their
motivation
to do the
tedious
arithmetic
problems,
whereas
task-faciiitating
implementation
intentions
were
effective
only
when
motivation
to do
the
problems
was
low
When
motivation
was
high,
task-facilitating
implementation
intentions
did not
shield
partici-
pants
againstthe
distractions
of the
cominercials,
and
performance
on
the arithmetic
tasks was
poor.
These
findings
suggest
that
task_
facilitating
implementation
intentions
may
result
in
overmotivation
in
distracting
conditions
and
thus
undermine
performance.
A.
Achtziger
and
p.
M.
Gollwi[er
coNTRoLLrNe pR_EJrJDrcE.
Researchers
have also iavesti_
gated the function of implementation intentions as shate-
gies for controlling i:nwanted stereotJ1)es
in impression foi-
mation. In general, models of impression formation (e.g.,
Brewer, l9BB; Devine, 1989) assume that the effects of social
stereot54)es
and prejudices on the way people judge others
are governed by processes that require attention, cognitive
resources, and conscious effort. Until recently, stereotlpe
research
assumed thatthe apptcation of stereotypes
-butnot
their activation - can be intentionally controlled (cf. Brewer,
l9BB; Devine, 1989). Stereotwe activation was thought to
be an unavoidable, automatic process; stereogpe use) ro
be controllable by effortfi:l correctional strategies.
Based on
the studies of the automaticity of implementation intentions
described above, Gollwitzer's research group conducted a
series of experiments to test whether implementation inten_
tions can inhibit the automatic activation of stereotypes and
prejudice, ald not just their application. The assumption was
that an automatic process such as the activation of a stereo-
tipe can be blocked by other automatic processes such as
those triggered by implementation intentions. Experiments
using different priming paradigms showed that the auto:
matic activation of the stereot)?e "old persorl, was inhib-
ited when participants formed an implementation intention
("\Atren I see an old person, then I wili tell myself: dont :
stereotype!"), butwas still observedin a group of partlcipants
who had formed a goal intenrion only (,,I intend to judge
fairly") and in a control group who were simply instructed '
to form an impression of the people presented (Goliwitzer,
Achtziger, & Schaal,
forthcoming). Analo gous
results emerged
from a study in which male participants were asked to
inhibit the stereotype "women," and studies in which par-
ticipants of both sexes were asked
to inhibit the stereotypes
"homeless person' or "soccer fans" (Achtziger & Goilwitzer,
2005).
Other studies investigated the extent to which implemen-
tation intentions cal pievent the application of stereot]4res.
Seifert (2001, Study 1) tested whether the d.iscrimination of
female job seekers appiynC for jobs in technical domains
carl be controlled by implementation intentions. Computer.
science students were presented with a number of applica-
tions for the position of computer scientist and a proflle of
the job's requirements. Haif the fictional applicants had a
woman's name, the other half aman's name. In apreliminary
and female narnes were assigned
to the applications at ran-
dom, however, the computer science
students were consider-
abJ.5r
more likeiy to hire male candidates, thus discriminating
against the female caadidates. Oniy a group of snldents who
had formed the implementation intention "14hen I evaluate
an application, then Iwill ignore the candidate,s
gender', man-
agedto orrercome
this bias.
Siereoq,peiesearchhas evidenced
that individuals under cognitive load are unable to process

Moiivation
and
Vüliiion in
the Course
0fAction
stereofrrpe-ilconsistent injormatj.on about unknor,r'n others
(cf. ![acrae, Hervstone, & Griffrths, 1993). Srereorype-
6consistent information is nct generallv
aüributed to repre-
seniatives
oi certain soci.al categories. For example, "machos"
are not usually chaJacterized as "tolerant." Successful
pro-
cessing of stereoq/pe-inconsistent informadon results in
nonstereotypica-l impressions. In two studies, Achuiger
et aI. (forthcoming) replicated ttre finding ttrat stereofire-
inconsistent information is poorlyprocessed under co
gnitive
load, and showed that peopie who formed implementation
-j-föntinnc are ahic t^ nrnaacc ciara^fihojn.^nc
!r.v-'--- --'
mation and hence to evaluate others fairly, even under cog-
nitive 1oad.
SI'PPRESSION OF EMOTIONAL RESPONSES. RCSCATCh hAS
shovrn that, apart from regulating unwanted behavioral
lqsponses (e.9.,
to distractions) and precludiag unfair eval-
, uations of others, implementation intentions can also inhibit
.-ll-=:unwanted emotional responses. For example, Schweiger
, Gallo,
Achtziger, ald Goilwitzer (2003)
report a study exam-
I ining howimplementation intentions can be used to inhibit
disgust.
Female
participants were presentedwithpictrue cues
, ftom.the InternationalAffective Picture System (IAPS;
CSEA,
' : 1999).
Some of these pictu-res show photographs of injured
,l and mutilated individuals, and activate the emotion "dis-
i. , gust." Participants were able to suppress their disgust by
.. : . means of an implementation intention, but not by means
. ofa goal intention alone.
SUMMARY
.Suppression-oriented implementation intentions have
proved effective in inhibiting spontaneous attentional
responses, stereotypical and prejudicial responses, and
reflexive
negative emotiönal responses.
11".6.2 Blocking Detrimental Self-States by Planning
Wanted Behavior
In the research
presented in Section 1I.6.1, the critical sit-
uation specified in the if-part of an implementation inten-
tion was ünked to a then-pait that served to suppress
unwanted responses. Implementation intentions may also
protect against unwanted responses in another way, how-
ever.
Instead of focusing on anticipäted obstacies and the
-unwanted responses they tri.gger, implementation intentioäs
may be designed to stabilize an ongoing goal pursuit. For
example,
an exchange of opinions can soon develop into an
arg!-rnent
if the parties are tired and worn out, even if they
did not intend the situation to escalate. However, if the parties
piannedin advance howto respond constructivelyto conflict-
ing opinions, the self-states offatigue and exhaustion should
not have a negative impact on the discussion. These assump-
tions have been tested in a series of studies, one of r,vhich is
describeri
below.
289
Study on Blocking Negative Self-States
one of the studies
on the use of
implementation
intentions io block
negative
self-states
(Gollwitzer
& Bayer, 2000, Study
1)
was based
on
the theory
of symbolic self-completion
(Wicklund
& Gollwiüer,
1982) and
tesied the extent
to which
the negative
effects ofself-
definitional
incompleteness on
social sensitivity
(cf.
Gollwiüer
&
Wicklund, 1985a) can
be attenuated
by forming
implementation
intentions.
Participants were
iaw students
who were
highly commit-
g,successful-lawyers.
As a cover storgthey-wele
told
thatthe study
had been designed
to analyze
how
goals
affect
how
people
getto
know each other.
To
this
end, theywould
be
introduced
to another
student;
their
goal
was
io take
iiai person's perspective
during
the conversation.
Half of the partjcipants
were
instructed
to furnish this
goal
with the following
implementation.
intention:
'And if my partner
expresses
a preference
for a certain
topic of
conversation,
then
I will directthe
conversation
to thattopic."
They
were then administered
a questionnaire
on
how
they
approached
their studies
('no
sense
of incompleteness"
condition)
or the
same
q
uestionnaire
with three supplementary
questions
rdrawing
atten-
ti0n to shortcomings
in
their
current skills
and experience
(e.9.,
"Do
you
have courtroom
experience as a judge
or district
attorney?").
This second
questionnaire
was designed to create
a sense of self-
defi nitional
incomoleteness.
Finally,
all
participants
were informed thatthe
peßon
they
were
to meet
was called Nadia, and that she had already
indicated
her preferences
for potential
topics of conversation.
Participants
were then handed
a sheet 0f paper
listing these
preferences.
lt
was
quite
clear that
Nadia did not
want
to discuss
law but
would
preferto
talk about
her lastvacation and
popular
movies.
To assess
whether self-definitional
concerns
would
increase the likelihood
of participants'
choosing law
as a preferred
topic of
conversation
despite Nadia's
preferences,
all participants
were asked to note
down
their own
preferred
topics
for Nadia.
In the control
condition,
a self-completion
effect
was clearly apparent
participants
with
an
incomplete self-definition
were more
Iikely to want to talk about
law than
participants
with a complete
self-definition,
even though
Nadia
was clearly not
interested in discussing
this topic.
The same
effectwas
notobserved in the
group
ofparticipants
who had formed
an
implementation
intention, however
- these
participants
showed
the
same low
preference
for law as a potential
conversation
topic,
whether
their
self-definitions
were complete or
incomplete.
These findings show
that implementation
intentions are able
t0
block the
negative
effects
ofthe self-state
"self-definitional
incom-
pleteness"
on goal-directed
action
(specifically,
taking someone
olco'c norcncntirio\
Implementation Intentions and Self-Regulatory
Performance
According to ego-depletion theory (Baumeister, 2000;
Muraven, Tice, & Baumeister, l99B), performing a task
that demands a high levet of self-regulation will encroach
on performaace on a second task that also requires
,lli
,lii
r,i'.lt.
lil[[l
iiiiii
,ril
I'ii.1,
iiiii
i]iirii
irrli
'lir
,.i
,1,,.
',,,11
,..

290
. self-regulation. Goliwitzer and Bayer (2000, Srudy 3) were
' interested in whether this effect could be co""t.*d;;;r;;:
meiitation intentions. In a cl.assic
ego_d.epletion
#;;;,
participalts were fust shown a humorous movie and
instucted either to erpress theü emotions freely, or to show
no emotions at all. Theywere then presented with a number
of diffrcr:it anagrams. All parricipaats had formed the goal
intention to solve as many anagrams as possible. Half the
participaats had furnished this goal.
intention witl an imple_
mentation intention: ,,And
if I have solved one anagram, then
Iwill move on immediatelyto the nex.,, participants who had
onlyformed a
goalintention showed t}le classic
ego_depietion.
effect, with those who had been instructea not ä show their
emotions duringthe flimperformingless well on the anagram
task than thcse.r,r'ho
had gi-u,en
free
rein to their emotions. This
effect was not observed
in participants who had furnished the
goal intention to perform well with an implementation inten_
tion, however.
Webb and Sheeraa (2003,
Study2) also demonstrated that
implementation intentions can offset ego_depletion effects.
First, half the participants were instructed to balance on their
"weaker" leg while counting down in sevens
from'l, 000 (ego_
depletion manipuiation). participants in the control condi_
tion counted to 1,000
in fives while standing nomally on two
legs. All participants were rhen given theloal intention of
naming the ink color of words presented in a Stroop test as
quickiy as possible. Half the participants fumished this goal
intention with an implementation intention: ,.\Alhen
I see a
word, then I wül ignore its meaning and name the color in
which it is printed." No ego-depletion effectwas observed for
implementation intention participants; those who had been
ego-depleted in the initiat taskperformed as
wenin the strooo
test as those in the nondepleted control condition. However,
participants who had onJy
formed a goal intention showed a
marked ego-depletion effect, with those who had been ego_
depleted scoring notably lower on the Stroop task than their
nondepleted counterparts.
SUMMARY
The negative effects of both sel_f_definitional
incompleteness
artd ego-depletion cal be blocked by forming implementa_
tion intentions.
l-1.6.3 Blocking Adverse Contextual lnfluences
by flqlling Wanted Behavior
People maysee the outcomes oftheir actions in terms of gains
or of losses (Kahneman & Tversky, 1929).
Conflict_resolution
research suggests
that cognitive processes
triggered by,,loss
framiag" or "gain framing,' have a strong UpJ., bn negoti-
ation processes
and theü outcomes (De Dreu et al., 19g4).
Loss framing resu.lts in comparatively unfair agreements arrd
othernegative
effects.
Trötschel
and
Gojh,ri,r., (200+.)
*rr.ro-
gated whether these negaCve
toss
ftaming effects
can be or,,er_
A. Achüiger
and
p.
M.
Gollwitzer
come if prosocial goals, such as firdirg a fair or integtative
solution, are fu rnished with corresp
ondin g implemenration
intentions. This hypothesis was tested.
in two e4periments,
the first of which is describäd belor,v.
Intentions and Performance Feedback
Goal attainment can also be negatively affected by unfa_
vorable performaace feedback conditions. One example
here is the 'lsocial loa_fing,,
phenomenon often obsewed at
workplaces where employees are
given collective, rather thali
individual performance feedback (cf. Latan6, Williams, &
Harkins, 1979;
Karau &\,\rüiams, 1993):
pecple r,vhenworking
Overcoming Loss Framing Effects by Means
of lmplementation Intentions
Pairs
of
participants
were
assigned
the roles
of
heads
ofstate
oftwo
rival
countries
and
asked
to
negotiate
ihe
partitioning
of a
disputed
island.
Tle island
was
made
up
of 25 regions,
each
representing
one
of four
terrains:
mountains,
cornfields, pastures,
or forests.
Within
each
pair
of negotiators,
one
participant
was
subjected
t0
ioss
framing
as foiiows:
r Loss
framing
condition:
The
participant
was
handed
a
übte
listing
the four
different
types
of regions,
and
specifytng
the
loss
that
would
be incurred
if each
were
relinquished
to the
other
participant
in terms of a negative
score.
The
other
par_
ticipant
in each
pair of negotiators
was subjected
to gain
framing.
I Gain
framing
condition:
In
this
conditjon,
the
regions
listed
in
the table
were
allocated
positive
scores,
indicating
the
gain
that
would
be incuned
if that
region
were
appropriated.
Both
participants
were
told
that
they
had
to
come
to
an agreement
on
the
distribution
of
the
25 regions
within
15
minutes.
A fairness
goal
was
instilled
in
some
participants
by handing
them
a sheet
of paper
informing
them
that
fair
negotiation
outcomes
are
often
.
very
difficult
to achieve,
and instructing
them
to set
themselves
the
following
goal
shoruy
before
entering
the
negotiations:
"l want
to find
a fair
solution."
Half
the
partjcipants
with
a fairness
goal
were
additionally
instructed
to fümish
this
goal
intention
with
an
implementation
intention:
,And
if my
opponent
makes
a proposal,
then
I will
make
a fair
counterproposal.,,
participants
in
the controt
condition
were
not instructed
to speciry
either
a fairness
goal
or
an implementation
intention.
Outcomes
were
assessed
in terms
of individual
"profits,,
within
each
pair
of negotiators.
In each
of
the
three
conditions,
the
authors
tested
whether
the
difference in
profits
within
each
dyad
was
significantly
different
from
zero.
In
both
the
goal
intention
condition
and
the control
condi!!gn,_
significant
differences
in
profits
were
observed
as
a
function
ofthe
framing
condition.
participants
who
had been
subjected
to loss
framing
made
higher
profits
than
those
subjected
to gain
framrng.
Unfair
outcomes
of
this
kind
were
not
observed
in
the imDlemen-
tation
intention
condition,
where
profits
were
equally
distributed
u--

291
rn groups where indir,idual perfoimalce cannot be moni- Before pariicipanis raiere
allorved to drive -tl:e flnal nvo
tored have been obseru.--ed
to show lor.ver
performance lev- circuits of the üack, auto-motive priming was used to acti-
els. Golltviüer a]ld.
Ba.ver
(2000,
Srudy 4) testeci
whether this vate fi,vo goa-Is
beyond ihe participants' conscious aware-
phenomenoncanbe cormteractedbymeans of i:rpiementa- ness.
AIl participants were asked to join the numbered dots
don intentions. Their participants were asked
to generate as presented on different sheets of paper as quickly as possi-
many uses as
possible for a common knife under one of two Lte to produce various shapes (flowers' anima-ls' and other
conditions: objects)' Those in the "move quicklf' priming condition
E ..Collective performance feedbaclc' condition: Partic- were instructed. to complete as many figures as possible in
ipants were told that their responses would be pooled flve minutes. Those in the "move slowly" priming condition
with those of seven other participants, and that thJexper- were told. to.join the dois as
careftill;r-and neatly as
possible'
--imenter would not-be-able to,tell-hor.\,-lraJr}leses-eaeh-ta-€-rg.as Il}ueh E:ne as thsy neeced fcr each-shape' Find-
individual had generated. ings showed that this auto-motive priming had prono'nced
e ,.IndivrCual perfornance feedbaclc' condition: Partic- effects on goal intention participants' driving in the last tlvo
ipants were told that the experimenter would be able to circuits: those in the "move quickly'' condition <irove
faster
assess
each participantt performance separateiy. and made more mistakes than those in the "move slow1y"
^L-^*,^.1 f^' imnlc-
4DJCOö cauu y<! t
. Before beginning the task, ali participants formed the goal condition. No such priming effect was observed for impie-
-:-.:'. intention,,Iintendtonameasmanyusesaspossible."Halfof mentation intentions participants, who drove at a moder-
r:::.r;-:-
-,ü1s palticipants furnished this goal intention with the imple- ate speed and made few mistakes in both priming condi-
. mentation intention: .,And
when I have noted down a use, tions. These findings indicate that goal pursuits furnished
r ,, thenlwilllirnmediatelygoontothenext."Thenumberofuses with implementation intentions are not affected by com-
: . eenerated
in 12
minuteswas taken as the dependentvariabie. peting, ionconscious goals that are activated by situational
' . "' äoa] intention participants generated notably fewer uses in cues'
' - ;J.;n;;Jä.-uo"" feedbacK, condition than in the Table ll.3 documents all effects of implementationinten-
. 'individual performance feedback' condition. This pattem tions that have been identified to date'
of results, which replicates the classic social loaflng effect'
was not observed in implementation iatention participants'
who generated an equal volume ofresponses, regardless of
the feedback condition.
As we have shor,nin,
implementation intentions facilitate goal
pursuit in various ways. It seems reasonable to hypothesize
that such an eft-ective
means of self-regulation may have cer-
tain unJoreseen costs.
This section examines
the three follow-
ing potential costs of implementation intentions:
- i. ft it possible that implementation intentions tread
to a
certain rigidity of behavior that may be detrimental when
task performance requires high leveis of flexibility'
2. It is possible that implementation intentions cause a
high degree of ego-depletion and thus undermine self-
regulatory resources.
3. It is possible that thoughts, feelings, and actions may
resurface later in a different context (rebound effects)'
although implementation intentions successfully sup-
presses tmwanted thoughts, feelings, and actions in a
given context.
Ll.7 .t lmplementation lntentions
and Behavioral Ri$iditY
Do people who have formed implemeniation intentions also
'recogrrize a.Iternative opporfunities to act towaid their goal'
or do they insist on acting only when the critical situation
specified in the implementation intention is enccurrtered?
The strategic .automaticity created by implementation
Motivaticn
and
Volition
in the
Course
of Action
Formation of Implementation Intentions
., 'andComPetingGoals
'' tAuto-motive thäory @argh, 1990;
Bargh & Gollwitzer, 1994)
holds ttrat when goal striving is activated repeatedly and con-
:, ' sistently in response to a given situation, this situation will
': eventuaily acquire the potential to trigger the critical goal
oursuitwithout conscious intent (Bargh, f 990;
Bargh & GolI-
witzer, 1994).
A goal intention that can be activated in tl-is
way is called a "chronic goal." Gollwitzer (1998) conducted
t!v:o experiments to test whether implementation intentions
can shield ongoing goal pwsuit against the effects of directly
. activated chronic goals.
' ln the fust study participants had to navigate a car along a
race trackin a simulator. The mean drivingspeed andnumber
of errors
were measured in two baseline circuits' Participants
' were then given precise instructions on how to drive thenext
tlvo circuits.
q Participants in the goa-l intention condition were
instructed to set themselves the goal of reaching the fil-
ishing post as quickly andwith as few errors as
possible'
o Participants in the implementation iatention condition
were additionatly instructed to form the following imple-
mentation iaientions: "And when I enter a curve, then l
will reduce my speed. And when I enier a straight section
of the track, then I wiJI speed
up again."
11.7 Potential
Costs
of lmplementation
lntentions

292 F.
I
:.:..
A.
Achtiger
and P M.
Gollwitzer
!:i
intentions - i.e., the delegation of behavioral control to
situational cues - can be assumed to free up cognitive
resources, thus allowing effective processing of information
about alternative opportunities. This assumption has been
confirmed in a number of studies showing that individuals
who had formed an implementation intention were not biind
to changed situational contexts or unexpected opportuni-
ties to achieve their goal. Instead of sticking rigidly to their
plaas, participants responded appropriately to new situa-
tions.
For instance, Achtziger (2003,
Study 2) showed that par-
ticipants are abie to form implementation intentions that are
only applied in certain contexts. A study on prejudice toward
were sho',ryn
two synrbols (e.g.,
flower, heart) on a monitor'
and asked
to select the s)rynbolwith the highest score.
Before'--'
the study began, they had been told the score of each syn-
bol, and some participants had formed the implementation
intention to select the qrmbol with the highest score espe-
ciaily quick by pressing the button as soon as it appeared'
A-fter a while, a new slnnbol with an even higher score was
presented on the screen. Participants in the implementa:
tion intention condition succeeded in selecting tfris new
symbol rather than the one that previously had the highest
score.
a1-.7.2 lmplementation Intentions
and Ego
Depletion -
soccer fans showed that participants were able to appiy the The assumption that implementation intentions autoDaa!9-.
impiementation intention "And if I see a soccer fan, then I'll the control of goal-directed behavior implies efflcient and
not evaluate him negatively'' flexibly, dependent on the con- relativeiy effort-ftee behavioral control. In other words, the
tert. Inttrisstudy,thepresenceofasignaltoneindicatedthat seif is not implicated - arid should therefore not become.
theimplementationintentionshouldbeapplied,whereasthe depleted - when behavior is controlled by impiementadon
absence of the tone indicated that it should not. In Line
with intentions. Empirical support for this assumption has been-
rhe.assumption that impiementation intentions cio nofne-c- provräeti by the studies of Gollwitzer and Bayer (2000) ano
essarilylead to behavioral rigidit5r the ilhibition of preju- WebbandSheeran(2003)reporiedinSectionll.5.2.Whether
dice.toward "soccer fans" was on-ly observed when pictures the initiat self-regulating task was to control one's emodons
of soccer fans were accompanied by a signal tone. Likewise, (Go[witzer & Bayer,
2000)
or to performwel] on a challenging
ariotJrerstudy (Jaudas
& Gollwitzer,2004) showedthatpa;tic- task(the Snoop taskWebb &Sheeran,2003),
impiementation
ipants who encountered an unexpected oppofuniry*to pur- intentions successfuliypreservedself-regulatoryresources' It
sue a goal intention - i.e., an opportunity other thal the one would thus seem that self-regulation based on implemen-
specifiedintheif-partoftheimplementaiionintention-were iaiion inrentions is not costiy in terms of self-regulatory
able to recognize arrd seize
tJ.is neu' opportu:rity. Pal
iicipants resources.
Conüolling
uriwanted
behavior Promoting
wanted behavior
Suppressing
unwanted
thoughts,
feelings, and actions
("suppression-oriented
implementation
intentions")
Inhibiting
automatic
activation of
siereotypes
(e.9.,
age stereotypes,
gender
stereotypes)
Inhibiting
prejudjce (e.9.,
discrimination
of
women in male-dominated
orofessions)
Shielding
against
distractibn during
complex tasks
(e.9.,
distracting
effects
of commercials
while
\vorking on
arithmetic
problems)
Controlling
impulsive behavior
in children
with ADHD
(e.9.,
enhancing
resoonse inhibition
in a reaction-time
task)
Replacing
unwanted
behavior by
other behavior
lnhibiting the
automatic
activation of
prejudice
(e.9.,
toward homeless
lnhibiting
negative emotions
(e.g.,
disgust)
Inhibiting behavior
that is detrimental
to
health
(e.g.,
cigarette
and
alcohol
consumption)
Shielding
wanted
behavior
frorn unwanted
internal and
external
influences
Blocking unfavorable
contextual
infl uences
(e.9.,
deindividuaiization,
competing
gcjal
activations,
framing effects)
Blocking detrimental
self-states
(e.9.,
self-definitional
incompleteness,
mood,
ego-depletion)
Fostering
the initiation
and execution
of goal-directed
actions
lncreasing
the
latency of
counterarguments
to iacist remarlG
Increasing
the
probability
of
participation
in cancer screening
(e'9.,
mammography)
Facilitating
the
processing
of stereotype-inconsistent
information
despite
cognitive
load
(e.9.,
on the central
executive)
Fosteri
n
g persistence of goal-directed
actions
Supporting
the
regular intake
of
vitamin tablets and
essential
medication
Helping
challenged
patient
groups
to
perform
difficult everyday
actions
(e.g.,
drug addicts
under
withdrawal to
write a
CV)
Fostering
engagement
in
physical
exercise
(e.g.,
after
hip
replacement
surgery)

Motivation
and
Volition
in
the Course of A'ction 293
-.
t,r 7.3 lmplementation inientions s implementation intentions do not lead to behavioral
lio'*""o"r"o
Errects äTliTjh;ä.t-LlfiDression
orprejudice
orinperror-
,,,v?sner
(1994) observed that consci.ous attempts to control x implementadonintentionsdonotleadtoego-depletion
^..r"ooorur. one's thoughts - e.g', "I will not think about pink (e.g.,
performance levels are not reduced when emotions
"*ofr*"," - lead to rebound effects in the sense that the are controlled. by means of implementation intentions),
tiroughts
controileci
become more readüy accessible
and thus and
-or. tit.ty to surface in subsequent thoughts and behav- * implementation intentions d.o not lead to rebound
ior. participants in his st:dies set themselves suppression effects(e.g.,whenstereotypicatthoughtsaresuppressed).
;ar oi trri,s
trna ana
*":".:lT:..d ,: nlc_" beil when-_
iuCr tliei,- *,iou ghts'turned-i:*-the Proscdbeö-dir-€en:ön-
?ar- -
6.ip*rc with the goal of not thinking about pink elephants
trcrv@r'"
"--- " ..o*"i.o trI"^ -;:..:_^-',,: _:-: 11.8 Discussion and Future Perspectives
,n
tuf4r succeeded
in suppressing
ttrese
thoughts.
However,
findings from a second phase of tJre experiment, in which
oarticipants engaged in free association and wrote down all
äf th"ir tho.tghts, showed that participants who had resolved
not to think ab out pink elephants in the fu st part of the exp er-
jinentwere now considerably more likely to report ttroughts
relating to pink elephar}ts thal participants who had not
set a suppression goal. This is effect is termed the rebound
effect:
O The
rebound
effect
involves
a marked
increase
in certain
thoughts
following
the "eKinction"
of a goal
to suppress
or inhibit
those
thoughts'
Against the background of these research findings, it would
seem reasonable to hypothesize that suppression-oriented
implementation intentions may inhibit unwanted thoughts
and
feelings to beginwith, but that these suppressedthoughts
or feeiings resurface later, i.e., that rebound effects occur.
Gollwitzer et al. (2004) conducted two experiments to test
this hypothesis. The participants in these studies were first
asked to suppress stereotypical thoughts about a carefirlty
described homeless person in an impression formation
task. Rebound was measured either in terms of subsequent
erpression of stereot!4)es in a questionnaire tapping partic-
ipants' evaluation of homeless peopie in general (Gollwitzer
et at., 2004, Study 1) or in a lexical decision task assessing
the cognitive accessibility of stereotypical contents regardtrg
homeless people (Gollwitzer et a1.,
2004,
Study 2). It emerged
that the participants who.had only set themselves the goal
of suppressing stereotypical thoughts when forrn-ing a11
impression of the homeless person experiencedpronognced
rebound effects in both studies' showing more stereot],'pi-
cal judgments of homeless people in general (Study 1) and
a higher accessibility of homeless stereotypes (Study 2). par-
ticipants who had furnished this goal intention with a cor-
responding implementation intention did not experience
rebound effects.
SUMMARY
Findings on dre potential costs
of implementation intentions
canbe summarüed as follolvs:
t1-8.t lmplementation Intentions:
A Foolproof Self-Regulatory Strategy?
Although implementation intentions seem to function effec-
tivelywithout significant costs in terms of behavioral rigidity,
ego-depletion, or rebound, they do not always result in the
desired
outcome. First, the behaviorspecifiedin the then-part
of an implementation intention may be beyond the person's
control. For example, somebodywho intends to eat healthily
may plan to order vegetarian food, but then find themselves
in a restaurant with no vegetarian options. Second, it makes
no sense to specify situations that barely, if ever, occur in the
if-part of implementation intentions. For example, it would
be pointless for someone to plan to eat healthily by ordering
vegetadan food the next time they go to a good restaurant if
they usually eat in ca-feterias
or at home. Ttrird, the behaviors
specifled in the then-part of the implementation intention
mäy not be instrumentai to reaching the goal. For example,
someone who plans to eat healthily may order a vegetarian
meal in a restaurant, not knowing that the dish chosen is ft:ll
offatty cheese.
t!.4.2 Prospective Memory and Neuronal Substrates
In the past 20 years, implementation intention research
has focused on motivational and volitional processes and
their effects on impression formation and behavior. In the
coming years, the focus should be shifted to cognitive
and neuroscientific aspects.
From the cognitive perspective,
implementation intention research stands to benefit from
prospective memory research (cf-
Smith, 2003),
which exam-
ines the processes by which intentions are stored in and
retrieved from long-term memory as well as from ongoing
attempts to examine the different components of working
memory (e.g.,
the central executive, the phonological loop,
andthe episodicbuffer
as
proposedbyBaddeley,
1986; Badde-
le-v, 2000) and their functions in the realization of goal inten-
tions and implementation intentions (Achtziger et aL., forth-
coming). From the neuroscientiflc perspective, researchers
have already used magnetic encephalography to exam-
ine neuronal activity in the deüberative and irriplemental

2g4 A.
Achüiger
and P M. Gollwitzer '--l.tt N
miadsets, and found that the implemental miadset is associ- but aspüe to goals that help them realize their fir1l poten- i
ated with higher posterior gamma activity than the delibera- tial. Future research shouid exami.ne
the postactional phase I
tivemindset.Theseflndingsindicatethatmoreintensiveand in which completed goal strivings are evaluated, and seek i
complex brain activity is involved in plaaning the implemen- to identify self-regulatory strategies that are conducive to a i
tation of a goal in terms of when, where, and how to per- person's goal striving in subsequent endeavors. The r:ltimate I
form a specific goal-directed action than in weighing up the goal of this research is to develop intervention programs that i
positive ald negative consequences of a potential course of will provide individuals with action control strategies that I
actionandtheprobabiJitythattheseconsequenceswilloccur. enable them to address the problems that beset goal striv- :
Moreover, the kind of brain activity generated by the imple- ing in the different action phases
of the Rubicon model more ;
menta-l mindset seems to be associated with preparation of successfull.v. t
actions (Achtziger,
Rockstroh, Oettingen, & Gollwitzer,2003). i
Research has also found the control of negadve emorions REVTEW OUESTIONS i
(e.g., arxiety Schweiger Galio, Keil, Mc Culloch, Rocksuoh i
& Gollwitzer, forthcoming) and automatic stereotype acti- 1. Which four phases are distinguished in the Rubicon i
vation (Achtziger, Moratti, Jaudas, Rockstroh, & Gollwitzer, modeiof actionphases.z i
forthcoming)bymeansofimplementationintentionsvs.goal The predecisional, preaclional, actional, and postac- i'
. ,.
..:
intentions to involve different electroencephalogram @EG) tional phase.
responses. Generaily speaking, howeve! ttrere is still much to i '
belearnedabouttheneuronalsubstratesofactioncontrolby 2. AttheendofwhichphaseofttreRubiconmodeldoesthe I
means of goa_l
intentions vs. implementation intentions, anä individual "cross the Rubicon" by committing to a goal , | ,'
indeed about intentional states
in general. intention? ,, '1.
At the end of the predecisional phase. . ,,:,t, ltij
SUMMARY - -L^ ^^---^- -!^^t'^- L- 3. whateffectsdothedeliberativevs.implementalmind- .. .l:i
: il
The study of motivation in the cor:rse of action has made it sets have on self_evaluation? |
possible
to distinguish phenomena of goal
setting (motiva- _----
-- _-
-- -
-- - " ----- ,,::;;:!-:,.r
''
tion) from phenomena of goal strivnj (voLition). whereas Studieshaveshownthatanimplementalmindsetisasso- . 'f-'r ;
research to date has focused on the cognitive orienta- ciated with more positive self-evaluations than a delib- , i ,,,
tions associated with the respective action phases (mind- erativemindset' -' -= f':-- '
set research), the aim of future research will be to identify 4. How are the implemental and deliberative mindsets 1,
,
self-regulatory strategies thai facilitate effective accomplish- experimentally manipulated? ..i- , .
ment of the tasks necessary at each phase
in the course of TL __^ ^_
-:*"- "- Therearerwomernoosormoucrngeacnrrunoser: i
action. The theory of intentional action control (Gollwiuer, r__r ^_ _ , I
' IlnDlementatmmoset: r.varnclDantsareasKeorocnoose I
i993, 1999) hastakenfuststepsiathisdirection,showinghow .^:.-^^_ |
oetween atternatrves, r.e., to maKe a oec$ron; z. varrJc- I
implementation intentions can facilitate the performance of !
rDanrs are asKeo to pran Ine steps requrreo to uansrare
taskst}ratnecessitatetheinitiationofgoaI.directedbehavior,
the shierding of rhat behavior asainst disracrions, ;" dm;tr i^*::::i?:::l:::o"" specifving when' where' and i
I
the shielding of that behavior against distractions, the timely , ^:-.- -^ :, i
nowro raKe
eacn srep- I
termination of
goalstriving,
andmeasuresto ensurethatthe n^r:L^-^#-.^_,-r^^i., D^-+i^i-^-. I
capacity for acrion contror is not overstr.t.h"d d;;;;;; 3-:}'"^'io* -indset I' Participants are
interrupted dur- : I
g me decision-making process; 2. Participarnts
weigh I
DllrvulB'' r +^,_^ ^ the positive and negative
short- and long-term conse-
Futureresearchshouldtakeatwo-prongedapproach.On rkeachangedecisiön.,-,. ,., i , ,
the one hand, it shor:ld seek to identify further self-regulatory quences or maKrng or laung to mi l,
strategies that help to address these kinds of difficulties and 5. What effects do the deliberative vs. implemental mind- I
thereby help people to attain theü goais; on the other hand, ' sets have on information processing? :
the search for effective self-regulatory strategies shor:Id be Individuals in the deliberative mindset generaJly
engage :: .
extended to other action phases.
The predecisional phase of in more ,,deliberativ_e]
tho-gghts,
4:g a-bie to recal delit:_-_ :_=_+-; _
goa1settinghasa1readybeenexamined.Fantasyrealization*"d*t}'oughtsbetterthanimplementaIthoughts,aldi
theory (Oettingen, 1996, 2000) distinguishes three different
goal setting stategies (mental contrasting of desired future
and actual presen! indulging in positive fantasies about the
future, and <iwelling on negaiive aspects of the preseni), and
has found that only meniaL contrasting guarantees that rtre
goals
people set are in iine with their perceived expeciations
of success. ln other lr.ords, menial contrasting ensures that
people do not pursue goals rhat are excessivei-v
high or lor^/,
tend to be open-minded (i.e.,
to process information in an
objective and unbiased manner); moreove! their atten-
tion is not centrally focused. The opposite effects are
observed for individuals in the implemental mindset.
6. After induction of rrvhich rnindset are goals more likeiy
to be attained?
After induction of the impiemental mindset.

--s€
Motivation
and
Volition in the Course of
Action
7. What arethe effects of adeliberative rrrinjr"toop"opl.'s
evaluations of their romantic relationships?
It depends on the person's comnitment to ttre relation-
ship. If commiüneni is high, the partner is rated more
positively after induction of a deliberative rnindset than
after induction of an implemental mindset; if cornmit-
. ment is iow the effects are reversed.
B. What is a "goal intentionJ'?
Goal intentions speciff desired end states that peopie wish
9. What is an "implementation'intention"?
t' l*pl"mentation intentions are "if-then" statements that
specifythe conditions underwhich goal-directed behav-
ior is to be initiated.
10. lvhat function do implementation intentions serve?
:." '- Implementation intentions facilitate the enacbTrent of
goal intentions that are particuiariy dif0cuit to attain.
11. Which factors moderate the effects of implementation
intentions?
' The foitowing moderatorvariabies have beeqidentiied:
. rlifnculty of the goal intention,
.: cornrnitment to the goal intention,
:il"'-'' '
295
co ramitme nt to the irnp lementaiio n intenLion'
degree of activadon ofthe goal intendon.
12. Are cop.itive resourcesrequiredto putirnplementatiorl
intentions into practice?
Irnplementation intentions are initiated automatically
and thus do not tax cognitive resources.
13. What positive effects can implenentation intentions
have on health-related behavior?
Exampi6e reguiar intake ofvitamin tabiets; parriciparion
in calcer screening; regular exercise after hip replace-
ment sugery.
14. How can implementation intentioas inhibit unwanted
effects, such as stereotypicalviews ofothers?
Unwanted behavior can be inhitited by forming an
implementation intention that inhibits either its activa-
tion or its application. The if part of ttle implementation
intention shor:ld specify a situation or a stimu-lus that is
likelyto trigger activation or application ofthe stereotype;
the then part should specify a goal-directed behavior
with the potentialto inhibitthe stereotype (e'g.,
byinitiat-
ing or upholding individualized processes of impression
formation).
ll
ll
.l
.,1