Você está na página 1de 7

ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES

Vol. 67, No. 2, August, pp. 222–228, 1996


ARTICLE NO. 0075

The Effects of Time Constraints on the Prechoice


Screening of Decision Options
LEHMAN BENSON III AND LEE ROY BEACH
University of Arizona, Tucson

consistent: decision makers speed up execution of their


The purpose of this research was to see if the effects decision strategies or switch to simpler strategies, some-
of time constraints on screening are similar to those times speeding up after having switched (Eland & Sven-
that have been reported for choice: switching to sim- son, 1993; Maule & Mackie, 1990; Payne, Bettman &
pler decision strategies and/or speeding up strategy
Johnson, 1988; Svenson & Benson, 1993a, 1993b; Sven-
execution. The first of two studies found that under
son, Edland, & Slovic, 1990; Svenson & Maule, 1993;
time constraints, subjects did not switch strategies but
that some subjects increased the speed of execution Smith, Mitchell & Beach, 1982; Wright, 1974; Zakay,
of their existing strategy while others became more 1985).
inconsistent in executing their existing strategy. The Because choice can be made using either kind of
second study found that instructions that stressed the strategy, switching frequently entails rejecting a com-
importance of the task led more subjects to speed up pensatory choice strategy for a simpler noncompensa-
and reduced the number who were inconsistent. tory choice strategy. In contrast, research indicates
Again, there was no evidence of use of a simpler ver- that screening is accomplished using one generic non-
sion of the strategy. The results show that, like choice, compensatory screening strategy called the compatibil-
screening under time constraints can lead to speeding ity test (reviewed in Beach, 1993). Therefore, time con-
up of execution, particularly if the screening decision
strained subjects have four ways of approaching a
is important. However, switching to a simpler strategy,
screening task: (1) They can proceed with execution of
which occurs in choice, does not appear to occur in
screening. Instead, subjects appear to become less ex- the compatibility test as though there were no con-
acting in the execution of their existing screening straints, which should result in a high rate of failures
strategy, which results in inconsistencies in accep- to complete the task. (2) They can speed up execution
tance and rejection of options. q 1996 Academic Press, Inc. of the compatibility test, which should result in a
higher rate of completion of the task. (3) They can resist
speeding up but execute the compatibility test less
Most decisions are made under time constraints im- carefully, which should result in inconsistencies— re-
posed by the pace of ongoing events, by explicit dead- jection of options that otherwise would be accepted and
lines, or by others becoming impatient with the deci- vice versa. (4) They can resist speeding up but execute
sion maker’s indecisiveness. Previous research has fo- an abbreviated version of the compatibility test, which
cused upon the effects of time constraints on choice. should result in completion of the task and consistent
Assuming the two-stage decision procedure described patterns of rejection and acceptance, but not the pat-
by Image Theory (Beach, 1990; Beach & Mitchell, 1990, terns called for by the unabbreviated version of the
see also Payne, 1978), in which screening serves to compatibility test. The research problem is to differen-
weed out wholly unacceptable options and choice serves tiate between inconsistent application and abbreviated
to select the best from among the survivors of screen- application of the compatibility test, and to differenti-
ing, the present research seeks to see if the previously ate this from the effects of increased speed of execution.
obtained results of time constraints on choice also are
obtained for screening. SCREENING
The effects of time constraints on choice are remarkably
The most systematic research on screening has been
done in the context of Image Theory (Beach, 1990,
Address correspondence and reprint requests to Professor Lehman
Benson III, Department of Management and Policy, College of Busi-
1993; Beach & Mitchell, 1990). The results indicate
ness and Public Administration, University of Arizona, Tucson, AZ that it is the failure of options to meet decision makers’
85721. standards, rather than the failure to offer desired out-

0749-5978/96 $18.00 222


Copyright q 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


TIME CONSTRAINTS 223

comes, that leads to their rejection. With such options and rejections might look inconsistent, it could be ra-
out of the way, the decision maker can then move on tionalized by presuming an abbreviated version of Eq.
to select the best option from among the survivors of (1). In light of previous research (Benson & Svenson,
screening. 1992), we would expect abbreviation to consist of at-
tending to only the most important features or, per-
The Compatibility Test haps, of attending to only the first features about which
information is received.
Screening involves serial consideration of available Methodologically, differentiating between inconsis-
options in which each option’s features are compared tencies and abbreviation consists of a subject-by-sub-
to a relevant set of standards. Failures of the option’s ject examination of the pattern of retentions and rejec-
features to meet the standards are called violations, tions. Patterns that show a clear threshold are counted
and the larger the number of violations the less compat- as full execution of the strategy. Patterns that show no
ible the option is with the decision maker’s standards clear threshold and that cannot be accounted for by
and the less acceptable it is as a possible choice. At iterative applications of increasingly abbreviated ver-
some level of incompatibility the option is rejected, sions of the strategy are counted as inconsistent execu-
called the rejection threshold. If only one option sur- tion of the strategy. Patterns that can be accounted for
vives screening, it becomes the decided-upon option by an abbreviated version of the strategy are counted
and choice does not take place. If two or more options as an abbreviated execution of the strategy.
survive screening, the decision maker undertakes to
choose the best from among them. Speeding Up
The research supports the following equation as de-
scriptive of the compatibility test for screening: This still leaves the question of how to detect speed-
n m ing up of strategy execution. Following the lead of Sven-
I Å ∑ ∑ Wc Vtc , (1) son and Benson (1993b), the time constraint used in
tÅ1 cÅ1 this experiment was determined by a pretest to see
how long subjects took to perform the screening task
in which incompatibility, I, is zero when an option has
when there was no time constraint. Because the distri-
no violations and decreases (i.e., is more and more neg-
bution of times was normal, the time constraint for
ative) as the number of violations increases; W is the
the experiment was then set at one standard deviation
importance weight for each of the relevant decision
below the pretest subjects’ mean time. This permits us
standards; V is a violation of a relevant standard c by
to detect speeding up, at the group level, by examining
feature t of the candidate (Vtc Å 01 or 0).
how many subjects in the time constrained experimen-
The pattern of option rejection and acceptance dic-
tal group actually finished the task within the permit-
tated by this strategy, and the pattern found in all of
ted time. If they executed the strategy at the same
the research thus far, consists of the subject retaining
pace as the nonconstrained pretest subjects, we would
all options that have an I less than some critical value
expect about 84% of the subjects who do not resort to
and rejecting all options that have an I greater than
abbreviated execution of the strategy to fail to finish
that value. (Note that I becomes increasingly negative
the task. Fewer failures to complete would imply that
as violations increase; for ease of communication we
at least some of the subjects increased execution speed
talk of large negative numbers as being larger than
in order to compensate for the reduced time for doing
small negative numbers). The critical value of I is the
the task.
rejection threshold, and while it differs from one sub-
ject to another, it has been found to be stable for each
individual subject over the duration of an experiment. RESEARCH STRATEGY

Switching Subjects were instructed to screen a set of options


(potential jobs) in anticipation of choosing one job from
Recall that switching consists of inconsistent or ab- among the survivors of screening. One group screened
breviated execution of the compatibility test. Inconsis- the jobs without time constraints and another group
tency would be revealed by the absence of a clear rejec- did so with time constraints. The hypothesis in Experi-
tion threshold, i.e., no single value of I below which all ment 1 was that more of the time-constrained subjects
options are retained and above which all options are would resort to inconsistent or abbreviated execution
rejected. Abbreviation would be revealed by decision of the compatibility test than would the nonconstrained
makers examining fewer features of the options. This subjects. In addition, if time-constrained subjects speed
would mean that while a subject’s pattern of retentions up execution of the compatibility test we would expect

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


224 BENSON AND BEACH

fewer than 84% of those who did not resort to abbrevi- in terms of their own perceptions about how important
ated execution of the test to fail to complete the task each feature ought to be in screening jobs.
in the prescribed time. Each job was described on a single sheet of paper by
a list of seven features corresponding to the job seeker’s
EXPERIMENT 1 seven requirements. The features were always listed
in the same order, and the job’s status was given for
Method each feature. Thus, for example a job might be de-
scribed by Firm Size: Large; Location: Out of State;
Subjects. Forty-two undergraduate business stu- Creative Freedom: High; Administrative Responsibil-
dents were each assigned to one of two experimental ity: Low; Initial Training: No; Travel: No; Vacation: 1
groups, a time-constrained group (n Å 19 Ss) or a non- week. Each feature of a job that was different from the
constrained group (n Å 23 Ss). In addition, there was job seeker’s requirements, listed above, was regarded
a pretest group (n Å 21 Ss). as a violation. Table 1 contains the features of each of
the eight jobs, indication of which features are viola-
Procedure. All subjects were told: ‘‘Imagine that tions of the job seeker’s requirements (standards), and
you are a 22 year old, single, student who will soon the value of I for each job.
graduate with a bachelor’s degree in Marketing. You
have gone to the Placement Center to look at available Pretest. In order to establish the time constraint for
jobs. The Center has provided you with descriptions of the main experiment, 21 subjects individually per-
8 jobs in marketing, all of which pay the entry level formed the screening part of the task while being
salary. Each of the jobs is described in terms of 7 fea- timed. The mean time for completion was 3 min, 11 s.
tures: (1) firm size, (2) location, (3) degree creative free- The distribution of times was roughly normal with a
dom encouraged, (4) degree of administrative responsi- standard deviation of 1 min, 7 s. Following the practice
bility, (4) whether there is an initial training period, (5) prescribed by Svenson and Benson (1993b), one stan-
whether extensive travel is required, and (7) amount of dard deviation below the mean was adopted as the time
annual vacation granted during the first two years with constraint, i.e., 2 min, 4 s to complete the task in the
the firm.’’ time-constrained condition of the experiment.
All subjects further were told: ‘‘You have strong re- The 23 subjects in the unconstrained (UC) group
quirements in regard to each of these 7 job features: screened the eight jobs at their own speed, chose the
You want to work for a small firm, preferably in Tuc- best from among the survivors, then ranked the seven
son. You want a high degree of creative freedom, but a features according to importance. The 19 subjects in
low degree of administrative responsibility until you the time constraint (TC) group screened as many of the
have been on the job for a few years. You want an eight jobs as they could in 2 min and 4 s, chose the
initial training period so you can more easily fit into best of the survivors, and ranked the features according
the firm, you want as little travel as possible, and you to their own perceptions of the features’ importance in
want at least 2 weeks of vacation. Of course, you may job screening. Because subjects nearly always chose as
not get precisely what you want, but these require- best the survivor that had the fewest violations, the
ments reflect your preferences.’’ choice data are not particularly interesting. This, to-
All subjects were told: ‘‘In light of your requirements, gether with the fact that choices were not the focus of
please screen the 8 job descriptions, rejecting those that the present study, makes it unnecessary to consider
are of no further interest and retaining those that you the choice data further.
would apply for. Do this by marking one or the other
blank ( Reject or Apply) at the end of each Results
description.’’ Subjects in the time constrained group
were further told: ‘‘In order to make sure everyone does Let us deal first with the importance rankings. The
the screening part of this task in a timely manner, we rankings were similar for both groups. Each individual
will allow you 2 minutes and 4 seconds to screen all 8 subject’s ranks were treated as weights in equation 1
jobs, after which we will, as a group move on to the and I was computed for each of the eight jobs. The
second part of the task. When the experimenter gives ordering of the jobs was then compared to the ordering
the signal, begin screening the jobs. Try to complete when weights were ignored and violations were merely
all 8 jobs in the allotted time.’’ After screening was counted (01) for each job. For virtually all subjects, the
completed, all subjects were instructed to go back and ordering was the same either way, indicating that the
choose a job from among those jobs that they had not ranks added little information, a common finding
rejected. Then they rank-ordered the seven job features (Beach & Strom, 1989). This does not necessarily mean

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


TIME CONSTRAINTS 225

TABLE 1
Features of Each of the Eight Jobs

Features

Job Size Location Creativity Admin. Training Travel Vacation I

A Small Tucson High Low Yes No 2 wks 0


B Large Tucson High Low Yes No 2 wks 01
C Small Tucson Low Low Yes No 1 wk 02
D Large Tucson Low Low Yes Yes 2 wks 03
E Large Other High Low No No 1 wk 04
F Large Other High High Yes Yes 1 wk 05
G Large Other Low High No No 1 wk 06
H Large Other Low High No Yes 1 wk 07

Note. Violations of the standards are in italics. The column on the right contains I for each job.

that subjects regarded all features of the jobs as equally Of the 19 subjects in the TC group, 2 (10%) failed
important or that their screening decisions were not to complete the task but it was unclear whether they
differentially affected by the features. Indeed, research ignored the time constraint and plodded on with the
specifically designed to detect differential weighting is execution of the compatibility test or whether they sim-
able to do so (Beach, Puto, Heckler, Naylor, & Marble, ply failed to perform the task appropriately. Of the
1996). However, because it is a linear equation, the remaining 17 subjects who completed the task in the
compatibility test is not sufficiently sensitive to reveal constrained time, only 7 (41%) exhibited consistent re-
such effects unless experimental conditions are spe- jection patterns, which is significantly different from
cifically designed to do so (Dawes, 1979; Dawes & Corri- the 87% for the UC group who exhibited consistent
gan, 1974). As a result, even though the rankings are rejection thresholds (Z Å 3.21, p õ .001), demonstra-
valuable for the analysis of abbreviated strategy use, ting an effect of the time constraint. The mean rejection
they are not of help for the analysis of inconsistencies threshold for the 7 who both completed the task and
in execution of the compatibility test and we will act had consistent patterns was 2.71, which is not signifi-
as though the subjects were using unit weights. The cantly different from the threshold of 2.89 for the UC
values of I in the right-hand column of Table 1 are group.
calculated using unit weights.
Abbreviation. The rejection patterns of each of the
Inconsistency. When the jobs are listed according 10 subjects in the TC group who completed the task
to increasing number of violations (last column in Table but had inconsistent rejection patterns were subjected
1) previous research leads to the expectation that a to individual analyses in an attempt to account for their
subject will retain all jobs having fewer violations than rejection pattern with an abbreviation of the compati-
his or her rejection threshold and will reject all jobs bility test. The first abbreviation consisted of recomput-
having more violations than his or her rejection thresh- ing Eq. (1) after deleting the feature the subject had
old. (That is, the observed pattern of responses as one ranked as least important, then recomputing again de-
proceeds down the list should contain no retentions leting the least and next least important features, and
after the first rejection.) This will be called a consistent so on until only the most important feature remained.
rejection pattern. Because different subjects will have The question was whether any abbreviated set of fea-
different rejection thresholds there need not be a simi- tures would account for the subject’s rejection pattern.
lar ‘break point’ in the pattern of retentions and rejec- The second abbreviation consisted of using only the
tions across subjects, but the consistent rejection pat- first few features in a job’s description. Because the
tern must hold for individual subjects if we are to con- seven features were listed in the same order for all
clude that they are performing consistently. eight jobs, it was possible to examine this version of
All of subjects in the UC group completed the task, abbreviation for each subject in the same way the ver-
of course. Of the 23 subjects in this group, 20 (87%) sion described above was examined.
exhibited consistent rejection patterns. The mean re- Neither version of the abbreviated compatibility test
jection threshold for the 20 who had consistent patterns accounted for the rejection patterns of any of the seven
was 2.89 violations. subjects.

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


226 BENSON AND BEACH

Speed. Recall that because the time constraint of 2 TABLE 2


min, 4 s is one standard deviation below the mean time Percentage of Subjects with Consistent Rejection Patterns,
taken by the pretest subjects to complete the task, it Mean Thresholds, and Percentage of Subjects Who Completed
the Task for Each of the Time Constraint Conditions in Exper-
is to be expected that roughly 84% of the subjects in
iments 1 and 2
the TC group would fail to complete the task if they
did not speed up or abbreviate execution of the compati- Experiment 1 Experiment 2
bility test. As we saw above, no subject abbreviated the Time
test. Only 3 of the 19 (16% versus the expected 84%, Constraint: UC TC UC TC
Å 1.89, p Å .03) of subjects in the TC group failed to % Consistent 87 41 72 65
complete the task, implying that relative to the pretest Mean threshold 2.89 2.71 3.61 3.65
subjects, most of the TC subjects in Experiment 1 were % Complete 100 84 100 100
able to speed up in order to complete the task within
the allotted time.
EXPERIMENT 2 significantly different from the 87% for the UC group
The results of Experiment 1 show that time con- in experiment 1 (Table 2). The mean rejection threshold
straints interfere with screening, as revealed by the for the 18 subjects who evidenced consistent rejection
significantly greater number of inconsistent rejection patterns was 3.61, which is significantly different from
patterns in the TC group. Moreover, the effect does not the mean rejection threshold of 2.89 for the UC group
lie in use of an abbreviated version of the test. in Experiment 1 (t Å 2.43, p Å .02). In short, when
The 2/19 (10%) of the subjects in the TC group who there was no time constraint, stressing the importance
did not complete the task may have been unwilling or of the task did not influence the percentage of subjects
unable to compensate for the time constraint by speed- evidencing consistent rejection patterns but it signifi-
ing up execution of the compatibility test, but the data cantly increased the subjects’ willingness to overlook
do not settle the question of whether they could have flaws in the job options. That is, the instructions about
compensated by speeding up, it merely suggests that importance raised the rejection thresholds so that sub-
they might not have done so. The question remains, jects retained more jobs than they did when importance
therefore, whether the subjects in the TC group, both was not stressed.
those who completed and those who did not, were op- Seventeen of the 26 subjects in the TC group (65%)
erating at their capacity or whether they could have evidenced consistent rejection patterns. This percent-
increased execution speed, and consistent execution of age of consistent patterns is significantly greater than
the strategy, even more if they had a reason to do so. the 41% for the TC group in Experiment 1 (Z Å 1.60,
Previous research (e.g., McAllister, Mitchell, & p Å .05). The mean rejection threshold for the 17 sub-
Beach, 1979) has demonstrated that stressing the im- jects who evidenced consistent rejection patterns was
portance of the decision can increase the level of perfor- 3.65, which is a significantly higher threshold than the
mance. Therefore, Experiment 2 consisted of rerunning mean rejection threshold of 2.71 for the TC group in
Experiment 1 with the addition of the following to the Experiment 1 (t Å 2.21, p Å .04), but not significantly
initial instructions for both groups: ‘‘Because you are different from the mean threshold of 3.61 for the UC
low on money and have bills that are past due, it is group in Experiment 2 (above). For none of the 9 sub-
important for you to find a job very soon. Moreover, jects who exhibited inconsistent patterns could the pat-
whatever job you end up taking, you almost certainly terns be accounted for by an abbreviated version of the
will have to stay with it for a while because you cannot compatibility test.
risk going without a paycheck while you move to an- All of the 26 TC subjects (100%) completed the task,
other job.’’ It was expected that stressing the impor- which is significantly greater than the 84% observed in
tance of the task would, compared to the subjects in Experiment 1 (z Å 2.66, p Å .004) and significantly
Experiment 1, lead subjects in this experiment to in- greater than the 16% that would be expected if subjects
crease execution speed if they could, while maintaining had not increased processing speed beyond that of the
consistency. pretest subjects (Z Å 9.71, p õ .001). This implies that
There were 25 subjects in the UC group and 26 in the instructions resulted in even the slowest subjects in-
the TC group. creasing execution speed in order to complete the task.
In summary, in contrast to the TC group in Experi-
Results ment 1, stressing the importance of the task led to all
Eighteen of the 25 subjects in the UC group (72%) of the subjects completing the task, implying increased
evidenced consistent rejection patterns, which is not speed of execution for even the slowest subjects. It also

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


TIME CONSTRAINTS 227

led to a significantly greater percentage of the subjects constraint in their experiments. Indeed, in most re-
evidencing consistent rejection patterns and to a sig- ports no rationale at all is given. In contrast, by timing
nificant rise in rejection thresholds for those who evi- ‘‘normal’’ execution time in a pretest and selecting an
denced consistent patterns such that they retained interval that is precisely one standard deviation below
more jobs than they did when task importance was not the mean of the distribution of observed times, we are
stressed. on much firmer ground. The use of one standard devia-
tion below the mean enables us to make a strong predic-
tion about what ought to happen if subjects can not or
DISCUSSION
do not increase their speed of execution; 84% of the
subjects in the time constrained group ought to fail to
These experiments were designed to examine the ef- complete the task. In fact, in Experiment 1 we found
fects of time constraints on screening, to complement that only 16% did not complete the task and in Experi-
earlier studies of the effects of time constraints on ment 2, when importance was stressed, none of the
choice. Experiment 1 demonstrated that time con- subjects failed to complete, implying that the instruc-
straints result in an increase in the number of subjects tions motivated them to speed up even more than in
exhibiting inconsistent rejection patterns, suggesting Experiment 1. We concur with Svenson and Benson
less careful execution of the compatibility test, al- (1993a, 1993b) that, when times are normally distrib-
though the exact reason is not clear. Even at that, more uted, this statistical definition of time constraint ought
than the expected number of subjects completed the to become the standard way of defining time con-
task, implying that they increased the speed of execu- straints in future research.
tion of the test. There was no evidence of subjects re- Perhaps the most puzzling result of this research is
sorting to abbreviated versions of the test in order to the absence of evidence for use of abbreviated versions
finish the task within the prescribed time. of the compatibility test by the time constrained sub-
These results are only in part similar to the results jects. Future research should be designed to more
for choice. Time constraints lead some subjects to speed closely examine this question. If for no other reason
up strategy execution in both screening and choice, but than that common sense suggests there ought to be a
there did not appear to be a counterpart in screening point at which motivation to complete the task is so
to the finding that time constraints lead to switching great, and the time allowed to do the task is so short,
to simpler choice strategies. Instead, some subjects ap- that subjects must resort to using only the most im-
peared simply to be less consistent in executing the portant or the first information they receive about op-
compatibility test in screening, leading to inconsistent tions in order to screen them. Presumably the present
rejection patterns and inconsistent rejection thresh- experiments fell short of that point, but even at that it
olds. is striking that subjects attempted to use all of the
Experiment 2 demonstrated that, in comparison to information about each option they screened, even at
the time constrained condition of Experiment 1, in- the price of inconsistent application of the compatibility
structions that stressed the importance of the task re- test or of failing to complete the task. Future studies
sulted in fewer inconsistent rejection patterns, perhaps must examine task conditions that do or do not encour-
suggesting more careful execution of the compatibility age subjects to compromise precision and completion
test, but in addition the rejection threshold was raised in order to preserve thoroughness.
so fewer jobs were rejected. The importance instruc-
tions also resulted in all of the subjects completing the REFERENCES
task, suggesting an increase in the speed of execution
by even the slowest subjects. Thus, importance Beach, L. R. (1990). Image theory: Decision making in personal and
organizational contexts. Chichester, UK: Wiley.
prompted subjects to operate more quickly but to be
Beach, L. R. (1993). Broadening the definition of decision making:
more tolerant of incompatibility (raised rejection The role of prechoice screening of options. Psychological Science,
threshold), thereby screening out fewer jobs and pass- 4, 215–220.
ing on the more exacting part of the decision to the Beach, L. R., & Mitchell, T. R. (1990). Image theory: A behavioral
choice stage of the process. theory of decisions in organizations. In B. M. Staw & L. L. Cum-
We should note that the manner in which the time mings (Eds.), Research in organizational behavior (Vol. 12). Green-
constraint was selected for these experiments is all- wich, CT: JAI Press.
Beach, L. R., Puto, C. P., Heckler, S. E., Naylor, G., & Marble, T. A.
important in detecting increased speed of strategy exe-
(1996). Differential versus unit weighting of violations, framing,
cution. As noted by Svenson and Benson (1993a, and the role of probability in image theory’s compatibility test.
1993b), most researchers are rather cavalier about se- Organizational Behavior and Human Decision Processes, 65, 77–
lecting the prescribed interval to be used for the time 82.

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP


228 BENSON AND BEACH

Beach, L. R., & Strom, E. (1989). A toadstool among the mushrooms: strategy selection in decision making. Journal of Experimental
Screening decisions and image theory’s compatibility test. Acta Psychology: Learning, Memory & Cognition, 14, 534–552.
Psychologica, 72, 1– 12. Svenson, O., & Benson, L., III. (1993a). Framing and time pressure
Benson, L., III, & Svenson, O. (1992). Post-decision consolidation in decision making. In O. Svenson & A. J. Maule (Eds.), Time
following the debriefing of subjects about experimental manipula- pressure and stress in human judgment and decision making. New
tions affecting their prior decisions. Psychological Research Bulle- York: Plenum.
tin, 32, 1– 13. Svenson, O., & Benson, L., III. (1993b). On Experimental instruc-
Dawes, R. (1979). The robust beauty of improper linear models in tions and the inducement of time pressure behavior. In O. Sven-
decision making. American Psychologist, 34, 571–582. son & A. J. Maule (Eds.), Time pressure and stress in human judg-
Dawes, R., & Corrigan, B. (1974). Linear models in decision making. ment and decision making. New York: Plenum.
Psychological Bulletin, 81, 94– 106. Svenson, O., Edland, A., & Slovic, P. (1990). Choices between incom-
Edland, A., & Svenson, O. (1993). Judgment and decision making pletely described alternatives under time stress. Acta Psycholog-
under time pressure: Studies and findings. In O. Svenson & A. J. ica, 75, 153– 169.
Maule (Eds.), Time pressure and stress in human judgment and Svenson, O. & Maule, A. J. (Eds.) (1993), Time pressure and stress
decision making. New York: Plenum. in human judgment and decision making. New York: Plenum.
McAllister, D., Mitchell, T. R., & Beach, L. R. (1979). The contingency Smith, J. F., Mitchell, T. R., & Beach, L. R. (1982). A cost benefit
model for selection of decision strategies: An empirical test of the mechanism for selecting problem-solving strategies: Some exten-
effects of significance, accountability, and reversibility. Organiza- sions and empirical tests. Organizational Behavior and Human
tional Behavior and Human Performance, 24, 228–244. Performance, 29, 370–396.
Maule, J., & Mackie, P. (1990). A componential investigation of the Wright, P. (1974). The harassed decision maker: Time pressure, dis-
effects of deadlines on individual decision making. In K. Borcher- traction and the use of evidence. Journal of Applied Psychology,
ing, O. I. Larichev & D. M. Messick (Eds.), Contemporary issues 59, 555–561.
in decision making. Amsterdam: North-Holland. Zakay, D. (1985). Post-decision confidence and conflict experienced
Payne, J. W., Bettman, J. R., & Johnson, E. J. (1988). Adaptive in a choice process. Acta Psychologica, 58, 75– 80.

Received: October 30, 1995

/ a706$$2636 08-02-96 13:25:02 obha AP: OBHDP

Você também pode gostar