Upload
roxy-shira-adi
View
221
Download
0
Embed Size (px)
Citation preview
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
1/12
aleE Zand
This paper presents
a
model of trust and its interaction with information flow,
influence, and control, and reports on an experiment based on the model to test
several hypotheses about problem-solving effectiveness. The subjects were man-
agers and the independent variable was the individual manager s initial level of
trust. G roups of business executives were given identical factual information about
a difficult manufacturing-marketing policy problem; half the groups were briefed
to expect trusting behavior, the other half
to
expect untrusting behavior. There
were highly significant differences in effectiveness between the high-trust groups
and the low-trust groups
in
the clarification
of
goals, the reality
of
information
exchanged, the scope of search for solutions, and the comm itment of managers to
implement solutions. The findings indicate that shared trust or lack of trust ap-
parently are a significant determinant of m anagerial problem-solving effectiveness.^
There is increasing researeh eyidenee that INTRODU TION
is a salient faetor inde termining the j ^ ^ . (^gg^ f ^ ^ j ^ ^ effeetiye
ofman y relationships, such as helping relationship, one participa nt (coun-
andchild (Baldwm selor, therapist, help er) behaye d in ways tha
al, 1945), psychotherapist and elient deyelop ed trust and the other experienc ed an
53; Seeman , 1954 ), and mem bers i^^.^ease in trust, and concluded tha t the de -
and ^elopmentoftrust is a crucia l initial factor
.^^^^
neeessary co ntinuing eleme nt
in
such
ofexpression, ^relationship. H e summ arized extensiye re
istrust eyokes interperson al rejec- .^^^ch in which an increase in trust app eare d
on and arouses defensive beha vior (G ibb , ^^ ^^ eausalK- related to mo re rap id intellec-
^
tual deyelopment, increased originality,
in-
Du ring the past fifteen years ma n)' man - c e a se d em otional stabilit)% increased' seH-
to
programs,
^oj^^j.^
^nd deereased physiological arousa
ously called sen sitivity tra ining (Bra d- ^o defend against threa t.
et al, 1964 ), grid laboratories (Blake j h ^ ig,,gi f^^^.tin arelatio nsh ip affeets
or
group workshops
he
degree
of
defensiveness. Gibb (1961 )
n and Bennis, 1965 ), to improve their f^^nd tha t m em bers of small groups tha
indeveloping trust and thus, presuma- developed a "defensive elimate," had diffi
It has ^ulty eoncentrating onmessages, perceiv ed
to show a direet j e mo tives, values, and emotionsof other
and managerial i^^s aeeurately, and increased the distortion
in a wo rking organization ^f messages. Oth er studies suggest tha t some
and Campbell , 1968;House, interpersonal trust is required for effeetiy
is aneedtoclarify the problem solying in a group. Parloffand
for assertions abou t trust Ha ndlo n (1966 ) found tha t intensiye, pe
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
2/12
230 ADMINISTRATIVE SCIENCE QUARTERLY
in problem-solving effectiveness. They found information he provides will be low in ac
that groups penalized for poor ideas and ad- racy, comprehensiveness, and timeliness; a
monished to produce only good ideas while therefore have low congruence with real
working on early problems produced poorer He will also resist or deflect the attempts
solutions to later problems when these restric- others to exert influence. He will be sus
tions were removed than groups that were cious of their views, and not receptive
not penahzed and admonished during their their proposals of goals, their suggestions
earlv problem assignments. reaching goals, and their definition of crite
This paper: (1) analyzes the concept of and methods for evaluating progress.
trust, (2) presents a model of the interaction though he rejects the influence of others,
of trust and problem-solving behavior, and will expect them to accept his views. Fma
(3) reports the results of an experiment that one who does not trust will try to mmm
attempted to test several hvpotheses derived his dependence on others. He will feel
from the model. cannot rely on them to abide by agreeme
and will try to impose controls on their
N LYSIS OF C O N C E P T
havior when coordination is necessary
Trusting behavior, following Deutsch attain common goals, but will resist and
(1962), is defined here as consisting of alarmed at their attempts to control
actions that (a) increase one s vulnerability, behavior.
(b) to another whose behavior is not under When others encounter low-trust behav
one s control, (c) in a situation in which the initially they will hesitate to reveal mtor
penaltv (disutilitv) one suffers if the other tion, reject influence, and evade control. T
abuses that vulnerabihty is greater than the short cycle feedback will reinforce the o
benefit (utility) one gains if the other does inator s low trust, and unless there
not abuse that vulnerability. For example, a changes in behavior, the relationship
parent is exhibiting trusting behavior in stabiHze at a low level of trust,
hiring a babv sitter so he can see a movie. All of this behavior, following from a
The action significantly increases his vulner- of trust, will be deleterious to informa
abilit^ since he cannot control the baby exchange, to reciprocity of influenc>e an
sitte/s behavior after leaving the house. If the exercise of self-control, and w ill dimm
the baby sitter abuses that vulnerability, the the effectiveness of joint problem-sol
penalt\^ mav be a tragedy that may adversely efforts.
affect the rest of his Hfe; if the baby sitter To the objective uncertainty mherent
does not abuse that vulnerability, the benefit problem, for example, unavailable facts
will be the pleasure of seeing a movie. Thus unknown causal relationships between
trust as the term wiU be used in this paper, tions and results, low trust will add s
is not a global feeling of warmth or affection, uncertainty- that
is
uncertainty^ introduce
but the conscious regulation of one s dcpen- individuals wdthholding or distorting rele
dence on another that will vary with the task, information and concepts,
the situation, and the other person. Persons lacking trust attempting to so
problem jointly will attempt to minim
MO EL their vulnerability. There will be an incr
The following model, based on Cibb, in the likelihood of misunderstanding or
(1964) conceptualizes the transforming of interpretation. The social uncertainty
one s inner state of trust (or mistrust) into duced by their low trust will increase
behavior that is trusting (or mistrusting) probability that underlying problems ma
through (1) information, (2) influence, and undetected or be avoided, and that inap
(3) control. priate solutions may be more difficu
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
3/12
and PROBLEM SOLVING
231
, comprehensive, aecu rate, and timely dance w ith agreem ents, an d therefore wdl
less fear th at the ir exposure yyill be
have less need to impose controls on others
(see Figure 1). Consequently they will con-
tribute to a decrease in social uncertainty
and be less likely to misinterpret the inten-
tions and the beha\dor of others. As a result
endence because of confidenee tha t underlying problems are mo re likely to be
s will control their behavior in accor- identified and examined, and solutions more
TRUST
Increase one s
vulnerabilit)
to others whose
behavior one
cannot control
\
CONTROL
Accept more inter-
dependence with others.
Impose less procedure
to c>ontrol others.
Greater confidence
others will do what
they agree to.
Greater commitment to
do what one agreed to.
INFORMATION
Disclose more
accm-ate, rele-
vant, and
complete data
about the prob-
lem, one s
thoughts and
one s feelings.
INFLUENCE
Accept more influence
from others in
selection of goals,
choice of methods,
evaluation of progress. 1, Initial level of
trust predisposes
information flow,
influence, and
control.
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
4/12
ADMINISTRATIVE SCIENCE QUARTERLY
likely to be appropriate, creative, and long-
range.
Hypotheses
It
is not assumed here that trust alone will
solve a technical problem; it is assumed that
group members collectively have adequate
knowledge, experience, and creativity to de-
fine and solve a complex problem. It is also
assumed that it is possible to increase or de-
crease trust in members of a problem-solving
group.
On the basis of the model described, the
following differences can be predicted in the
problem-solving behavior of groups with
high and low trust.
An increase in trust will increase the ex-
change of accurate, comprehensive, and
timely information. Problem-solving groups
with high trust will:
Hypothesis 1 Exchange relevant ideas and
feelings more openly.
Hypothesis 2
Develop greater clarification
of goals and problems.
Hypothesis 3
Search more extensively
alternative courses of acti
HypotJiesis 4 Have greater influence
solutions.
Finally, an increase in trust will incre
willingness to control one's own behavi
will increase confidence in the reliability
others, and will decrease efforts to control
behavior of others, all of which will c
tribute to increased satisfaction and moti
tion. Hence, problem-solving groups w
high trust will:
Hypothesis 5
Be more satisfied with th
problem-solving efforts.
Hypothesis 6 Have greater motivation
implement conclusions.
Hypothesis 7 See themselves as closer a
more of a team.
Hypothesis 8 Ha ve less desire to leave th
group to join another.
ynamics of Trust
Trust takes form in the interaction of t
(or more) people, and the dynamics of t
interaction is illustrated in Figure 2.
P's intention s &
expectations
Not
trusting
O
iintrustwortliv
P's belmvior
Restricts information
Resists
influence
Seeks
to impose controls
P's conclusion
O\s behavior confiiTns
P's expectation and
justifiesP s lack of trust
P s
perceptir
O s behavior seen as
untiTJSting
O's perception
P's b(>havior seen as
untrusting
O's conclusion
P's behavior confirms
O's expectations and
justifies O's lack of tiust
O's behavior
Restricts information
Resists
influence
Seeks
to impose controls
O's intentions 6c
expectations
Not trusting
P
untnistwortlu '
F I G U R E 2 . A M O D E L O F T H E I N T E R A C T I O N O F T W O P E R S O N S W I T H S I M I L A R I N T E N T I O N S A N D
P E C T A T I O N S REGARDING TRUST
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
5/12
and PROBLEM SOLVINC 233
t to control O. (3 ) Assume O also laeks trust, but this study did not focus on a tes
(4 ) pereeiyes P s initial beh ayio r as of the spiral reinforcement mo del. Th e aim o
ly untrusting, an d (5 ) concludes he this study was to examine the relation be
righ t to expeet P to be untru stw orth y; twee n trus t and problem-solying effectiyenes
6 ) he will feel justified in his mistrust as formu lated in the eigh t hypothese s,
f P. Sinee (7 ) P sees O s beh avior as un - To test the hypotheses deriyed from the
he (8 ) will b e confirmed in h is mo del, the research was designed so th at half
tial expectation that O wo uld not be trust- of the experime ntal groups started work on a
and (2) P will beh aye with less trust business-managem ent problem with a me nta
wh en he entered . set toward low trust and half with a men ta
The interaetion will eontinue around the set towa rd high trust. M ental set, as used
indu cing O an d P to beh ay e with less here , ineludes intentions as to one s beh ayior
ess trust un til they arriye at an e quilib- expectations as to the beha yior of others, feel
um leyel of low tru st, eaeh atte m ptin g to ings such as anxiety^ or diseomfort, and cogni
yuln erability and to maximize tiye frame used to interp ret eyents and form
s control of the other. In the proeess the pereep tions. In this researeh trust was no
eetiyeness of pro blem solying will dec rease , exam ined as a personalit)^ ya riab le; that is
er interaetion has continu ed, each will an element of indiyidu al cha racter, bu t as an
nd to hold more firmly to his entering indueed attitude, one that the indiyidua
iefs. Th ey will no t hay e a reliable basis for eould alter in a situation in which he w as led
or sharing inffuenee, an d the mu- to intend and to expect trust (or m istru st)
resistan ce to inffuenee will arouse feel- from others as he atte m pte d to solye a
s of frustration in both . If they hay e a problem . Beeause trust as a person ality trai
hne , each wdll attem pt to impose con- was not releyant, and to ayoid alerting th e
ols on the other. If P is O s organ izational subjects to the issue of trust, no prior me a
or, he may com ma nd O s com pliance, sures we re taken of the subjects attitu de
will reinforce O s mistrust. Usually, by tow ard trust.
Subjects
Because of their high potentia
ll be lower than the initial leyel. for top man ageme nt, upp er-midd le m anagers
Cib b (1964 ) offers supp ort for the dyn a- from all functions and pro du ct diyisions of a
cs of this interaetion . In obserying small large, internation al electronics com pany we r
beh ayior he noted th at the defensiye periodically seleeted by their superiors, afte
f a listener genera ted cues whieh discussion with the corpo rate person nel staff
ntly inereased t he defensiveness of to attend an off-site, four week, in-residence
e comm unicator, resulting, if unehecked, in program in man agem ent developm ent given
circular pa ttern of esealating defensiyeness. seyeral times a year tha t accep ted sixteen
The p atte rn of spiral reinforcemen t illus- m anage rs at a time. Eig ht m anag ers in each
ed in Figure 2 wou ld operate eonstrue- program, were randomly chosen to be sub
yely if it is assumed tha t bo th P and O en- jeets and distribu ted into two problem
the relationsh ip with trust in the other, solying groups, each with four m em bers. Th
b (1964) obseryed that wh en defensiye- remaining eight m anagers were obseryers
ss was reduce d, memb ers were better able eaeh was randomly assigned to a group with
o eoncen trate on the content and me aning of four m embers and each group ob seryed on
message, beeam e more problem o riented, problem-solying group. Da ta were gathered
w ere less eonc erned abo ut imposing con- in eight program s proyid ing a total of sixty
ols on each other s behay ior. four subjeets in sixteen prob lem-solying
groups and fifty-nine obseryers (fiye pro
M THO
grams were short one manager) in sixtee
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
6/12
234 ADMINISTRATIVE SCIENCE QUARTERLY
ment, which was designed asalearning event to resign. Furth erm ore, he wastold t
embeddedinthe program . expansion wasnot feasible becauseitwo
Problem Thecentral problems involved redu ce short-term profits, take more tha
(1) developingastrategy toincrease short- yeartobuildandstartupa new plant,a
term profits without undermining long-term theboard wasnot likely toapproveth
growth
of a
medium-sized electronics com- nancing,
so as a
first step toward increas
pany with very lowreturn on inve stme nt, profits,he would haveto announce his d
ou tda ted ma nufa cturing facilities, w hose sion against expansion. Thevice preside
labor force hadbeen cut 25percent and ha d no know ledge of the president's dilem
whosetopmanagem ent personnelhadbeen when they started their thirty-minute pr
changed andreorganized twoyears before, lem-solving m eeting.
and 2) obtaining commitment to implement
Induction
of
conditions
of
trust
The ind
such a program despite strong man agerial tion of thetwolevels of trust wasacc
disappointment because expectations ofim- plishedbyoperatingonthe following en
mediate investment forexpansion andmod- ing beliefs of subjects: 1) the
emization would not be met. The situation,a competence of others, 2) norms onin
variation
of
one
described
by
Maier
et al
ducing information and new ideas,
3)
no
(1959),
involved four executive roles: presi- onattem ptstoinfluence ma nag ers outsid
dent and vice presidentsforma rketing , man - one's prim ary responsibility, 4) likelih
ufacturing , and personn el. Subjects we re ran- tha t others w ould abuse trusting behav
domly assigned totheroles. and 5)competitivenessorcollaborativen
Procedure All subjects and observers were forrewards.
givenawTitten descriptionofthe production. Inahigh-trust group,a manager's ente
marketing, financial,
and
perso nne l difficul- m en tal
set
toward trust w as shaped
by
tiesofthe company. following para grap h, which followed
In thepresenceoftheobservers, subjects factual information inthe role statemen
were told they were
to
conduct
a
m ee ting you have learned from your experiences du
lasting thirty minutesmthe president
oihce ^^e past two years that you can trust tbe
to mak e app ropriate ma nagem ent decisions, membersoftop management. You and the o
Ostensibly, they w^ereto dem onstrate their topmanagers openly express your differe
decision-making competence tothe ir fellow andyour feelings ofencouragement or of
managers,theobservers. appointment. You andtheothers shareall
Each subject was then givenanad di tiona l vant information andfreely explore ideas
w ritten statem ent with factual and attitud inal feelings that maybe in oroutof your def
information relevanttohis function. He had responsibil^y. The resulthasbeena high l
1 1 J r .1 1 r . of give and take and mutual coniidence m
no knowledgeortherole lnrorm ation given , , _ j V-T,.
4U
u- .. T-u
u ^
i. 1 J
others support
and
ability,
to other subjects. The subjects privately read
^
and absorbed this problem information
for
Subjects
in
low-trust groupshad
a
sim
twenty-five minutessothere wouldbemini- paragraph in their role information,
mal needtorefertoitduring the meeting. worded to induceadecreaseintrust.
Treatments Subjects were randomly as- The reward system wasoperated o
signed tooneoftwogroup conditions: an information placed only in the presid
entering mental set toward highorlow trust, statement. In thehigh-trust condition,
The factual data about production, market- president wasled to see hisrelation t
ing, finance,and so onwas identicalinboth vice presidents as collaborative. His
conditions,and allvice presidents we re led statem ent said that although the Bo
to expect tha t the president wou ld annou nce decision considered
you
specifically, s
approvalof a long-stu died plan t expansion, y^ou app ointed thecurrent topmanagem
In both conditions the presiden t's state- team it is likely that the Board willgo
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
7/12
Zand:PROBLEM SOLVING 235
as potentially competitive. His role in Table 1, with the chi-square value for each
said that since the board's ultima- item.
m pertained to him, it was possible that The subjects' rating of level of trust con
might appoint one of the vice presidents firms that the induction of high and low trus
his successor. The vice presidents in both was successful (
p
< .001) after one-half hou
ditions were given no information about of problem discussion. This result, although
on to the president was not a direet test of the spiral reinforcemen
competitive or collaborative. model, does offer support for it.
All subjeets were told that "whenever in- Since the observers used only their per
ation is incomplete, introduce whatever sonal standards for their ratings, it is note
ets and experiences seem reasonable under worthy that they had little difficulty recog
circumstances." nizing the behavior indicative of low or high
Observers.
In addition to reading the writ- ^
P ^
, ,. ,
general description of the company's ^he hypotheses about differences between
s, before observing the problem- groups with high or low trust were confinned
the
observers were told
of
^>\*^^f ,^P ^ t
ft^^ ^^
o^^ ^'
m^
for seeking P
7/26/2019 1972_ Zand_Trust and Managerial Problem Solving
8/12
6
ADMINISTRATIVE SCIENCE QUARTERLY
TA BL E 1. FREQ UEN CY OF RESPONSE TO EACH ITE M BY SUBJECTS AND OBSERVERS UNDER
HIGH TRUST AND LOW TRUST WITH CHI-SQUARE VALUES FOR DIFFERENCES
1.
2 .
3 .
4.
5.
6.
Y_
8.
9.
Item
Trust
Openness about feelings
Clarification of
problems and goals
Search
Influence
Satisfaction
Motivation to imple-
ment conclusions
Closeness as a team
Desire to take a job
in another company
* p < .001 for all X values
p < .001 for all X- values
Subjects
Condition
High
Response trust
Much
Little
Much
Little
Much
Little
Much
Little
Much
Little
Much
Little
Much
Little
Much
Little
Much
Little
in this
in this
30
2
31
1
24
8
21
11
29
3
28
4
30
2
27
5
8
24
Low
trust
9
23
15
17
10
22
6
26
6
26
2 5
10
22
9
23
22
10
column,
column except item 2 for
28.1
26.4
10.8
13.8
30.8
25.6
26.6
19.1
10.8
which p