行为决策理论：判断与选择过程解析

需积分: 10 146 浏览量更新于2024-07-25 收藏 1.22MB PDF 举报

行为决策理论是社会科学和经济学领域的重要研究分支，它关注个体在面对复杂情境时如何进行判断和选择的过程。这一理论强调人类决策并不总是基于完全理性的计算，而是受到情感、认知偏差、社会影响等多因素的综合影响。Hillel J. Einhorn 和 Robin M. Hogarth 在他们的经典论文《行为决策理论：判断与选择的过程》（Behavioral Decision Theory: Processes of Judgment and Choice）中，深入探讨了这些非理性因素如何影响决策制定，并将研究聚焦于会计学领域，发表在1981年春季号的《会计研究期刊》(Journal of Accounting Research) 第19卷第1期，页码1-31。该论文由Wiley-Blackwell出版社代表位于芝加哥大学布斯商学院的会计研究中心（Accounting Research Center）发布，通过JSTOR在线数据库提供访问。JSTOR是一个非营利组织，致力于将学术期刊、书籍和其他研究资料数字化保存，以便学者、研究人员和学生能够方便地查找、利用和扩展这些珍贵的信息资源。使用JSTOR服务需接受其使用条款和条件，详情可访问<http://www.jstor.org/page/info/about/policies/terms.jsp>。论文作者Einhorn和Hogarth的研究成果对于理解决策学中的心理经济学视角至关重要，它不仅适用于会计领域的实践和理论探讨，也对管理、市场营销、金融等多个学科有着广泛的影响。他们通过实证分析和理论模型，揭示了决策者在面临不确定性时如何处理信息、权衡利益，以及如何被心理偏见所左右。此外，论文可能还讨论了如何通过教育、培训或设计更合理的决策框架来减少这些偏差，提高决策的效率和有效性。阅读这篇论文，读者可以深入了解行为决策理论的最新进展，以及如何将这些理论应用到实际决策场景中，提升决策的质量和一致性。同时，这也为跨学科的研究提供了桥梁，促进了心理学、经济学与商业实践之间的知识交流。

BEHAVIORAL DECISION THEORY

horizon is specified, optimality

can also be problematic in what might

seem

be simple situations.

Environment vs Problem

Space

The importance to behavior of

the cognitive representation

of the

task,

i.e.

"problem

space,"

has

been

emphasized by

Newell

& Simon (1972).

It is

now

clear

that

the process of

representation, and the

factors

that

affect

it,

are

major importance

judgment

and choice. Illustrations of

the effects

problem representation on

behavior are found

work

estimating

probabilities via fault trees

(Fischhoff, Slovic & Lichtenstein

1978);

re-

sponse

mode effects

inducing

preference

reversals

(Grether

& Plott

1979);

coding processes

risky choice

(Kahneman

Tversky 1979a); "problem

isomorphs" in problem

solving (Simon & Hayes 1976); context

effects in

choice (Aschenbrenner 1978,

Tversky & Sattath 1979)

and

agenda

setting

(Plott & Levine 1978);

purchasing behavior (Russo 1977);

and

causal sche-

mas

probability judgments

(Tversky & Kahneman 1980a).

essential to emphasize

that the cognitive approach

has been con-

cerned

primarily

with how tasks

are represented.

The

issue of why tasks are

represented in particular ways

has not yet been addressed. However,

given

functional arguments, this is

a crucial issue

view of the way

minor

contextual changes can lead

to the violation

the

most

intuitively

appeal-

ing normative principles, e.g.

transitivity.

The

reconciliation of

persistent errors

and biases with functional

argu-

ments

has

taken two forms.

First, it has been claimed that such effects can

be overcome

by increasing

incentives

(through higher payoffs and/or

pun-

ishments).

one sense, this

argument is irrefutable since it can

always

claimed that the incentive wasn't

high enough. However,

direct

evidence

shows that

increased

payoffs

not

necessarily

decrease extreme overconfi-

dence

(Fischhoff,

Slovic &

Lichtenstein

1977)

nor

prevent preference

rever-

sals

(Grether

Plott 1979).

Furthermore,

the indirect evidence from

clinical

judgment

studies

naturally occurring settings,

where

payoffs

are

presumably high enough

motivating,

continues

indicate

low

validity

and

inferiority

statistical models

(Dawes 1979).

addition,

claims

that

people

will

seek

aids

and/or

experts

when the stakes are

high (Edwards

1975) are predicated on the

assumptions that: (a) people know that

they

don't know; and (b) they

know (or believe) that others do. On the

other

hand,

it is

foolish to deny that

payoffs, and thus motivation, have no

effect

processes

judgment

and choice.

Indeed,

one

only

needs

recall the

fundamental

insight

signal detection

theory (Green

& Swets

1966),

which

that

both

cognitive

and

motivational components

affect

judgment

(also

see

Killeen

1978).

second

way

reconciling

biases

with

functional

arguments

involves

enlarging

the context

which

performance

is evaluated.

This has

taken

four

forms: 1.

One

view

evolutionary theory (as espoused by

the sociobi-

ologists;

for

example,

Wilson

1978)

could

lead to the belief that the human

H. J.

EINHORN

AND R. M.

HOGARTH

system

represents the

optimal design

for a

complex

environment.

Heuristics

exist because

they

serve

useful

functions and

their

benefits

outweigh their

costs. While

this view is

often

espoused, there

surprisingly little

evidence

support

it.

important exception

is the

simulation

study by

Thorngate

(1980), where

it was

shown how

heuristics can

often pick

the best of

several

alternatives across

range

tasks.

However, neither

this

study

nor

any

other that we

are aware of

has considered the distribution of

tasks

the

natural

environment

which

heuristics

would work

well or

poorly. 2.

Hogarth

(1980a) has

argued that

most

judgments and

choices occur

sequen-

tially

and

that

many

biases

reflect

response

tendencies which

are functional

dynamic

environments.

Furthermore, the

static tasks

typically

investi-

gated

reflect a

preoccupation with

those

relatively

simple

situations for

which

optimal

models can

constructed. 3.

Toda

(1962) has claimed that

the

coordination of

behavior

that reflects an

organism's

efficiency, not

individual and

thus

isolated

actions.

Furthermore,

coordination between

functions

requires

trade-offs and these

can be facilitated

limitations

(e.g.

limited

memory facilitates efficient

forgetting of

needless

detail).

Cost/

benefit

analyses

can be

expanded

include "the cost of

thinking"

(Shugan

1980), which

seems

compatible with

notions of

bounded

rationality.

While there

is much

merit in the

above

arguments,

care must

be taken

since

they

can

easily become

tautological; i.e. costs

and

benefits can

defined

post

hoc

accord

with a

presumption

optimality.

However,

can

there be actual

dysfunctional behavior

(rather

than

seeming dysfunctional

behavior)

that

persists,

and

so,

what

mechanism(s)?

Since

functional

arguments

rest

evolutionary

theory,

it is

easy

overlook

the

fact

that

nonadaptive

behavior can also be

compatible

with

principles of

natural

selection:

Biological

evolution is

directly related to

the amount of

variance

the

genotype (Lewontin

1979).

For

example,

the

development of

wings

could

be functional for humans on

many

occasions.

However,

without an

appropriate

mutation

(the chance of which

minis-

cule),

such

evolution cannot take

place.

While it is evident

that

physical

limitations

preclude certain

types

of behavior

regardless

incentives to the

contrary,

biological

limitations can also

preclude

certain

cognitive

opera-

tions

(Russo

1978).

For

example,

the

study

memory

indicates

limitations

short-term

storage

and

retrieval.

Furthermore,

Seligman

(1970)

has

explicated

biological

limitations

the

learning process

itself.

Cognitive

limitations

can

therefore

persist

and

dysfunctional

(relative

given

goals)

for

the same reasons

that account for

physical limitations.

The

time-frame of

human

biological

evolution is such that it can be

considered

constant

over

many generations. It

is thus difficult

to determine whether

any

current

trait

mechanism

becoming

more or less

adaptive,

or is a

vestige without

apparent

function

(e.g. the human

appendix:

see

also

Skin-

ner

1966).

Therefore,

without

denying general

cost/benefit considerations

over

the

very

long run,

dysfunctional

behaviors

may persist

for

extremely

long

periods by human

standards. The demise of the

dinosaur,

for

example,

剩余31页未读，继续阅读

jonathan70

粉丝: 0
资源: 2

行为决策理论：判断与选择过程解析

决策理论与方法

决策与决策方法决策理论与决策过程.ppt

辅助决策平台的理论依据

value activated

马尔可夫决策过程理论与应用 pdf

自动驾驶决策算法中的行为规划算法

有没有关于冲动消费的消费者行为理论

用户行为预测依据的基本理论与方法

请比较常用的激励理论(马斯洛的需求层次、x理论、Y理论、Z理论和期望理论)的优缺点和适用范围

弗鲁姆的期望理论是什么

最新资源