Copyright
Susan Fox Schiro.

Effective goals for complex tasks : towards a broader conceptualization of specific-difficult goals online

. (page 1 of 4)
Online LibrarySusan Fox SchiroEffective goals for complex tasks : towards a broader conceptualization of specific-difficult goals → online text (page 1 of 4)
Font size
QR-code for this ebook


^K^^orrtf^



^



tiUZO

.M414
no.

^7






CENTER FOR COMPUTATIONAL RESEARCH
IN ECONOMICS AND MANAGEMENT SCIENCE



EFFECTIVE GOALS FOR COMPLEX TASKS:

TOWARDS A BROADER CONCEPTUALIZATIOJ

OF SPECIFIC-DIFFICULT GOALS

Susan F. Schiro



October, 1989



WP #: 3096-89-BPS



ALFRED P. SLOAN SCHOOL OF MANAGEMENT

MASSACHUSETTS

INSTITUTE OF TECHNOLOGY

CAMBRIDGE, MASSACHUSETTS 02139



EFFECTIVE GOALS FOR COMPLEX TASKS:

TOWARDS A BROADER CONCEPTUALIZATION

OF SPECIFIC-DIFFICULT GOALS

Susan F. Schiro
October, 1989 WP #: 3096-89-BPS



—• ^




EFFECTIVE GOALS FOR COMPLEX TASKS:
TOWARDS A BROADER CONCEPTUALIZATION OF SPECIFIC-DIFFICULT GOALS

Susan F. Schiro

A key finding of goal theory is that specific-difficult goals, i.e.,
quantitative goals which are challenging to the individual, lead to improved
performance. However, recent research suggests that these goals are
ineffective for highly complex tasks such as managerial jobs. This paper
argues that if a broader conceptualization of specific-difficult goals is adopted,
then specific-difficult goals will be seen to improve the performance of complex
tasks.

Difficult quantitative goals improve task performance by providing a
precise definition of the desired outcome; the individual therefore considers
only a few strategies for dealing with the situation. When the assigned task
is of low to moderate complexity, this leads to improved performance. On a
highly complex task, however, such a goal leads to the selection of an inferior
action plan and, therefore, to reduced performance.

A highly complex task requires the individual to consider the relative
importance of various goals and to choose the best strategy for achieving the
selected goal or goals. Consequently, an effective goal for such tasks will
both provide a precise definition of the desired outcome and encourage the
individual to consider a broad variety of action plans.

This paper proposes that a goal system is an effective type of specific-
difficult goal for a highly complex task. A goal system is a goal hierarchy
which includes both a superordinate goal (or goals) and subgoals. The
superordinate goal defines the individual's primary objective(s) . Subgoals may
specify particularly desirable features of the superordinate goal, identify several
steps towards the superordinate goal as separate goals, or describe other
desirable but less critical goals. Thus, subgoals increase the specificity with
which the superordinate goal is defined thereby increasing the individual's goal
comprehension .

The superordinate goal identifies non-optional goals while subgoals are
presented as optional objectives. Because the subgoals are presented as
optional goals, the individual is more likely to consider a broad variety of
action plans. The superordinate goal then serves as a template against which
the possible strategies can be evaluated.

A research project which measures the relative effectiveness of
quantitative and non-quantitative goal systems versus simple
quantitative goals and do-your-best goals in both a simple and a complex
version of a managerial task is proposed. The project also includes
measurement of the key cognitive processes which determine goal effectiveness.
No previous study has included multiple types of goals, two levels of
complexity, and the key cognitive steps in the process by which goals influence
performance. This study therefore provides a more complete test of a model of
the process by which goals influence performance than is currently available.

Thesis Committee: Deborah Ancona

John Carroll (chair)
Robert Wood



Table of Contents



Page



Introduction 4

Statement of the Problem 4

Design of the Paper and Proposed Research 5

Contribution of the Papaer 6

Literature 7
Task Complexity and the Effectiveness of Specific-Difficult Goals 7

Goal Systems 11

Hypothesized Relationships among Task Complexity, Goal Type, and the

Steps by Which Goals Influence Performance 12

Scope of the Model 13

Step 1

Step 2:

Action or Strategy Planning 19



Goal Comprehension 14

Goal Acceptance 17



Step 3

Outcome: Performance 22

Methodology: Experimental Design 23

Criteria for Selection of Experiment 23

Description of Proposed Experiment 24

Suitability of the Simulation 25

Goals 27

Subjects 28

Manipulation Checks 28

Tests of Hypotheses 29

Hypothesis #1 29

Hypothesis #2 31

Hypotheses #3a and #3b 31

Hypotheses #4a and #4b 32

Hypothesis #5 32

Hypothesis #6 33

Hypothesis #7 33

Hypothesis #8 34

Contribution of the Research 34

Theoretical 34

Practical 35

References 37

Appendix

1 - The Effect of Quantitative Specificity on the Performance

of Complex Tasks 44

2 - Proposed Questions for Measuring Type of Action Plan 45



Tables

1 - Studies Measuring Two or more Steps in the Process by which
Goals Influence Performance or Including Two Measures of Action
Planning 48

2 - Studies Measuring the Effect of Goals on Tasks of Varying
Complexity 50

Figures

1 - A Simplified Model of the Key Cognitive Processes by which
Goals Influence Performance 52

2 - Hypothesized Strength of Relationships among Task Complexity,
Goal Type, and Steps by which Goals Influence Performance 53

3 - Research Design 54



Introduction

Statement of the Problem

A key finding of goal theory is that specific-difficult goals, i.e.,
quantitative goals which are challenging to the individual (Locke, Shaw, Saari,
& Latham, 1981), are more effective than vague do-your-best goals in improving
performance; this has been confirmed by research in both laboratory and field
settings (Latham & Lee, 1985; Mento, Steel, & Karren, 1987; Tubbs, 1986).
However, a recent meta-analysis demonstrates that specific-difficult goals are
decreasingly effective in improving performance as task complexity increases,
although usually still significant (Wood, Mento, & Locke, 1987). Additionally,
for some highly complex heuristic tasks specific-difficult goals even lead to
poorer performance than a non-specific do-your-best goal (Earley, Connolly, &
Ekegren, 1989; Huber, 1985). This suggests that specific-difficult goals may be
ineffective or counterproductive for managerial tasks. If that is true, then we
have lost a potentially valuable tool for improving managerial performance; goal
theory is widely recognized as one of the most useful theories in organizational
behavior (Miner, 1984; Pinder, 1984; Schneider, 1985; Staw, 1984).

This paper argues that specific-difficult goals are relevant to improving
the performance of complex tasks, even heuristic managerial tasks. The
problem is the operationalization of specific-difficult goals, not the theory.
Goal theory virtually always operationalizes specific-difficult goals as
quantitative goals. This has happened because most of the research on goals
has been done with simple tasks (Huber, 1985; Locke et al. 1981) or tasks of
only moderate complexity (Locke Chah, Harrison, & Lustgarten, 1989);
quantitative goals are highly effective for such tasks. However, goals do not
have to be quantitative to be difficult or specific. Goal difficulty is defined as



the extent to which the goal is challenging to the individual (Locke et al. 1981)
or the probability that the individual will achieve the goal (Naylor & Mgen,
1984). Goal specificity is defined as the explicitness or clarity with which the
desired outcome is defined (Locke & Latham, forthcoming; Locke et al. 1989;
Naylor & llgen, 1984). Thus, it is legitimate to use an alternative
operationalization of specific-difficult goals.

Naylor and llgen (1984) propose that there are two types of goal
specificity. One type of goal specificity is quantitative specificity (i.e., the
degree of quantitative precision with which the level of performance of the
desired outcome is defined); however, another type of goal specificity is
content or outcome specificity (i.e., the explicitness with which the content of
the desired outcome is defined). For example, the goal of writing two
publishable papers a year is specific with regard to both quantity and content.
The goal of writing publishable papers is specific with regard to content but is
not specific regarding quantity. The goal of making two scientific
contributions a year is specific with regard to quantity but is not specific
regarding the content of the goal (Naylor & Mgen, 1984). Although there is a
substantial body of research on the effect of quantitative specificity on task
performance, there is little consideration of the effect of content specificity on
performance (exceptions include Campbell & Gingrich, 1986; Earley 1985, 1986;
Erez 5- Arad, 1986). This paper therefore discusses how content specificity
influences the performance of simple versus complex tasks.
Design of the Paper and Proposed Research

The paper does this by considering two operationalizations of specific-
difficult goals. One operationalization is called a simple goal; this includes both
the traditional type of quantitative specific-difficult goal and vague do-your-
best goals (henceforth these will be referred to as simple quantitative goals and



simple non-quantitative goals, respectively). The second operationalization is a

goal system which includes a superordinate goal plus subgoals (Bandura, 1988).

The paper supports the selection of this operationalization by discussing the

characterisitics of effective goals for complex tasks and by reviewing current

theory and research on the cognitive processes by which specific-difficult goals

influence performance. The paper hypothesizes that simple goals will lead to

superior performance on a simple task, but that goal systems will lead to

better performance on a complex task. The paper supports these hypotheses

with hypotheses about the effect of simple goals versus goal systems on each of

the steps' in the process by which goals influence performance.

The paper then proposes a research project to measure the effect of the

two types of goals on overall performance and on the steps in the process by

which goals influence performance for both simple and complex tasks. Thus,

the approach is consistent with the dictum that:

Any theory that describes a process responsible for a given effect (be it
motivational or cognitive) must attempt to demonstrate this intervening
process in addition to the effect... (Bavelas & Lee, 1978, p. 226).

Contribution of the Paper

This project will represent a particularly valuable contribution to the
literature since research on the process is limited (Bandura & Cervone, 1983;
Locke et al. 1981; Mento et al. 1987; Riedel Nebeker, & Cooper, 1988; Steers &
Porter, 1987; Wood et al. 1987; Wood & Locke, forthcoming). Although many
researchers have studied one of the steps in the process by which goals
influence performance, only a handful of studies have considered two steps (see
Table #1), and no study has attempted to measure all of the steps in the
process. Additionally, few studies have included tasks at more than one level



' These steps are generally described as goal comprehension, goal
acceptance, and action or strategy planning (Locke, 1968; Locke et al. 1981;
Wood and Locke, forthcoming).



of complexity (for exceptions see Table #2).

Furthermore, although this paper is primarily concerned with the effect of
content specificity on task performance, the study design includes both
quantitative traditional specific-difficult goals and non-quantitative do-your-best
goals. These are necessary to provide evidence that the simple task is simple
and that the complex task is complex. Since these are included, the study also
includes both quantitative and non-quantitative versions of a goal system. The
study therefore also contributes to our knowledge of the way in which
quantitative specificity influences task performance.

This paper is intended to lead to a broader conceptualization of specific-
difficult goals. The proposed research will increase our knowledge of the
relative effectiveness of two types of specific-difficult goals on the
performance of simple and complex tasks. It will also provide a more complete
test of one model of the process by which goals influence performance than is
currently available.

Literature

This section discusses the key cognitive processes by which simple
quantitative goals lead to improved performance of tasks of low to moderate
complexity. The section then discusses why these processes, and therefore such
goals, are less effective for complex tasks. The section argues that a goal
system which includes both superordinate goals and relevant subgoals (i.e.,
subgoals which relate the superordinate goal to the source of task complexity)
will lead to improved performance of complex tasks.

Task Complexity and the Effectiveness of Simple Specific Goals



By definition, simple specific goals are goals with high content specificity.
Such goals provide a precise definition of the content of the desired outcome
(Naylor & llgen, 1984) thus focussing the individual's attention on a narrow
range of possible actions (Wood and Locke, forthcoming). This leads to
improved performance of simple tasks. For example, if a widget maker has
been successfully meeting the goal of producing 12 widgets per hour, that
individual is likely to continue using the approach which has been successful in
the past, if someone were to ask the widget maker why he/she always used
the same approach, the widget maker might shrug and say: "If it ain't broke,
don't fix it." in the short run at least, the widget maker is making the correct
decision. If the widget maker were to look for a better approach, performance
would initially suffer because the time used in experimenting with different
approaches would be unavailable for making widgets.

Performance of a complex task, however, is improved by a broader
consideration of action plan possibilities. Task complexity is defined by the
extent to which the task possesses one or more of the following attributes:
multiple components (including acts and information cues), high coordination
needs, and dynamic or uncertain conditions (Wood, 1986). As the complexity of
the task increases, it is decreasingly possible to predict the relationship
between an action and its consequence. Therefore, the performance of a highly
complex task is improved by having an individual with relevant specialized
knowledge decide on an appropriate action plan (Carroll & Tosi, 1973; Perrow,
1970; Wood & Locke, forthcoming). When the individual considers many options
or engages in extensive planning, the quality of these choices improves (Janis
and Mann, 1977).

These points are supported by a study done by Hackman, Brousseau, and
Weiss (1976) concerning the effect of time spent planning on both simple and



complex tasks. In this experiment performance was measured by the dollar
value of all components assembled by a group of students. In the simple
condition, all subjects knew what components the group was expected to
assemble. Thus, each subject could independently decide what components to
produce in order to maximize the dollar value of the group's production. in the
high complexity condition, however, each subject knew only some of the
components which were requested. Production of the most profitable
components therefore required coordination. The experimenters found that on
the simple task time spent planning hurt performance. On the complex task,
however, time spent planning improved performance. This suggests that
different types of goals will lead to superior performance on simple versus
complex tasks. Specifically, goals which lead to reduced planning should be
associated with superior performance on simple tasks; on complex tasks,
however, goals which lead to increased planning should be associated with
improved performance.

In the example of the widget maker we saw that a simple specific goal is
unlikely to lead the widget maker to look for a new production method or to
engage in planning. (This point is discussed in greater detail below.) Simple
specific goals lead to a reduced consideration of options and therefore decrease
the individual's ability to make good choices in complex situations'^. Goal
theory therefore needs to identify a different type of goal for complex
situations .

Goal theory makes two different suggestions of what constitutes an
appropriate goal for a complex task. Sometimes, goal theorists suggest that



Because this paper is concerned with the effect of content specificity
on the performance of complex tasks, it does not discuss the effect of
quantitative specificity. In the interest of completeness, however, a brief
discussion is included in Appendix #1.



vague goals may lead to better performance of complex tasks than specific

challenging goals:

Under [some circumstances] (e.g., managing in an uncertain environment)
vague goals could conceivably be more effective than specific goals in that
the manager would have more flexibility in responding to environmental
contingencies (Locke et al. 1989, p. 272).

However, while vague goals give the individual flexibility, they do not define

the desired outcome. This may create problems because different individuals

may have different understandings of the organization's goal (or goals); this

may lead to poor coordination and inefficiency as each individual pursues the

project which in his or her judgment is of greatest importance. Several people

may perform one task while other tasks are left undone.

Goal theorists also suggest that multiple quantitative goals lead to
improved performance in some situations. The effectiveness of such goals has
been demonstrated for proofreading tasks (ligen and Moore, 1987), for
salesclerks (Ivancevich, 1976; Kim, 1984), highly skilled technicians (ivancevich
1977; Ivancevich and McMahon, 1977a, 1977b, 1977c, 1982; Pritchard, Jones,
Roth, Stuebing, and Ekeberg, 1988), and materiel handling and storage units
(Pritchard et al. 1988). Multiple goals lead to improved performance of such
tasks because they increase the individual's awareness of a number of specific
objectives and suggest action plans which are appropriate for each objective.
However, these jobs are less complex than middle level managerial jobs.
Managers frequently need to choose between two different objectives where
achieving one objective precludes achievement of the other objective. Multiple
goals may indicate the various tasks which are to be performed; they do not,
however, provide information about the relative importance of the various tasks.

Therefore, what is needed is a type of goal which specifically defines one
or more desired outcomes and indicates their relative importance. This goal
should stimulate the creation of a new action plan by increasing the individual's

10



awareness of his/her freedom to select a course of action and contributing to
the individual's ability to consider multiple factors and objectives in creating an
appropriate action plan"^.

Goal Systems

A goal system can lead to improved performance of complex tasks because
it defines the desired outcome while simultaneously identifying areas within
which the individual has discretion. A goal system includes one or more
superordinate or overarching goals and multiple subgoals. The superordinate
goal defines the individual's primary objective(s) . Subgoals may specify
particularly desirable features of the superordinate goal, identify several steps
towards the superordinate goal as separate goals, or describe other desirable but
less critical goals. For example, the superordinate goal may be to clean the
house and subgoals may include dusting and making beds. In this example the
subgoals can be thought of either as means by which the superordinate goal is
accomplished or as additional specification as to the definition of a clean house.
Alternatively, the superordinate goal might be to clean the house and an
additional but unrelated subgoal might be to rearrange a vase of flowers.
Thus, subgoals can help to clarify the superordinate goal and to identify the
relative importance of other goals.

The presence of multiple goals does not necessarily create a goal system.
Multiple goals become a goal system only when there is a clearly identifiable
superordinate goal and, therefore, a goal hierarchy. It is the hierarchical
nature of the goal system which enables it simultaneously to provide clear
specification of the desired outcome and flexibility regarding selection of the
means by which that outcome is to be achieved. The superordinate goal is a



Ideally, such a goal will do this without leading to excessive arousal.

11



non-optional goal which identifies the individual's primary objective(s) .
Subgoals identify less critical but desirable outcomes; these represent the
individual's area of discretion.

A goal system is particularly effective when the subgoals relate the
superordinate goal to the factors which make the situation complex. This
focusses the individual's attention on the source of task complexity thereby
leading to an improved choice of actions. For example, the job of an air
traffic controller is complex because of component complexity (e.g., fuel
availability, multiple runways, and landing order), coordinative complexity (e.g.,
coordinating the plane's landing with that of other planes which are advised by
other controllers), and dynamic complexity (e.g., wind rate, wind direction, and
the plane's position in the holding pattern) (Wood, 1986). Clearly, the
controller's superordinate goal is the prevention of accidents. Ideally, subgoals
will increase the controller's awareness of all three types of complexity, thereby
leading to a broader consideration of options and, consequently, to improved
performance.

A superordinate goal can be an individual goal or an organizational goal.
When the superordinate goal is an individual goal, then the goal system is a
hierarchical individual goal system as described by Lord & Kernan (1987) and
Newell E, Simon (1972). Goal systems with superordinate organizational goals
and individual level subgoals have been described by Carroll & Tosi (1973),
Cyert & March (1963), Drucker (1989), March t Simon (1958), and Quinn (1980).
Both kinds of goal systems are thought to lead to improved performance. The
next sertjon discusses some of the processes by which goal systems improve the
performance of complex tasks.

Hvpothesized Relationships among Task Complexity, Goal Type,

12



and the Steps by which Goals Influence Performance

This section discusses a model of the major steps by which goals influence
performance. It argues that goal systems lead to improved performance of
complex tasks because they combine superordinate goals with subgoais. Bandura
states that superordinate goals "... give purpose to an activity and serve a
general directive function, but subgoais are better suited to serve as the
proximal determinants of specific choice of activities and how much effort is
devoted to them" (1988, p. 50). This section describes how these two functions
improve the goal comprehension and action planning of complex tasks.
Conversely, the section argues that goal systems provide superfluous
information which may hinder goal comprehension and action planning for simple
tasks; simple quantitative goals are hypothesized to lead to improved
performance of simple tasks.

Scope of the Model

The model is concerned with the process by which externally imposed
goals, such as those assigned by employers to employees, influence the initial
understanding and, therefore, the performance of cognitive tasks'' by
experienced employees. It is a highly simplified model (as shown below), which
includes only the key cognitive processes by which specific goals influence the
performance of cognitive tasks.

[insert Figure 1 about here]

The model in Figure 1 omits the effect of goals on effort. This is an
appropriate choice because effort is considered to be of less importance in



Wood et al. (1987) state that it is possible that there are significant
differences in the process by which goals influence the performance of
cognitive versus psychomotor tasks.

13



influencing the performance of cognitive as compared with psychomotor tasks
(Wood and Locke, forthcoming) . This is a limited model intended to facilitate a
discussion of the process by which goals influence the performance of cognitive
tasks.

The model is designed to clarify the initial steps by which goals influence
performance. It therefore assumes that each step of the process precedes the


1 3 4

Online LibrarySusan Fox SchiroEffective goals for complex tasks : towards a broader conceptualization of specific-difficult goals → online text (page 1 of 4)