ANNUAL

REVIEWS Further
Quick links to online content
Annu. Rev. Psychol. 1992. 43:399-441
Copyright © 1992 by Annual Reviews 1nc. All rights reserved

TRAINING AND DEVELOPMENT IN

WORK ORGANIZATIONS

Scott I. Tannenbaum and Gary Yuki

Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Management Department, State University of New York at Albany, Albany, New

York 12222
by University of Connecticut on 02/24/13. For personal use only.

KEY WORDS: management development, training design, training evaluation, simulation,

team training

CONTENTS

INTRODUCTION .......... . . . . . . . . .......... .. . .............. . . . ............. . . . . ....... . . ... . . . . . . . . 400

TRAINING NEEDS ANALySIS ......................................................... . . ...... . 400
Organizational Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . .. . .. . . . . . . . . . . . . . . .. . . . . . . . . 401
Task Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Person Analysis . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . .. . .... . . . . . . . . . . . . . .... . . . . . . . . . .. . . .... . . . . . . 403
DESIGN OF TRAINING ........................................................................... . 403
TRAINING METHODS ... . ................ . . .. ...... , ... ... . . ..... ........ . . . . . .. . . . ...... . . .. . . . . . 406
Simulations and Games . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
High-Technology Methods . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
Behavior Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
TRAINEE CHARACTERISTICS ........... . . . . ... . .......... . . . . . .................. . . ........... . 412
T rainee Ability and Sk ills . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
T rainee Mot ivation, Attitude, and Expectations ............ . . . ................ . . . . . . ....... 414
Aptitude- Treatment Interactions . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
PRETRAINING ENVIRONMENT ........... . ................... . . . ..................... . ....... . 417
Environmental Cues and Signals ...... . . . . . . . ... . . . . . . . . . . . . . . . . . . 417
c 418
i:r�;�:r {::u;. �;� �7�� � : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :: : : : : : : : : :
n r r t . 420
POSTTRAINING ENVIRONMENT . . . ................ . . . . . . . ...................... . . ............ . 420
Transfer En vironment ... . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Posttraining Activities . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
TRAINING EVALUATION ................... . . . . .... .................... .. . . . . . . ................ . 423
E valuation Design and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . 423
Criterion Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
TRAINING FOR SPECIFIC POPULAnONS ... . . . . . ............. . . . .. . ...... . ..... . ........ . 427
Management Development . . . . . . . . . . . . .. . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . 427
Team T raining . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
CONCLUDING COMMENTS ................. .................................................... . 432

399
0066-4308/92/0201-0399$02.00
 400 TANNENBAUM & YUKL

INTRODUCTION

This is the fourth review of training and development to appear in the Annual
Review of Psychology. The earlier reviews were by Campbell (1971), Gold­
stein (1980), Wexley (1984), and Latham (1988). Our review focuses on the
scientific literature on training and development in an organizational context
during the years from 1987 through February, 1991. We describe current
practices and emerging trends in training in work organizations . We describe
continuing trends and prevailing approaches but pay special attention to new
concepts and methods. OUI review is selective rather than comprehensive. We
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

do not attempt to review literature on some of the subjects emphasized four

years ago by Latham (1988), including the historical background of training,
by University of Connecticut on 02/24/13. For personal use only.

training in other cultures, training of raters , and leadership training. We focus

on training design, selected training methods, trainee characteristics, the
pretraining and posttraining environments, and to a lessor extent, on training
needs analysis and evaluation . Recent developments that receive special
attention include cognitive learning theories, high technology training
methods, team training, and on-the-job managerial development.
Following Latham's (1988) example, we do not attempt to make a systema­
tic review of the voluminous literature in practitioner-oriented publications.
Nor do we review the literature on related but distinct subjects such as
organizational development, socialization in organizations , motivational in­
terventions to change behavior, and training administration. Literature in
instructional psychology, management education, and motor skill training is
cited when especially relevant, but we do not examine the literature in those
related disciplines in detail.
Our review is primarily descriptive regarding current training theory and
research, but it is prescriptive regarding future training research . Whenever
appropriate, we identify research gaps and highlight issues that need more
research or a different research orientation.

TRAINING NEEDS ANALYSIS

The importance of conducting a thorough needs analysis is well accepted in

the training literature . A properly conducted needs analysis yields information
helpful to the development of instructional objectives and training criteria.
Unfortunately, only 27% of companies surveyed by Saari et al (1988) re­
ported having procedures for determining the training and educational needs
of their managers. A thorough description of training needs-assessment
methodology can be found in Goldstein et al (1991). Only a limited amount of
empirical work on training needs analysis has appeared in the last several
years, although research in job analysis (e. g . Cranny & Doherty 1988) and
performance assessment (e . g. Sackett et al 1988) has implications for assess-
 TRAINING & DEVELOPMENT 40 1

ing training needs . The most prevalent framework for considering training
needs continues to be McGehee & Thayer' s ( 1 96 1 ) categorization of organi­
zational , task, and person analysis. In a conceptual development, Ostroff &
Ford ( 1 989) suggested extending McGehee & Thayer' s framework to in­
corporate different levels of analysis explicitly . Their extension provides
ideas to guide future research, but for purposes of summarizing the recent
research on training needs analysis we rely on McGehee & Thayer's original
categorization.
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Organizational Analysis

The original purpose of organizational analysis as described by McGehee &

by University of Connecticut on 02/24/13. For personal use only.

Thayer ( 1 96 1 ) was to provide information about where and when training was
needed in an organization . Over the last several years , organizational analysis
has been reconceptualized �s an examination of systemwide components that
determine whether a training program can yield behavior change back on the
'
job (Goldstein 1 99 1 ) . Along these lines, Rouillier & Goldstein ( 1 99 1 ) have
conducted research on assessing an organization's or unit's transfer climate.
Their work is discussed in the section below on the posttraining environment.
One recurring theme, particularly in the practitioner literature, is the need
to link training and organizational strategy (see Sonnenfeld & Peiperl 1 988;
Schuler & Jackson 1987). Training courses should support the strategic
direction of the organization, and training objectives should be aligned with
organizational goals. For example, an organization undergoing downsizing
should have a different training and development focus than an organization
actively involved in geographic expansion, product development, or quality
enhancement (see London 1987). The American Society for Training and
Development (ASTD) and the Work in America Institute completed projects
that examined the link between training and company strategy in many
organizations. Examples and case descriptions can be found in Carnevale et al
(l990b), Casner-Lotto and Associates (1989), and Rosow & Zager ( 1 988).
As more organizations assume global strategies and operate in international
arenas, the nced for effcctive cross-cultural development continues to grow
(Ronen 1 989; Tung 1 988; Von Glinow & Milliman 1 990). Research has
shown that employees who receive overseas assignments often return pre­
maturely or show low levels of effectiveness (e.g. Black 1 988) . However, the
use of cross-cultural training and development in US organizations has been
limited (Finney & Von Glinow 1 988; Tung 1988). Black & Mendenhall
( 1 990) reviewed the empirical research on cross-cultural training and con­
cluded that it can be effective in developing skills, facilitating cross-cultural
adjustment, and enhancing job performance. Noting that most of the previous
research has lacked theoretical grounding, they presented a framework based
on social learning theory to guide future research efforts.
 402 TANNENBAUM & YUKL

Task Analysis
A task analysis identifies the nature of the tasks to be performed on the job
and the knowledge, skills, and abilities (KSAs) needed to perform these tasks.
Increasingly rapid technological changes can modify task requirements,
which in tum can influence knowledge, skill, and ability requirements.
Staying ahead of the technology curve can be difficult. Schneider & Konz
(1989) described a procedure for anticipating future training needs by having
subject matter experts project how the job will change and how KSA require­
ments will be affected. Campbell (1988) suggested collecting hypothetical
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

critical incidents that might be expected under future conditions. He noted

that this approach would require judges to express desired and undesired
by University of Connecticut on 02/24/13. For personal use only.

performance in concrete terms. While the logic behind future-oriented job

analysis is clear, its applicability and usefulness are still untested.
Rapid changes in job requirements are increasing the importance of crosS­
job retraining. Lance and his associates (e.g. Lance et al 1991) examined
methods for estimating cross-job retraining times for different tasks. Mum­
ford et al (1987) examined the validity of a measure of occupational learning
difficulty . Sparrow (1989) reported on the use of a measure of Position
Analysis Questionnaire (PAQ) profile similarity to predict the transfer of
learning across jobs. Each of these methods is designed to facilitate planning
by assessing how easily employees can be trained to assume new responsi­
bilities.
Howell & Cooke (1989:123) noted that work requirements at all levels are
becoming cognitively more demanding. "What were once highly structured
tasks may now call for inference, diagnosis, judgment, and decision­
making." Howell & Cooke called for an expanded view of task analysis to
include an examination of the cognitive processing and learning requirements
necessary to perform job tasks. Some authors have even suggested that the
Instructional System Development (ISD) methodology so prevalent in the
military should be modified to reflect what is known about changing cognitive
demands (Ryder et al 1987).
Campbell (1988) emphasized the need for a better understanding of what is
meant by competent or expert performance. Identifying differences in the way
experts and novices approach a task and process information could yield
insights valuable in determining what to include in training and how it should
be presented (Kraiger 1988). The most common methods used in the cogni­
tive psychology and artificial intelligence literatures for eliciting this informa­
tion from experts are the analysis of "think-aloud" verbal protocols and
psychometric scaling techniques (e.g. Cooke & Schvaneveldt 1988). Green &
Gilhooly (1990) studied how novices learned to use a statistical software
package. They used verbal protocols to identify differences in learning strat­
egies employed by fast and slow learners. A group trained to use the "effec­
tive" learning strategy performed better and retained more information than an
 TRAINING & DEVELOPMENT 403

alternate training group or a control group. The application and translation of

cognitive-task analysis concepts for training purposes are not as well de­
veloped as traditional methods. However, this approach holds great promise
as a supplemental tool for training analysts.

Person Analysis

Person analysis focuses on identifying who should be trained and what

training is needed by an individual. Inadequate person analysis can result in
training targeted to an inappropriate level or to the wrong people. Ford & Noe
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

( 1 987) studied self-assessed training needs and found small but significant
differences for managers with different job levels, functions , and attitudes
towards the utility of training. Lower-level managers reported higher needs
by University of Connecticut on 02/24/13. For personal use only.

for administrative skills than did middle managers. Managers who perceived
training to be worthwhile reported greater need for training in quality-control
skills.
The training needs of older employees have received more attention lately.
Stems & Doverspike ( 1 988, 1989) reviewed the literature on training and
retraining of older employees. They identified some of the factors that seem to
result in better performance for older individuals and suggested areas for
further research. Similarly , London & Bassman ( 1 989) examined the retrain­
ing of midcareer workers , including those experiencing career plateauing,
midlife crisis, and job loss.
New employees also have unique training needs , and formal training
programs for new employees often fail to present material at the appropriate
level of difficulty and job specificity (Feldman 1 988) . When the diagnosis of
recruits' strengths and weaknesses is inadequate, many organizations will
provide training at the lowest common denominator, resulting in suboptimiza­
tion of training effectiveness (Feldman 1 989) .
Finally, in addition to determining who needs training, person analysis can
be used to assess whether employees have the prerequisite attitude, knowl­
edge, and motivation to benefit from training. Individuals who lack basic
skills or motivation prior to training are less likely to succeed and may require
remedial preparation prior to entering a specific training program.

DESIGN OF TRAINING

After the training needs analysis is complete and training objectives have been
identified, the next step is to determine how training will be accomplished.
Design of training should take into account learning objectives , trainee char­
acteristics, current knowledge about learning processes, and practical con­
siderations such as constraints and costs in relation to benefits.
In a comprehensive review of the organizational training literature, Bald-
 404 TANNENBAUM & YUKL

win & Ford ( 1 988) found that much of the research on learning principles
(e.g. identical elements, stimulus variability, conditions of practice) was done
with college students on short-term memory tasks and simple motor tasks. As
a result, these learning principles have limited utility for designing training to
develop the complex skills required in most organization jobs. Likewise,
Campbell ( 1988) concluded that the current state of the art regarding training
principles does not provide precise guidelines for design of training. How­
ever, he noted that even the consistent use of the guidelines we already have
would contribute significantly to training effectiveness. The following guide­
lines appear most useful:
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

1 . The instructional events that comprise the training method should be

consistent with the cognitive, physical, or psychomotor processes that lead to
by University of Connecticut on 02/24/13. For personal use only.

mastery. For example, the training method should guide the learner to the
most appropriate encoding operations for storing information in memory.
2. The learner should be induced to produce the capability actively (e . g .
practice behaviors , recall information from memory, apply principles i n doing
a task) . The more active the production the greater the retention and transfer
(e.g . restating or applying principles rather than just recalling them, adapting
behavior to varying situations rather than just imitating it repeatedly in the
same situation. )
3 . All available sources o f relevant feedback should b e used, and feedback
should be accurate, credible , timely, and constructive.
4. The instructional processes should enhance trainee self-efficacy and
trainee expectations that the training will be successful and will lead to valued
outcomes. For example, training should begin with simple behaviors that can
be mastered easily, then progress to more complex behaviors as trainees
become more confident.
5. Training methods should be adapted to differences in trainee aptitudes
and prior knowledge .
Our knowledge about training design has been advanced by recent de­
velopments in the fields of cognitive psychology (Anderson 1985) and in­
structional psychology (Gagne & Glaser 1 987; Pintrich et aI 1 986) . Cognitive
approaches to learning supplement behavioral approaches and provide addi­
tional insights into the way trainees acquire knowledge and learn skills
(Howell & Cooke 1 989). Ackerman & Kyllonen ( 1 99 1 ) summarized the work
of Anderson ( 1 985, 1987) and others on the stages of skill acquisition,
highlighting the progression through declarative knowledge (knowledge about
facts and things, "what to do") , knowledge compilation (integration of facts)
and procedural knowledge (knowledge about "how to do things"). Others
have addressed the importance of conditional or tacit knowledge (knowledge
about when and why to do things) which should facilitate application of
training back on the job (Cassidy-Schmitt & Newby 1 986) . Research on a
 TRAINING & DEVELOPMENT 405

taxonomy of learning skills has been conducted by Kyllonen & Shute ( 1 989) .
Other writers have described research on metacognition , a term that refers to
mental processes involved in acquiring knowledge, interpreting feedback, and
learning from experience (Clark 1 988; Kanfer & Ackerman 1 989, Swanson
1 990) . In reviewing the developments in the field Thayer ( 1 989:46 1 ) stated
that "it will be hard to write training texts in the future without the use of such
cognitive expressions as declarative knowledge, procedural knowledge, auto­
maticity, schema, scripts , and frames ."
The cognitive approaches are especially useful for guiding the design of
training on tasks involving cognitive processes such as monitoring , problem
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

solving , and decision making. Howell & Cooke ( 1 989) described some of the
insights for training design gained from understanding the nature of cognitive
by University of Connecticut on 02/24/13. For personal use only.

processes involved in learning complex tasks .

AUTOMATIC PROCESSING Performance in a complex task with a mix of

routine and nonroutine elements can be enhanced by having a trainec initially
overlearn the routine task elements. Repeated practice of only these elements
under conditions that ensure a correct response causes the behavior to shift
from conscious to automatic control, thereby conserving mental capacity to
perform the nonroutine tasks that require more mental capacity and cannot be
automated easily. If the routine elements are not initially automatized, they
will compete for attention with the complex elements and inhibit learning of
these elements . If the routine elements are automatized early but are learned
incorrectly ("bad habits") , they will likewise inhibit development of com­
petence in performing the task.

MENTAL MODELS AND SCHEMATA Learning of a complex task can be

facilitated by helping the learner develop an accurate and efficient mental
conceptualization of the material that must be understood before the task can
be performed correctly. Retention can be facilitated by use of encoding
methods such as mnemonics , imagery, and cues that relate information to the
leamer's existing knowledge . Conceptual learning can be increased by pro­
viding relevant category systems, coding guidelines, diagrams, analogies,
and conceptual models . The manner in which material is presented can foster
the formation of appropriate mental models (e. g. Caplan & Schooler 1 990;
Zeitz & Spoehr 1989) .

MET ACOGNITION AND LEARNING SKILLS Learning of tasks that require

analytical processing is facilitated by helping learners monitor their own
progress and evaluate what they know and don't know . An important discov­
ery of the research on metacognition is that feedback about performance
outcomes may be of little value if the learner is using an inappropriate mental
 406 TANNENBAUM & YUKL

model and the feedback does not provide guidance for recognizing why it is
inappropriate and developing a better one. Learning can be facilitated by
showing learners how to seek and utilize relevant feedback about their
strategy for doing a task (e.g. what strategy was used, what was done
correctly and what mistakes were made, what might have been done instead).
Meta-cognitive skills, along with other learning strategies (e .g. if-then rules,
working backwards) distinguish proficient from nonproficient learners. There
is some evidence that these skills can be taught or that learning can be
facilitated by embedding within the training diagnostic probes and clues to
help the learner analyze and interpret performance feedback (Clark 1988;
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Derry & Murphy 1986) . Additional work is needed to examine the utility of
cognitive concepts for facilitating transfer of training in addition to skill
by University of Connecticut on 02/24/13. For personal use only.

acquisition.
Training design is advanced by continuing developments in learning theo­
ries of all types, and these developments have been reviewed in detail by
Weiss ( 1 99 1 ) . Research on motor skills has continued to investigate the
traditional design parameters (e.g. distribution of practice, knowledge of
results , conditions facilitating retention and transfer) , and a detailed review of
this literature can be found in Adams (1987). Our knowledge has been
advanced also by the ongoing work in the military on design of training
systems. An example is the work by Morrison & Holding ( 1990), who
described recent efforts to refine design instructional guidelines developed for
the military in the 1 970s. Morrison & Holding reviewed theory and research,
including cognitive psychology , relevant to several design issues such as
identification of relationships among training objectives , sequencing of train­
ing components, selection of training methods , and allocation of training time
to various training activities . Tentative design guidelines based on the current
state of knowledge were offered as hypotheses to guide future research.
The effectiveness of different forms and amounts of learner self-control
over training is an issue that has become increasingly important owing to the
growing interest in self-management (Frayne & Latham 1987; Kanfer &
Hagerman 1987; Latham & Frayne 1989; Manz & Sims 1 989) and the
availability of interactive training methods that allow individualized instruc­
tion and increased trainee control over learning. Kinzie ( 1990) noted that
significant questions have emerged about the capability of learners to manage
their own learning in an effective way. (Will they make good choices, will
they be motivated to learn?) Kinzie reviewed the limited research relevant to
this question and identified areas where more research is needed.

TRAINING METHODS

A major question in the design of training is the selection of training methods .

The literature on evaluation of training methods continues to increase. As
 TR AINING & DEVELOPMENT 407

Campbell (1988) noted recently, by far the largest category of studies in­
volves comparison of a single training method to another method or to a
control condition with no training. The purpose of such studies is usually to
demonstrate that a particular method "works" or that it is "superior" to another
method. We agree with Campbell that research of this type has only marginal
utility for improving our understanding of training. All of the well-known
training methods have been shown to be effective for some learning objec­
tives, but "demonstration" studies do not reveal why a particular method or
combination of methods facilitates learning or how the method can be used
more effectively. Even a study that pits one method against another can be
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

inconclusive as it is likely that the relative effectiveness of different methods

will depend on the purpose and objective of the training , the attributes of the
by University of Connecticut on 02/24/13. For personal use only.

trainees, and the effectiveness criteria selected. Here we briefly review

research on training methods that have received the most attention in the past
few years, namely simulations and games , high technology methods , and
behavioral role modeling.

Simulations and Games

Faria (1989) found in a survey of training managers , consulting firms, and
business schools that simulations and games are now widely used as a training
method. Simulations are also a popular training method in the armed services .
Thornton & Cleveland (1990) pointed out that simulations vary widely in
complexity of issues and number of participants, ranging from relatively
simple simulations for an individual (e.g. in-basket, some computerized
business games) or small group (e. g . dyadic role play, leaderless group
discussion, group problem-solving exercises), to moderately complex com­
puterized business games , combat simulations , and large-scale simulations of
a particular type of organization. Looking Glass (McCall & Lombardo 1982;
Van Velsor et al 1989) is the best known of the large-scale behavior sim­
ulations, but the development of industry-specific simulations has increased
in recent years. There are now simulations for a wide variety of industries
(Keys 1987; Lepsinger et al 1988; Stumpf & Dunbar 1990).
The amount of research on simulations and games has increased dramati­
cally in the last several years . Research on simulations and games used for
management education has been reviewed by Hsu (1989), Keys & Wolfe
(1990), Thornton & Cleveland (1990), and Wolfe (1990). Aspects of sim­
ulations that may influence the success of the training include the complexity
of the simulation, scope of functional and environmental activities, size and
composition of teams, scheduling of activities (e.g. concentrated in a 2-3 day
period versus 1-2 hour sessions distributed over several weeks) , type of
preliminary preparation and conceptual learning, type of feedback and de­
briefing, use of decision support systems and tutoring aids based on expert
systems , and the quality of game administration.
 408 TANNENBAUM & YUKL

Despite the growing research literature on simulations, many important

research questions remain unresolved. More research is needed to determine
what types of learning occur and the conditions that facilitate it. Hsu ( 1 989)
reanalyzed the simulation research and concluded that most studies of multi­
team and large-scale behavioral simulations fail to evaluate the types of
learning for which these simulations are most appropriate, namely problem
solving and interpersonal skills. Furthermore, because simulations are typi­
cally used together with other training methods such as reading materials,
lecture-discussion, and demonstrations, it is difficult to disentangle their
unique contribution to trainee learning (Thornton & Cleveland 1 990).
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Evaluation is complicated also by the fact that simulations often have

multiple learning objectives, some of which are general and long-term rather
by University of Connecticut on 02/24/13. For personal use only.

than specific and immediate. There is little evidence that simulation training
results in long-term improvements in managerial effectiveness . A study by
Wolfe & Roberts ( 1 986) found that performance in a single-person game
predicted career success as a manager five years later, but the study did not
assess what, if anything, trainees learned in the game. Keys & Wolfe ( 1 990)
pointed out that game performance is not equivalent to learning; a team that
experiences difficulties and setbacks may learn more about interpersonal
processes and problem solving than a team for which financial success is
easy . There is clearly a need for more longitudinal research on how simulation
training affects immediate learning and longer-term performance.
Because management games have become something of a fad, some writers
have urged caution in their use (Jacobs & Baum 1 987) . Although it was
assumed initially that learning would occur automatically as a result of the
game experience, we are coming to realize that extensive preparation, plan­
ning, and debriefing are needed to realize the potential benefits from sim­
ulations (Keys & Wolfe 1 990) . Thornton & Cleveland ( 1 990) recommended
that complex simulations should be used only within a carefully planned
sequence of developmental experiences. In addition, the training should
include presentation of an appropriate model of effective management with
clear descriptions of relevant managerial skills, there should be an adequate
debriefing of the simulation experience, and there should be opportunities for
coaching, feedback, reinforcement, and practice of skills on the job .

High-Technology Methods
This section deals with a variety of high-technology training methods such as
computer-aided instruction, interactive videodisc instruction, and equipment
simulators. As development of technology proceeds at a rapid pace and the
cost of computers continues to decline, these high-technology training
methods are finding increasing use in industry , academia, and the military
(Lippert 1 989; London 1 989; Pursell & Russell 1 99 1 ) . Moreover, innovative
 TRAINING & DEVELOPMENT 409

media such as teleconferencing and satellite television networks are being

used increasingly to provide training to widely separated sites in a cost­
effective manner (see London 1989).
Computer aided instruction (CAl) allows individualized instruction with
the advantages of self-pacing by trainees, active practice or rehearsal, im­
mediate feedback, continuous monitoring and assessment of learning, di­
agnosis of learning problems, and remedial assistance when needed. One
major disadvantage of CAl is the extensive development time and expense,
but computer software has been developed to simplify and speed up the design
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

process and facilitate programming (e.g. CASE technology). Eberts & Brock
(1987) described recent developments in CAL They cautioned that computers
will not automatically improve training because success depends upon the
by University of Connecticut on 02/24/13. For personal use only.

adequacy of the needs assessment and instructional design, as well as on the

technology for delivering instruction. Eberts & Brock noted that improvement
of computer aided instruction depends heavily upon research in cognitive
processes.
Some progress is being made in improving computer-aided instruction,
particularly with respect to incorporating artificial intelligence, but it is too
early to reach any firm conclusions about its effectiveness (Lippert 1989).
Fletcher (1988) reviewed developments in the use of intelligent computer­
aided instruction (ICAI) by the military. He found almost no data on training
effectiveness for the nine ICAI systems he reviewed, and he urged more
evaluation of ICAI, not only to determine its effectiveness, but also to
discover why and how it facilitates learning. Despite the lack of rigorous
evaluation studies Fletcher, like most other writers who have examined ICAI,
is optimistic about its potential to improve training.
As new technology continues to be developed, progress has been made also
in linking various types of instructional technology. One promising innova­
tion of the 1980s is the linking of videodisc players with microcomputers to
create interactive videodisc instruction. Videodisc players present information
in the form of text, still photographS, videotaped pictures, computer graphics,
and multi-track sound recordings. The simpler versions of videodisc training
have a menu that allows thc trainee to selcct the type of information desired;
the more sophisticated versions permit an interactive experience with di­
agnosis, feedback, and remedial coaching for the trainee. Touch screens,
voice recognition, and optical readers provide promising alternatives to the
keyboard as a medium of interaction with the computer-tutor (De Bloois
1988; Pursell & Russell 1991). An example of the flexibility of interactive
videodisc training is provided by Alliger et al (1989), who described a
successful interactive videodisc training program that included behavior mod­
eling to train entry-level sales personnel at IBM.
Fletcher (1990) conducted a comprehensive review and meta-analysis of 47
 410 TANNENBAUM & YUKL

studies evaluating interactive videodisc instruction in the military, industry,

and higher education. The training content in these studies was very diverse,
including instruction in equipment maintenance, equipment operation , medi­
cal procedures, military operations, science education, and interpersonal
skills. Videodisc instruction was substantially more effective than con­
ventional instruction (e.g. lecture, videotaped demonstration, text, pro­
grammed text, on-the-job training) with respect to both knowledge acquisition
and job performance. The more the videodisc instruction included interactive
features such as tutorials, the greater its effectiveness . Within-group variabil­
ity in criterion scores was lower for videodisc instruction than for con­
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

ventional instruction, which indicates that the benefits were widespread rather
than attributable to large improvement in a few "star" trainees. However,
by University of Connecticut on 02/24/13. For personal use only.

Fletcher (1990) noted that aside from the question of interactivity, the studies
provide little insight into the relative contribution of the various features of
interactive videodisc technology to learning . He concluded that more research
is needed to identify design alternatives that contribute to various learning
Objectives .
Equipment simulators based on computers have been used extensively for
training aircraft pilots, and Jacobs et al ( 1 99 0) conducted a meta-analysis of
this research . Their analysis indicated that simulator training combined with
training on the actual aircraft was more effective than training on the aircraft
by itself. Simulator training was more effective when pacing was based on
each trainee' s individual progress as compared to having all trainees proceed
at the same pace. Progress in understanding how to use simulators effectively
requires evaluation of simulators in relation to specific learning objectives.
An example of this kind of research is a study by Drucker & Campshure
(1990) on army tank operations. A computer-based simulation was compared
to a table-top simulation with respect to their utility for training a variety of
skills needed by tank platoon officers. The computer simulation was superior
for some tasks but not for others .
The realism of simulators is being enhanced greatly by continuing de­
velopments in videodisc technology, speech simulation, and speech recogni­
tion. For example, Hamel et al (1989) found that realism could be increased in
microcomputer-based training for air controllers by using "off the shelf"
automated voice recognition and speech generation technology.
Advancements in networking technology in the last few years have opened
up new possibilities for large-scale simulator networking (SIMNET) .
Although simulators have been used in the past for training of small teams in
the military, simulator networking allows large groups of military trainees to
practice their skills in an interactive simulation of combat conditions (Thorpe
1987; Alluisi 1991). Since under these circumstances the opponent is not just
a computer but another military team, trainees are highly motivated by the
competition . As in real combat, the chain of command for each side plans
 TRAINING & DEVELOPMENT 411

operations, logistics, and combat support; issues commands; receives reports;

and reacts to unpredictable , rapidly changing developments. Learning occurs
at all levels, from the operators of several different weapon simulators (e.g.
tanks, helicopters, aircraft) to the commanders who plan and coordinate the
operations. Just as in the case of field simulations �ith real weapons, learning
is facilitated by feedback received in detailed "after action reviews" during
which actions and decisions taken during the simulation are analyzed to
discover what was done well, what mistakes were made, and how perform­
ance could be improved. A thorough history of the development of SIMNET
technology is provided by Alluisi (199 1 ) .
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Behavior Modeling
by University of Connecticut on 02/24/13. For personal use only.

Interest in behavioral role modeling has continued to be strong since the

pioneering research by Goldstein & Sorcher ( 1 974). A meta-analysis by
Burke & Day (1 986) of studies evaluating behavior modeling found that it
was one of the most effective training methods. Subsequent reviews continue
to support the utility of this training method (Latham 1 989; Mayer & Russell
1 987; Robertson 1 990) . However, several writers (e.g. Mayer & Russell
1 987; Parry & Reich, 1 984; Robertson 1 990; Russell et al 1 984) have
expressed concerns about the conclusions drawn from the research. One
concern is that claims for the superiority of behavior modeling are exagger­
ated and unsubstantiated (Mayer & Russell 1 987). Mayer & Russell found
little evidence that role modeling is more cost effective than other methods,
and they noted that most studies on behavior modeling examined only im­
mediate learning , not the application of the learned behaviors back on the job
or the effects on job performance. Another concern involves the type of
training for which behavior modeling is appropriate. Behavior modeling
appears useful for teaching concrete behaviors that are clearly optimal for a
particular type of task (e.g. operate equipment, assemble a machine, perform
a surgical procedure). However, it is not clear how effective behavior model­
ing is for teaching flexible adaptive behaviors or cognitive knowledge
(Robertson 1990). More research is needed to determine the types of tasks and
skills that can be trained effectively through behavior modeling.
For some types of learning, the effectiveness of behavior modeling may be
increased by using a modified form of the typical training method. Gist (1 989)
found that a variation called "cognitive modeling" may be useful for training
people to increase idea generation. Parry & Reich (1984) suggested a varia­
tion of the usual procedure of presenting concrete behavioral guidelines for
training interpersonal skills that require flexible , adaptive behavior. They
proposed that the training should emphasize general principles, and trainees
should be encouraged to devise alternative ways to apply the principles in
various situations .
 41 2 TANNENBAUM & YUKL

Some of the most useful research on behavior modeling examines how

learning is affected by different aspects of the training, such as characteristics
of the modeling demonstration, the way behavior is practiced or rehearsed,
and the type of feedback provided to participants (e.g. Hogan et a1 1 986; Mills
& Pace 1 989). One important question about model characteristics is whether
trainees should be exposed to a variety of models of varying competence.
Baldwin ( 1 99 1 ) found that exposing trainees to models of varying competencc
and models in varying situations increased generalization of skills. However,
Trimble et al ( 1 989) found that showing both positive and negative models, as
compared to showing only positive models, did not result in greater recall or
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

performance of the guideline behaviors. A negative model alone was less

effective than either a positive model or a combination of positive and
by University of Connecticut on 02/24/13. For personal use only.

negative models . Results from the two studies appear contradictory , and it is
not clear yet whether model variability improves discrimination of appropriate
behavior or increases ability to apply a general principle to varying con­
ditions.
Some progress has been made in understanding the conditions that facilitate
learning in behavior modeling, but more research is needed both on the
learning process and on the motivational process. The theoretical basis for
behavioral role modeling is Bandura's ( 1986) social learning theory . Robert­
son ( 1 990) noted that prior role modeling studies do not test all aspects of the
theory . For example, the motivational aspects of the theory , such as level of
trainee self-efficacy and perceived instrumentality of applying learned be­
haviors on the job, are rarely explored (Robertson 1 99 0; Mayer & Russell
1 987). More research is needed to identify characteristics of the training that
increase trainee self-efficacy. Some of the recent research on self-manage­
ment (e.g. Latham & Frayne 1 989; Frayne & Latham 1 987) suggests that it
would be feasible to measure changes in self-efficiency during behavior
modeling training.
In summary , some of the same concerns can be expressed regarding the
research on simulations, high-technology methods , and behavior modeling.
Each training method has demonstrated some utility, but more research is
needed to determine the types of content for which a training method is
appropriate and to discover how different aspects of the training method affect
training outcomes.

TRAINEE CHARACTERISTICS

Training effectiveness is determined in part by the thoroughness of the needs

analysis and the quality of the training design, but other factors also contribute
to training effectiveness, including the attributes of trainees. Historically,
research on trainee characteristics has focused more on selecting trainees who
 TRAINING & DEVELOPMENT 413

will pass training and less on placing individuals into appropriate training
programs, revising training to match trainee attributes, or understanding how
trainee characteristics influence training effectiveness. Empirical investiga­
tion of trainee characteristics in organizational settings is still limited, but
there appears to be a recent increase in research on the implications of trainee
characteristics for improving training effectiveness. In the next few sections
we summarize literature on trainee characteristics, including trainee abilities
and skills; trainee motivation , attitudes, and expectations; and aptitude treat­
ment interactions.
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Trainee Abilities and Skills

by University of Connecticut on 02/24/13. For personal use only.

Trainability testing attempts to predict whether a person will successfully

complete training. An example of this method is the minicourse approach
used in the AT&T system (Reilly & Israelski 1 988). In a minicourse, a
candidate is given a standardized sample of programmed training material
followed by a knowledge test to measure learning of the material. A similar
method is the work-sample test of trainability. In this method the period of
instruction is followed by a work sample test to predict subsequent trainee
performance. Unlike standard work samples, these can be used with in­
experienced applicants because they include a structured training experience.
The rationale behind these approaches is that a trainee's ability to learn
material can be assessed by placing the person in a context that approximates
the training setting. Robertson & Downs ( 1 989) conducted a meta-analysis of
studies on work-sample trainability tests and found that they predict subse­
quent training and job performance in most situations . Although the relation­
ship was positive for all subsets of validity coefficients, the greater the
interval between the work sample test and the criterion, the weaker the
relationship. In other words, trainability tests predict short-term training
success better than longer-term training success or subsequent job perform­
ance. In general, trainability testing research has value from a selection
perspective but does not shed much light on why training works or how to
improve training. Moreover, it is unclear whether trainability tests provide
any useful incremental gain over the use of cognitive ability tests in predicting
training success and subsequent job performance (Robertson & Downs 1 989) .
There is an ongoing debate among researchers about whether certain
abilities are more important at various points during the skill acquisition
process , and if so, which ones and under what circumstances (see Ackerman
1 989; Barrett et al 1 989; Fleishman & Mumford 1 989a,b; Henry & Hulin
1 987; Murphy 1989) . For example , Ackerman (1988) found that for novel,
moderately complex, but consistent psychomotor tasks, initial performance is
best predicted by general and broad content abilities; intermediate levels of
 414 TANNENBAUM & YUKL

skilled performance are best predicted by perceptual speed abilities; and late,
asymptotic performance levels are best predicted by psychomotor abilities .

Trainee Motivation, Attitudes, and Expectations

As Noe ( 1 986) suggested, a variety of factors besides ability can influence

training effectiveness , including trainee motivation, attitudes, and ex­
pectations . Some conceptual development has been done recently on the
effects of trainee characteristics. Kanfer & Ackerman ( 1 989) presented a
theoretical framework for integrating ability and motivation, and this theory
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

has important implications for the study of skill acquisition and training
performance. Their framework (based on Kahneman 1 973) depicts perform­
by University of Connecticut on 02/24/13. For personal use only.

ance as a joint function of the trainee's relative attentional capacity (cognitive

ability) and the proportion of the trainee' s total capacity actually devoted to
the task (motivation) . The model also differentiates distal (i .e. antecedent to
task engagement) and proximal motivational processes (i.e. distribution of
effort during task engagement) . Empirical work focusing on simulated air
traffic control tasks provided some support for their model. Mathieu et al
(1990) presented and tested a model that hypothesized linkages between
trainee characteristics and training outcomes based on the work of Noe
( 1986) . Additional attention should be given to understanding the rela­
tionships among trainee characteristics and their relative contribution in
facilitating learning, skill acquisition, and transfer.

MOTIV ATION It is widely accepted that learning and transfer will occur only
when trainees have both the ability ("can do") and volition ("will do") to
acquire and apply new skills (Wexley & Latham 1 99 1 ; Noe 1 986) . Yet until
recently there was only a limited amount of research on the "will do" factors
in the training context.
Baldwin & Ford ( 1 988) and Noe ( 1 986) suggested the use of an expectancy
framework for studying training motivation. Mathieu et al ( 1 990) used ex­
pectancy-type motivation measures in a study of proofreading training for
clerical employees. As predicted, they found that trainees with higher pre­
training motivation demonstrated greater learning and more positive reactions
to training , even after controlling for educational differences. Williams et al
( 1 99 1 ) , Tannenbaum et al ( 1 99 1 b), and Baldwin et al ( 1 99 1 ) all provided
additional empirical evidence for a positive relationship between trainee
motivation and training outcomes . Kanfer ( 1 99 1) reviewed and summarized
developments in motivation theory as they pertain to individuals in organiza­
tions. The application of motivation theory to the training environment could
enhance our understanding of training effectiveness. In an effort to clarify
what trainees believe are the rewards of training, Nordhaug ( 1 989) conducted
interviews with Norwegian employees and asked whether their participation
 TRAINING & DEVELOPMENT 415

i n training had contributed to certain outcomes . Factor analysis identified

three dimensions of payoffs from training including "motivation to learn"
(e.g. increased interest in the subjects of the course) , "career development"
(e.g. increased autonomy, promotion) , and "psychosocial development" (e. g.
increased self-confidence) . Although based strictly on retrospective recall ,
this exploratory study provides some information on employee perceptions
about the instrumentality of training . However, Nordhaug did not assess what
trainees expect or desire from the training.
Feldman ( 1 989) noted that "the whole issue of expectations about training
itself has to be explored more fully" (1989:38 1 ). Tannenbaum et al ( 1 99 I b)
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

examined the expectations and desires of naval recruits regarding their up­
coming training experience (e.g. degree of challenge, opportunity to prac­
by University of Connecticut on 02/24/13. For personal use only.

tice) . They found that trainees who had their pretraining expectations and
desires met (referred to as "training fulfillment") developed greater posttrain­
ing commitment, self-efficacy, and motivation even after controlling for
pretraining attitudes and ability, training performance, and trainee reactions .
Eden ( 1 990) summarized the research on the Pygmalion and Galatea effects
and concluded that enhancing a trainee's performance expectations can have
an impressive effect on trainee achievement. In fact, this research showed that
trainee achievement can be increased by heightening the trainer's expectations
(e.g. telling the trainer that a trainee has superior capabilities) .

SELF-EFFICACY Self-efficacy i s another important trainee construct that has

received increased attention in the training literature. Self-efficacy refers to a
belief in one's ability to perform a specific task, and it is a central concept in
social learning theory (Bandura 1986). Self-efficacy can be considered a
potential antecedent of training effectiveness, because individuals with high
self-efficacy tend to outperform individuals with low self-efficacy (Taylor et
al 1 984; Bouffard-Bouchard 1 990). That is, individuals who enter training
believing they are capable of mastering the training content are likely to learn
more during the training. Gist et al ( 1 989) studied managers and administra­
tors undergoing two types of training in the use of computer software.
Trainees with higher self-efficacy prior to training and those who reported
higher self-efficacy at the midpoint of training performed better than their
low-self-efficacy peers on a timed computer task at the end of training.
Individuals who leave training with the belief that they can successfully
perform the task they have been trained to do should be more resilient when
they encounter obstacles in the transfer environment (Marx 1 982) and more
willing to try new things (Hill et al 1 987). As noted earlier, increased
self-efficacy is probably one reason for the effect of behavior modeling on
trainee behavior. Thus, self-efficacy can be regarded as a predictor of training
success , as a process variable during training, or as a desirable outcome of
training.
 416 TANNENBAUM & YUKL

OTHER TRAINEE CHARACTERISTICS Other trainee characteristics that may

have relevance for understanding the training process include problem solving
style (e.g. Basadur et al 1 990) , action orientation (Kuhl 1 985) , openness to
experience (Barrick & Mount 199 1 ) , trainee attributions of success and failure
(Campbell 1 988) , need for achievement, and goal orientation (See Dweck
1 986; Elliott & Dweck 1 988). The last characteristic is a concept in the
educational literature that can be applied to the training context. Elliott &
Dweck ( 1 988) proposed that individuals can pursue two different types of
goals in achievement situations: (a) performance goals, in which individuals
seek to maintain positive judgments about their ability; and (b) learning goals,
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

in which individuals seek to increase their ability or to master new tasks.

Although most of the research on goal orientation has been conducted with
by University of Connecticut on 02/24/13. For personal use only.

school children, it is reasonable to assume that trainees with a learning goal

orientation will approach training quite differently from trainees with a per­
formance orientation . Research is needed on the influence of goal orientation
with adult trainees.
In sum, trainee self-efficacy, motivation , and expectations appear to be
central constructs in understanding training effectiveness, and other trainee
characteristics (e. g . goal orientation) hold promise as well. Dynamic trainee
characteristics that can be influenced before, during, or after the training
process should receive particular attention because these have the greatest
potential for enhancing our understanding of how and why training works.
However, measures of these constructs should be refined to fit the training
context. For example, self-efficacy and trainee motivation must be operation­
a1ized more clearly, and distinctions must be made between motivation to
attend , motivation to learn, and motivation to transfer, with the latter perhaps
focusing on behavioral intentions (see Tubbs & Ekeberg 1 99 1 ) . Additional
research on motivational antecedents to training is needed . We agree with
Campbell ( 1 988) that "we have barely scratched the surface here, and much
more remains to be done" (p. 209).

Aptitude-Treatment Interactions
An aptitude-treatment interaction occurs when one training program is not
uniformly superior to another, but instead the programs have a differential
impact on trainees of different aptitudes . In this context, an aptitude can be
broadly defined as any characteristic of trainees that determines their ability to
profit from instruction, including abilities, skills, knowledge, and even prev­
ious achievement (Cronbach & Snow 1 977) . Aptitude-treatment interactions
have important implications for the design of training . In order to adapt
training methods to differences in trainee aptitudes and prior knowledge, it is
necessary to understand the interactions between training methods and apti­
tudes. In addition , aptitude-treatment interactions provide information for
placing people into training conditions that will maximize training effective-
 TRAINING & DEVELOPMENT 417

ness. For example, Savage et al ( 1 982) found the use of regression-based

techniques to assign trainees to one of two training programs reduced training
time by 50% over random assignment of trainees .
Aptitude-treatment interactions have received considerable attention in the
instructional and education psychology literatures (Pintrick et al 1 986; Acker­
man et al 1989; Snow 1 989) , but they have not been studied extensively in
organizational settings. Overall, the empirical support for consistent interac­
tion effects has been limited. For the training context, one of the most
promising interactions may be between general academic ability and the
complexity or structure of the instructional program. According to Snow &
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Lohman ( 1984), high-ability students tend to benefit more from programs

with less structure (e.g. emphasizing independent acquisition of knowledge)
by University of Connecticut on 02/24/13. For personal use only.

and greater complexity, whereas low-ability students benefit more from

explicit, structured instruction. Campbell ( 1 988) suggested that within organi­
zational contexts , trainee achievement and experience may also interact with
training complexity or difficulty . As he noted, if this is true , we should pay
more attention to measuring the existing achievement level of potential
trainees and tailoring the training accordingly during the design process.
In sum, although aptitude-treatment interactions may have important im­
plications for personnel training, little empirical research has been conducted
in organizational settings. To date, the potential of aptitude-treatment in­
teractions to be useful in organizational training applications remains un­
fulfilled.

PRETRAINING ENVIRONMENT

Accumulating evidence suggests that events prior to training (i.e. the pretrain­
ing environment) can influence training effectiveness. Management actions
provide cues and signals that influence employee motivation. The pretraining
environment contains many cues about training; some are conveyed by man­
agers but others are conveyed by peers or reflected in organizational policies
and practices. Employees start to learn about the way training is viewed in the
organization early in the socialization process (Feldman 1989) and continue to
gather information with each course they attend . Some actions signal to
trainees whether training is important (e .g. supervisory and peer support,
resource availability, and posttraining follow-up) . Other actions reveal to
employees the amount of control, participation, or input they have in the
training process (e. g . advance notification, participation in needs assessment,
and degree of choice in attendance).

Environmental Cues and Signals

In a study across five companies, Cohen ( 1 990) found that trainees with more
supportive supervisors entered training with stronger beliefs that training
 418 TANNENBAUM & YUKL

would be useful. Supervisors can show their support for an upcoming training
course by discussing the training with the employee , establishing training
goals, providing release time to prepare , and generally encouraging the
employee . Cohen also found that trainees who set goals prior to training
entered training with higher levels of motivation to learn. Unfortunately , the
cross-sectional nature of the study makes it impossible to determine whether
setting goals enhanced motivation to learn or whether the highly motivated
trainees were more likely to set goals.
Baldwin & Magjuka ( 1 99 1) found that trainees who entered training ex­
pecting some form of follow-up activity or assessment afterward reported
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

stronger intentions to transfer what they learned back to the job. The fact that
their supervisor would require them to prepare a posttraining report or un­
by University of Connecticut on 02/24/13. For personal use only.

dergo an assessment meant that they were being held accountable for their
own learning and apparently conveyed the message that the training was
important. The pretraining environment can act also as an inhibitor of training
effectiveness . Mathieu et al ( 1990) found that trainees who reported many
situational constraints in their job (e.g . lack of time, equipment, and re­
sources) entered training with lower motivation to learn. These trainees had
little incentive to learn new skills in an environment where the skills could not
be applied.
A clearer distinction is needed between effects of cues associated with a
particular training course (e. g . Did my manager meet with me to develop
goals?) and effects of cues associated with the overall training culture (e.g.
Does my company tend to promote from within; is training considered a
reward or a punishment?) . Future research should examine the relative im­
portance of specific and global cues on trainee motivation.

Trainee Input and Choice

Wlodkowski ( 1 985) suggested that involving employees in decisions about

the training process could enhance their motivation to learn. In the pretraining
context participation may include informing trainees about training content in
advance , soliciting trainee preferences for training content and methods,
and/or allowing trainees to decide which courses to attend.
Saari et al ( 1 988) reported that about half to three quarters of all managers
who attend training are briefed prior to training. However, the depth and
nature of the pretraining information can vary. Baldwin & Magjuka ( 199 1 )
found that those trainees who had received information about the training
ahead of time reported a greater intention than others to apply what they
learned back on the job. In an interesting study of race relations training,
Alderfer et al ( 1 99 1 ) found that trainees who had received more information
prior to the training had more positive reactions at the conclusion of the
training. In a study of supervisory training, Hicks & Klimoski (1987)
 TRAINING & DEVELOPMENT 419

manipulated the information trainees received prior to training and found that
trainees who received a realistic description of the training reported more
motivation to learn than trainees who received a traditional positive portrayal
of the training. However, no difference was found in actual learning. In
general , it appears that providing trainees with advance notification may be
helpful, but it is not clear whether notification enhances feelings of involve­
ment, creates realistic expectations, indicates importance, or allows time for
trainees to align their personal goals with the training goals.
Wlodkowski ( 1985) suggested that trainee participation in the training
needs-assessment process should enhance motivation to learn, but we know
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

of no research that has examined this proposition directly. Noe ( 1 986) and
Campbell ( 1 988) proposed that trainees will be more motivated if they believe
by University of Connecticut on 02/24/13. For personal use only.

that the assessment of their strengths and weaknesses is accurate . Noe &
Schmitt ( 1986) tested this proposition and found a weak relationship between
perceived assessment accuracy and pretraining motivation. In contrast, Wil­
liams et al ( 1 99 1 ) reported a strong link between perceived assessment
accuracy and posttraining motivation to transfer.
An important question with respect to participation is whether trainees can
choose which training they attend. Voluntary participation has been shown to
be related to higher motivation to learn, greater learning, and more positive
trainee reactions than mandatory attendance (Cohen 1 990; Hicks & Klimoski
1 987; Mathieu et aI 1 990). In contrast, Baldwin & Magjuka ( 199 1 ) found that
engineers who perceived training to be mandatory reported greater intentions
to apply what they learned back on the job than engineers who viewed their
attendance as voluntary. One explanation for the discrepancy in results be­
tween this study and the others may be the different attitudes towards training
in the various companies. Over 80% of the engineers reported that their
previous training experiences with the company had been favorable or very
favorable . Baldwin & Magjuka suggested that by making training mandatory
the company was signaling to the employees which courses were most
important. In the study by Hicks & Klimoski only 1 7 % of the employees who
were given a choice whether to attend the training actually did so, suggesting
that employees in that organization perceived training to be of low value.
When training is not valued, mandatory attendance may be demoralizing.
Baldwin et al ( 1 99 1 ) showed that there is a clear difference between
soliciting a trainee' s opinion and taking that opinion into account. In a lab
experiment, they found that allowing trainees to specify what training they
wanted increased their motivation to learn , provided they were given the
training of their choice . However, trainees who were allowed to "choose" a
course but were then assigned to a different course were less motivated and
learned less than the trainees who did not participate at all in the choice of
training. B aldwin et al explained their findings based on procedural justice
concepts of "fair process" and "frustration effects" .
 420 TANNENBAUM & YUKL

Pretraining Preparation
In addition to the signals provided to trainees in the pretraining environment,
specific preparatory activities may also occur. Trainee deficiencies in self­
efficacy, learning skills, or reading skills can create problems during training
(Mumford et al 1 988). As an alternative to selecting only trainees with the
necessary aptitudes or designing training to the lowest common denominator,
pretraining activities can be used to enhance self-efficacy or prepare trainees
with the necessary learning , reading , meta-cognitive , or other basic skills
(Carnevale et al 1 990a).
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

POSTTRAINING ENVIRONMENT
by University of Connecticut on 02/24/13. For personal use only.

The effectiveness of a training program can be influenced by events that occur

after a trainee returns to the job. Some employees leave training with new
skills and with strong intentions to apply those skills to their job, but limita­
tions in the posttraining environment interfere with the actual transfer of
training. As Ajzen ( 1 985) noted, the degree to which intentions are converted
into acts and products is partially determined by various inhibiting and
facilitating control factors. The personal skills, ability, and willpower that
trainees possess at the conclusion of training are potential determinants of
transfer. In addition, elements of the posttraining environment can encourage
(e.g. rewards, job aids), discourage (e.g. ridicule from peers), or actually
prohibit the application of new skills and knowledge on the job (e.g. lack of
necessary equipment) . In this section we examine developments regarding the
influence of the posttraining environment.

Transfer Environment
Transfer of training can be defined as the extent to which trainees effectively
apply the knowledge, skills, and attitudes gained in a training context back to
the job. Several writers have discussed conceptual issues in understanding
transfer of training. Baldwin & Ford ( 1 988) emphasized the distinction
between (a) generalization, or the extent to which trained skills and behaviors
are exhibited in the transfer setting, and (b) maintenance, or the length of time
that trained skills and behaviors continue to be used on the job . They
presented several hypothetical maintenance or decay curves that illustrate the
changes that can occur in the level of knowledge, skills, and behaviors
exhibited on the job as a function of the time elapsed since the completion of
training.
Laker ( 1 990) suggested that the concept of transfer distance, prevalent in
the educational literature (e.g. Butterfield & Nelson 1 989) , may be useful for
examining generalization in organizational settings. Transfer distance refers
to the extent to which a trainee applies what is learned to job situations similar
 TRAINING & DEVELOPMENT 42 1

to the training situation (near transfer) or different from it (far transfer). Ford
( 1 990) noted that Laker did not consider other aspects of generalization that
might be equally valuable in examining transfer, such as lateral and vertical
transfer, literal and figural transfer, and specific and nonspecific transfer.
Increased conceptual clarity regarding transfer can help us understand how to
modify the work environment to facilitate transfer of training.
While practitioners acknowledge that transfer is dependent on the posttrain­
ing environment, little empirical research is available to guide practice.
Baldwin & Ford ( 1 988) found only seven studies that examined the influence
of the work environment on transfer of training and none that attempted to
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

change the work environment. In particular, they noted that supervisory

support is considered a key environmental factor that can affect the transfer
by University of Connecticut on 02/24/13. For personal use only.

process . In the posttraining environment, supervisor support could include

reinforcement, modeling of trained behaviors , and goal-setting activities .
Baldwin & Ford called for better identification and operationalization of key
work-environment variables that affect trainee perceptions and influence
transfer of training.
Along these lines, Rouillier & Goldstein ( 1 9 9 1 ) presented an encOl�raging
study that examined the organizational climate for transfer of training. They
identified transfer climate components and classified them into situational
cues and consequences , based on an organizational behavior model proposed
by Luthans & Kreitner ( 1985). Situational cues in the work environment
included (a) goal cues that serve to remind trainees to use their training; (b)
social cues, including the behavior and influence processes exhibited by
supervisors, peers and/or subordinates; and (c) task and structural cues,
including the design and nature of the job itself. Consequences included
positive and negative feedback and punishment. Trainees were assistant
managers who completed a week-long training program and were then ran­
domly assigned to one of 1 02 organizational units. In units with a more
positive transfer climate, trainecs demonstrated significantly more trained
behaviors even after controlling for learning and for unit performance .
S ituational cues and consequences each contributed additional explained
variance over and above the other. This study highlights the importance of
assessing transfer climate during the needs-analysis process and suggests the
potential benefits of manipulating the work environment to support subse­
quent transfer.
Another factor that could influence transfer is the extent to which the
posttraining environment provides opportunities for trainees to apply what
they have learned. Ford et al ( 199 1 ) studied Air Force technical trainees after
they completed training and found significant differences in opportunity to
apply the training and wide variations in the length of time before trainees first
performed the tasks for which they had been trained. Supervisor and peer
 422 TANNENBAUM & YUKL

support were related to the extent to which airmen had opportunities to

perform trained tasks. In a study of IRS managers , Pentland ( 1989) found that
attempts to practice trained computer skills immediately upon returning to the
job had a major impact on long-term retention. Clearly, opportunity to apply
and practice trained skills is an important construct worthy of further atten­
tion.
Situational constraints in the posttraining work environment can impede
transfer of training (Noe 1 986) . Goodman & Miller ( 1990) described several
facilitators and inhibitors of training in a high-tech manufacturing environ­
ment. Schoorman & Schneider's book ( 1 988) contains a great deal of useful
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

information about the nature and influence of environmental facilitators and

situational constraints-as does the work of Peters and his associates (e.g.
by University of Connecticut on 02/24/13. For personal use only.

Peters et al 1 988).

Posttraining Activities
Baldwin & Ford ( 1 988) noted that the counseling and psychotherapy litera­
tures are potentially rich sources of ideas for creating conducive transfer
environments . They mentioned the applicability of buddy systems, booster
sessions, and relapse-prevention programs to an organizational context . A
few recent empirical studies utilized aspects from related fields.
Fleming & Sulzer-Azaroff ( 1 990) conducted a study of paraprofessionals at
a facility for the disabled . After training, pecrs were paircd and instructed to
provide feedback and reinforcement to each other. Fleming & Sulzer-Azaroff
reported that this peer management approach increased the maintenance of
training behaviors .
Several studies have examined whether transfer is facilitated by relapse­
prevention training, an approach derived from research on physical addictions
(Marx 1 982). Relapse-prevention training is designed to prepare trainees for
the posttraining environment and to anticipate and cope with "high risk"
situations. Marx & Karren ( 1 988) found that trainees who received relapse­
prevention training after a regular training seminar demonstrated more of the
trained behaviors than trainees without relapse training. Gist et al ( 1 990)
found that relapse prevention training that incorporated self-management and
goal-setting techniques yielded higher rates of skill generalization and better
overall performance than training that used only goal-setting. Although this
approach holds promise, the research on relapse prevention training in organi­
zational settings has been mixed (cf Wexley & Baldwin 1 986; Marx & Karren
1 990) .
As we noted earlier, trainees who expected some form of posttraining
follow-up left training with stronger intentions to transfer (Baldwin & Mag­
juka 1 99 1 ) . Nevertheless , most companies report conducting no specific
follow-up of participants after program attendance (Saari et aI 1 988). Marx &
Karren ( 1990) found that trainees were more likely to apply time man agement
 TRAINING & DEVELOPMENT 423

skills when follow-up occurred three weeks after a time management course.
In this study, follow-up consisted of either a re-examination of high-risk
situations (for those who had also undergone relapse-prevention training) or a
discussion of long- and short-term career goals. Neither follow-up was speci­
fically related to the time management training per se, yet both improved
transfer. Thus, it appears that posttraining follow-up can be helpful. How­
ever, this study , like the one by Wexley & Baldwin ( 1 986), was conducted
with college students, and follow-up was conducted by a trainer and not a
supervisor. Future research should examine the impact of engaging the
trainee' s supervisor in follow-up activities . For example , a supervisor could
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

examine a summary developed by the trainee, discuss the relevance of the

training, or review posttraining action plans with the trainee in order to signal
by University of Connecticut on 02/24/13. For personal use only.

that transfer of training is important and that the trainee is accountable for it.
In summary , although research has not been plentiful, it appears that
limitations in the posttraining environment may inhibit the application of
skills acquired during training. Furthermore, supervisory actions taken after
one training course may become pretraining cues for subsequent training
courses. As such they can influence employee attitudes upon entering the next
program. To optimize training effectiveness we must consider some of the
diagnostic concerns expressed in the organizational development and socio­
technical literatures . We should examine the transfer environment carefully,
identify situational facilitators and inhibitors, and propose means either to
prepare trainees to deal with the inhibitors or to modify the posttraining
environment to encourage transfer.

TRAINING EVALUATION

As was the case with training-needs analysis, researchers generally agree that
training evaluation is an important part of the training system (Goldstein &
Gilliam 1 990). Unfortunately, although a great deal of information exists
regarding various evaluation designs (e.g. Cook et al 1 99 1 ; Goldstein 1 99 1 )
application still lags, and employee reactions are too frequently relied upon to
determine training effectiveness (Brinkerhoff 1 989; Saari et al 1 988). The
American Society of Training and Development ( 1 990) examined the evalua­
tion practices of several large organizations in detail. Almost all of the
companies reported that they evaluated trainee reactions, but only 10%
reported evaluating behavior change on the job. In this section we present
recent developments regarding training evaluation.

Evaluation Design and Analysis

Arvey & Cole ( 1 989) examined the power of posttest only, gain score , and
analysis of covariance (ANCOVA) options for analyzing classic experimental
designs for training evaluation, and they concluded that ANCOVA is at least
 424 TANNENBAUM & YUKL

as powerful as the other options under most conditions . They also reiterated
some concerns regarding the use of nonequivalent control groups. They
suggested use of an alternative-ranks procedure when sample sizes are small,
and use of latent-variable models when pretest differences are exhibited.
Maxwell et al ( 199 1 ) compared two methods for increasing power in random­
ized between-subject designs . They recommended using analysis of variance
(ANOV A) with a lengthened posttest in situations where pretest sensitization
or an inability to administer a posttest prohibit the traditional ANCOVA
approach. Additional studies that compare the tradeoffs in using various
analyses would be beneficial . Although our knowledge about randomized
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

evaluation designs continues to improve, organizational constraints often

limit the applicability of such designs (Brinkerhoff 1 989) . Research on the
by University of Connecticut on 02/24/13. For personal use only.

usefulness of alternative, quasi-experimental designs is sorely needed (Gold­

stein & Gessner 1 988). Latham ( 1 989) advocated the use of multiple-baseline
designs to assess training effectiveness. Woods & Tannenbaum ( 1 990) pre­
sented an application of a "multiple cohort design" in a large financial
organization, comparing learning and performance changes both within and
across groups of trainecs . The measurement of change across nonequivalent
groups remains a complex problem CArvey & Cole 1 989) . However, because
it is unlikely that organizational realities will allow for extensive use of
experimental designs, we must continue to explore alternatives more con­
sistent with typical constraints.
The collection of pre- and posttraining self-report data is common in
evaluating training effectiveness. Concerns about response-shift bias have
prompted researchers to consider the use of retrospective pretest measures
(sometimes called a "then score") (Howard & Dailey 1 979) . Most studies
have shown that change scores based on retrospective pretest measures are
more in agreement with objective indications of change than change scores
based on more traditional pre-test measures (Sprangers & Hoogstraten 1 989),
but retrospective recall measures may entail potential biases (Sprangers &
Hoogstraten 1 988).
The few evaluation studies that examined transfer tended to focus on initial
generalization to the job, not long-term retention or behavior maintenance.
However, initial retention is not necessarily equivalent to longer-term reten­
tion (Fendrich et al 1 988). Longitudinal follow-up, such as that done by
Latham & Frayne ( 1 989) , provides additional information about the mainte­
nance of trained skills. Pellum & Teachout ( 1 990) presented a longitudinal
evaluation of training Air Force jet engine mechanics. They collected work
samples and learning measures after introductory training, after advanced
training, and again after on-the-job training.
Despite the importance of assessing maintenance of skills there are no
guidelines for determining the appropriate length of time to wait before
 TRAINING & DEVELOPMENT 425

collecting posttraining measures . Research that collects outcome measures

over several points in time would allow us to understand skill decay better and
could help clarify when follow-up measures should be collected and refresher
training offered.

Criterion Issues
Kirkpatrick's ( 1 976) typology remains the prevalent framework for categoriz­
ing training criteria. His typology includes four levels of training effective­
ness: reactions , learning, behavior and organizational results. As trainee
reactions remain the most commonly used training criterion, it is important to
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

understand the extent to which positive reactions are indicative of the other
three types of criteria. In a meta-analysis of previous training studies, Alliger
by University of Connecticut on 02/24/13. For personal use only.

& Janak ( 1 989) examined the correlations among the four levels of training
effectiveness. They found virtually no relationship among trainee reactions
and the other levels , but slightly higher correlations among the other levels.
However, Alliger & Janak noted that their findings were based on a small
number of studies. A few recent studies add related information. Mathieu et al
( 1990) reported a significant relationship between learning and subsequent
performance (a proofreading work sample). They also found that training
motivation was positively related to learning among those individuals who
reacted positively to the training; that is, trainee reactions moderated the
relationship between motivation and learning. In a study of interview-skills
training, Campion & Campion ( 1987) found positive indications of trainee
reactions and learning, but no differences were observed between the trained
group and nonparticipants with regard to their interview behaviors or out­
comes.
In sum, these studies fail to support the direct causal relationship among
levels often assumed in Kirkpatrick's typology. In particular, trainee learning
appears to be a necessary but not sufficient prerequisite for behavior change.
As discussed earlier, the posttraining environment can play an important role
in determining whether transfer occurs . Furthermore, reactions do not appear
to be directly related to the other criteria. In other words , liking does not
imply learning, a finding that highlights the problem of using trainee reactions
as the sole criterion of training effectiveness . Future research could examine
whether trainee reactions that focus on the utility or applicability of training
(as opposed to assessing whether trainees liked the training) are related to any
of the other effectiveness criteria. At this point, however, we conclude that
reaction measures are not a suitable surrogate for other indexes of training
effectiveness.
The selection and development of appropriate criterion measures remains a
critical component of training evaluation. Goldstein ( 1 99 1 ) has reminded us
that the issues of criterion relevance, deficiency, and contamination are as
 426 TANNENBAUM & YUKL

applicable to measuring training effectiveness as they are to any other per­

formance evaluation context. Furthermore, the most convenient or easily
accessible measures of performance may prove to be inappropriate for assess­
ing training effectiveness. Campbell ( 1 988) noted problems with the use of
global measures of performance. S ince training is but one of many factors that
contribute to overall performance, global measures of performance can easily
suffer from criterion contamination. Campbell advocated the use of targeted
measures based on specific training objectives .
Sackett et al ( 1 988) found low correlations between measures of typical and
maximum performance across two samples, suggesting that the two indexes
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

measure distinct phenomena. Maximum performance is characterized by: (a)

an explicit awareness of being evaluated, (b) an acceptance of explicit in­
by University of Connecticut on 02/24/13. For personal use only.

structions to maximize effort, and (c) a short enough measurement period to

allow focused attention on the goal. Although Sackett et ai' s study was not
conducted in a training context it does have implications for evaluating
training effectiveness . For example, in a training context, a role-play con­
ducted at the end of training approximates a maximum performance measure
and indicates how well a trainee "can" exhibit the trained behavior, but it does
not indicate that the trainee "will" typically perform at that level.
Although skill development is often the primary goal of training, it is not
necessarily the sole purpose (Marx & Hamilton 1 99 1 ; Brinkerhoff 1 989).
Marx & Hamilton ( 1 99 1 ) noted that training can also contribute to team
building, coalition building, and culture building outcomes. Under certain
circumstances , variables such as self-efficacy (Latham 1 989) and organiza­
tional commitment (Tannenbaum et al 199 1 b) may be valuable outcomes in
their own right. For example , socialization training may be designed both to
transmit knowledge and to affect attitudes (Feldman 1 989) . Thus, it is impor­
tant to establish clear training objectives and develop outcome measures that
assess these objectives.
According to an ASTD study, approximately two thirds of training manag­
ers felt they were coming under additional pressure to show that their pro­
grams produce "bottom-line" results (Carnevale & Schulz 1 990). Methods for
estimating "bottom-line" impact continue to be refined. Mathieu & Leonard
( 1987) applied utility concepts to examine the effectiveness of training bank
supervisors . They included estimates of decay and employee flow as well as
economic parameters and break-even analyses to project the cost effectiveness
of the training over a 20-year period. Cascio ( 1 989) presented guidelines for
assessing the economic impact of training, incorporating net present value
adjustments , and explaining breakeven analysis. Finally, Deitchman ( 1 990)
attempted to quantify the military value of training expenditures in compari­
son to equipment improvements . He found that training and equipment
improved force effectiveness by comparable amounts , but training cost the
 TRAINING & DEVELOPMENT 427

same or less than most equipment improvements. Deitchman explained that

his findings only pertained to specific cases but demonstrated the feasibility of
the approach for military planning purposes.

TRAINING FOR SPECIFIC POPULATIONS

In the next two sections, we discuss the application of training to two unique
and especially important populations of trainees: managers and teams.

Management Development
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Training and development of managers continues to be a multi-billion dollar

business. Two recent articles report the results of large surveys on the amount
by University of Connecticut on 02/24/13. For personal use only.

and types of management training in companies (Gordon 1 988; Saari et al

1 988). Saari et al surveyed a representative sample of 6 1 1 companies of
different sizes and types to identify the type of management development
activities they use, the reasons for using them, procedures for selecting
managers to participate in training activities, demographic characteristics of
the participants , preparation of managers prior to program attendance , and
procedures for evaluating training effectiveness. The percentage of companies
that reported using various types of training was 93% for on-the-job training ,
90% for external short courses, 80% for special projects or task forces , 57%
for mentoring, 40% for job rotation, 3 1 % for university residential programs ,
and 25% for executive MBA programs. Significant differences in training
practices were found among various industries and between large and small
companies.
An extensive literature on management development within the discipline
of management education includes considerable theory and research on the
skills relevant for managers and the effectiveness of different approaches in
teaching these skills (see review by Keys & Wolfe 1 988). For example, a
recent special issue of the Journal of Management Development contained
articles on a variety of training methods , including cases (Osigweh 1 989) ,
lecture and discussion (Griffin & Cashin 1 989) , and simulations (Faria 1 989) .
A current concern in management education is to identify the skills and
competencies needed by managers and evaluate the capacity of MBA pro­
grams to teach these skills (Porter & McKibbin 1 989) .

NEEDS ASSESSMENT FOR MANAGERS Assessment of training needs for

managers is more difficult than for most jobs owing to the complexity of
managerial work. Identification of competencies likely to be relevant for a
particular type of managerial position is facilitated by continuing progress in
the research on managerial skills, activities, and behavior. This research is
reviewed in leadership books by Bass ( 1 990) and YukI ( 1989) , and additional
 428 TANNENBAUM & YUKL

findings are reported in recent articles (e. g . Jonas et al 1990; Kraut et al

1 989) .
Needs assessment should be based on accurate measures of managerial
behavior and skills. A recent book edited by Clark & Clark ( 1 990) includes
reports on the development and validation of several measures useful for
assessing training needs of managers . For example, Compass is a set of three
types of questionnaires used to provide managers with information about their
managerial behavior and the developmental needs related to this behavior
(YukI et al 1 990; YukI & Lepsinger 199 1 ) . Subordinates and peers describe
how much a manager uses each of fourteen managerial practices, and they
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

recommend whether the manager should do more, the same amount, or less of
each behavior. Managers compare this feedback to their own self-assessment
by University of Connecticut on 02/24/13. For personal use only.

of behavior and to norms for similar managers . Ratings of the importance of

each type of behavior for the manager's job provide additional information for
selecting relevant training.
Another example of a needs-assessment instrument is Benchmarks
(McCauley et al 1 989) , a rating form developed to measure relevant skills and
traits identified in research on managerial experiences and managerial derail­
ment. SeU-ratings and ratings made by subordinates, peers , and superiors are
used to assess a manager's strengths and weaknesses and identify develop­
mental needs.
Streufert et al ( 1 988) described research on assessment of managerial skills
with two computer-based simulations (Woodline County Disaster Control
Coordinator, and Governor of Shamba) . The simulations are designed for a
single individual (like an in-basket) rather than a group in order to regulate the
information available to each participant. Managerial behavior and skills in
handling crises can be assessed under standardized conditions . In other
business games, by contrast, each player faces a unique situation determined
by the joint actions of all the players and the parameters of the industry model
used in the simulation software.

LEADERSHIP TRAINING PROGRAMS An extensive review of leadership

training programs was done by Latham ( 1 988) in the last Annual Review
chapter on training, and that review is updated in the new edition of the
training book by Wexley & Latham ( 1 99 1 ) . Leadership training is also
reviewed by Bass ( 1 990) and by Tetrault et al ( 1 988). Most of the training
studies described in these reviews examined whether managerial effectiveness
is improved by training based on a particular leadership theory. The typical
design of these studies is to show improvement in performance ratings after
training, or to compare ratings for managers who receive the training to
ratings for untrained managers. It is difficult to interpret positive results in this
type of study, since a number of rival hypotheses are plausible (e.g. criterion
 TRAINING & DEVELOPMENT 429

contamination, increased self-confidence), and these rival hypotheses may

explain improved performance even though trainees did not learn the leader­
ship theory showcased in the training, or did not behave back on the job in
ways prescribed by the theory. Future research on the effects of leadership
training should be designed to evaluate rival hypotheses and to determine how
training affects managerial behavior as well as job performance.

DEVELOPMENTAL EXPERIENCES AND MENTORING In recent years, re­

searchers have paid increasing attention to management development that
occurs on the job. Much of this development is informal and unplanned. Most
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

organizations do not use developmental assignments and other forms of

on-the-job training for managers in a systematic way. Even when there is a
by University of Connecticut on 02/24/13. For personal use only.

program of job rotation or developmental assignments, it is usually not linked

to specific learning objectives or assessment of individual training needs
(Ruderman et al 1990).
Efforts to improve management development have focused on identifying
the types of experiences that facilitate development and the processes by
which managers are able to learn from these experiences. Research conducted
at the Center For Creative Leadership (McCall et al 1988) found that growth
and learning were greatest when challenging situations or adversity forced a
manager to come to terms with his or her limitations and overcome them.
C hallenging situations likely to encourage development included conflict with
the boss, problem subordinates, a forced merger or reorganization, problems
inherited from a predecessor, intense pressure from a difficult assignment
with high visibility, a job for which the manager lacks experience and
credibility , unfavorable business conditions or unusual external threats, and
turning around a wcak unit. A study by Kelleher et al (1986) examined
personal and organizational variables that jointly affect learning from experi­
ence by managers. A study by Dechant (1990) investigated the process by
which managers diagnose their learning needs and develop strategies to meet
them.
Ruderman et al (1990) described a new questionnaire called the Job Chal­
lenge Profile (JCP) designed to measure the potential of a job for management
development. In contrast to traditional job analysis, which focuses on duties
and responsibilities, the JCP focuses on demands and challenges in a job. The
JCP could be useful for determining which job rotation assignments, special
projects, or job changes have the greatest developmental potential for a
manager.
Kaplan et al (1987) found that self-development becomes more difficult at
high-level executive positions in the organization. Several obstacles inhibit
learning from experience by executives: the hectic pace and unrelenting
demands of the job make introspection difficult, success fosters resistance to
 430 TANNENBAUM & YUKL

changing behavior or recognizing personal weaknesses, executives tend to

become isolated, and executives are insulated from criticism from people
reluctant to risk offending them. Thus , even if an executive wants to discover
his or her weaknesses, it is difficult to do so. Kaplan et al ( 1 987) suggested
ways to develop sourCeS of accurate feedback and identify developmental
needs .
Interest continues in mentoring as a method for facilitating management
development (London & Mone 1 987; Noe 1 99 1 ) . Many organizations rely on
informal mentoring, although without any formal structure some managers
who need mentors are unlikely to establish a mentoring relationship. In most
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

formal programs individuals are assigned a mentor. However, problems such

as personality conflicts and lack of mentor commitment are more likely to
by University of Connecticut on 02/24/13. For personal use only.

occur with assigned mentors than with informal mentors unless the assigned
mentors are carefully selected and trained. Zey ( 1 988) recommended greater
use of mentors to facilitate adjustment during difficult job transitions, such as
with newly hired employees , employees transferred or promoted to a different
unit in the organization, employees in newly merged companies , and em­
ployees with a job assignment in a foreign country . Noe ( 1 988a) developed a
questionnaire to measure mentor behavior and found that mentors provide two
functions similar to those found in Kram's ( 1 985) research based on in­
terviews: a psychosocial function (acceptance, encouragement, coaching,
counseling) and a career facilitation function (sponsorship, protection,
challenging assignments , exposure and visibility). People with assigned men­
tors reported receiving psychosocial benefits but only limited career benefits.
Noe ( 1 988b) and Ragins ( 1 989) examined the difficulties encountered by
female managers in mentoring relationships. Willbur ( 1987) found that career
advancement and success were predicted by several aspects of mentoring,
including number of mentors and the functions provided by mentors. Un­
fortunately, there has been little empirical research on the ways mentors
facilitate managerial development , as opposed to their influence on job
satisfaction and career advancement.
Action learning is an approach widely used in Europe for combining formal
management training with learning from experience (Revans 1 982). A typical
program is conducted over a period of from 6 to 9 months and includes field
project work interspersed with skill training seminars . Teams of managers
with diverse backgrounds conduct field projects on complex organizational
problems requiring use of skills learned in the formal training sessions. The
teams meet periodically with a skilled facilitator to discuss, analyze, and learn
from their experiences (Marsick 1 990) .

Team Training
As organizations rely more on the use of teams to attain organizational goals
the need to enhance team effectiveness has increased (Hackman 1 989) . Two
 TR AINING & DEVELOPMENT 43 1

methods designed to enhance team effectiveness are team training and team
building. Although team building and team training interventions often focus
on similar concerns (e.g. enhancing communication , decision making,
coordination) the means of approaching the concerns differ (Tannenbaum et al
1 99 1 a) . In team training, the specific knowledge , skills, and attitudes to be
developed are determined prior to the start of training, and learning objectives
are established. In contrast, team building is more of a process intervention,
aimed at helping individuals and groups examine and act upon their behavior
and relationships. Recent reviews concluded that team building appears to
have a positive effect on the perceptions and attitudes of team members
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

although results for behavioral outcomes were more equivocal (Sundstrom et

al 1 990; Tannenbaum et al 199 1a,b) . A meta-analysis of team performance
research found few empirical studies of team training through the mid- 1 980s
by University of Connecticut on 02/24/13. For personal use only.

(Salas et al 199 1b). Interest in team training has recently increased, and this
section focuses on developments in the area.
Several writers have noted that an obstacle to developing an effective team
training program is a lack of methods for analyzing team tasks, behaviors , and
skills (Modrick 1 986; Morgan & Salas 1988). Earlier research on task analy­
sis has typically focused on the individual level of analysis, aggregated to the
job level (Ostroff & Ford 1989) . Recently, Levine & Baker ( 1 99 1) presented
a methodology to analyze team tasks, and Bowers et al ( 1 991 ) presented
preliminary work on assessing the coordination requirements of team tasks.
Glickman et al ( 1 987) studied military command and control teams in training
and reported that two separate tracks of behavior evolve during team training:
taskwork, involving the development of skills related to execution of the task;
and teamwork, focusing on the behaviors required to function effectively as a
team member. Teamwork behaviors that differentiate effective teams include
effective communication, coordination, compensatory behavior, mutual per­
formance monitoring, exchange of feedback, and adaption to varying
situational demands (Oser et al 1 989) . Other developments include Fleishman
& Zaccaro's ( 199 1 ) team-oriented taxonomy of performance, and Franz et
aI' s ( 1 990) identification of seven categories of team skills to be included in
aircrew coordination training.
Although preliminary progress has been made in the analysis of team tasks
and skills, little research has been done on the design of team training. Salas
et al ( 1 99 1 a) raised a number of questions about the applicability of traditional
learning principles to the team training context. Swezey & Salas ( 199 1 a,b)
examined the related research and presented a series of learning and in­
structional guidelines for use in developing team training. In one of the few
empirical studies comparing team training methods, Lassiter et al ( 1 990)
found that teams trained in a skill-oriented training program demonstrated
better communication skills than those in a lecture-based program or a control
group. A number of commercial airlines have established team training
 432 TANNENBAUM & YUKL

programs for their aircrews, but the preliminary results on training effective­
ness have been mixed (cf Cannon-Bowers et al 1 989; Helrnreich & Wilhelm
1 989; Helrnreich et al 1 990).
The examination of mental models has particular relevance in the team
training context. Cannon-Bowers et al ( 1 990) suggested that shared or over­
lapping mental models among team members should enhance the use of
implicit coordination strategies by enabling team members to anticipate be­
havior and information needs more accurately. Thus, training efforts to
enhance the development of common and accurate mental models (e . g .
crosstraining) should improve team effectiveness. Additional research i s also
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

needed on team task analysis, the nature of team performance, and the
applicability of traditional learning principles to team training. Finally, re­
by University of Connecticut on 02/24/13. For personal use only.

search is needed to determine the optimal configuration of individual and

team training to enhance team effectiveness .

CONCLUDING COMMENTS

In closing, we highlight a few general trends in the field and offer a few
general suggestions about future research. The practice of training continues
to grow , with employers spending approximately $30 billion on formal
training and approximately $ 1 80 billion on informal on-the-job training each
year (Carnevale et al 1 990b) . Along with the growth in practice, the quantity
and quality of research related to training have increased during the last
several years . There appears to be increased cross-pollination with other
disciplines, most notably cognitive, instructional, and social psychology.
Training researchers are drawing upon related conceptual work, including
social learning theory and self-efficacy (e.g. Gist et al 1 989) , cognitive or
attentional resources (Kanfer & Ackerman 1989) , realistic job previews
(Hicks & Klimoski 1 987), relapse prevention (Marx & Karren 1988) ,
situational cues and reinforcers (e.g. Rouillier & Goldstein 1 99 1 ) , and pro­
cedural justice (e.g. Baldwin et al 1 99 1 ) . The signs of improving integration
and conceptual development are encouraging , but a great deal more is possi­
ble and desirable.
For many years writers have discussed training as a system embedded
within an organizational context (e.g. Hinrichs 1976) . During the last few
years , researchers have begun to pay more attention to the pre- and posttrain­
ing environments as important determinants of training effectiveness. Re­
searchers are starting to consider trainees as active participants in the system
who interact with the environment before training, during training, and after
training. This trend is encouraging, and more research with a systems per­
spective is clearly desirable.
A recurring theme in several recent reviews and overviews of training is the
 TRAINING & DEVELOPMENT 433

need for a paradigm shift from research designed to show that a particular type
of training "works," to research designed to determine why, when, and for
whom a particular type of training is effective. Furthermore , we must do a
better job of describing the nature and purpose of the training being studied.
Researchers have tended to consider all training the same, without regard to
the purpose of the training or the type of learning involved. We need to be
clear not only about the training method(s) employed but also about the basic
content and purpose of the training. Making these distinctions should greatly
enhance the ability to compare studies in future reviews.
One of the most important changes in the field is the increased attention
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

being given to cognitive concepts . As the kinds of tasks performed by humans

in organizations become more complex owing to technological change , in­
by University of Connecticut on 02/24/13. For personal use only.

ternationalization , and other significant developments (Goldstein & Gilliam

1 990) , cognitive models of learning are becoming increasingly important.
Cognitive developments have already made some contributions to training­
needs analysis and training design. Cognitive approaches hold great promise
for the field of training, but we should be careful they do not become the latest
research fad. Cognitive models only supplement existing behavioral training
models , they do not replace them . Although significant work remains in
applying and evaluating cognitive concepts in organizational settings, the
prospects are exciting .
Improvements in high-technology training methods such as intelligent
computer-aided instruction, interactive-videodisc instruction, and equipment
simulators , continue at a rapid pace. It is now possible to link several different
training methods into an integrative, technology-based, training program with
the capacity to provide the benefits of self-pacing, active involvement and
expert tutoring for each trainee . The new developments promise to
revolutionize the practice of training and education in the next century .
Unfortunately, empirical research to determine how different features of the
high-technology methods facilitate training has lagged far behind develop­
ment of the technology itself.
The distinction between on-the-job training and off-site training is becom­
ing blurred. New technology allows training to occur "on-line" as expert
systems provide diagnostic clues and feedback to workers operating sophisti­
cated equipment, monitoring operations , and troubleshooting problems. As
organizations decentralize to become "closer to their customers ," training
responsibilities often become decentralized as well. Approaches that integrate
on-the-job practice with formal training are becoming more popular. Research
on the developmental experiences of managers is providing information with
which to improve on-the-job learning of relevant skills and allow better
integration of developmental experiences and formal training.
In general we are optimistic about the field of training. In the past, the field
 434 TANNENBAUM & YUKL

has been labeled faddish and atheoretical and it is still prone to those tenden­
cies. Nevertheless, progress during the past decade has been encouraging.
Several areas of current research hold particular promise for improving our
understanding of training effectiveness, and the field appears poised to move
forward rapidly.

ACKNOWLEDGMENTS

Lisa Dupuree-Bruno and Bruce Tracey, doctoral students at the University of

Albany's Organizational Studies Program, provided valuable assistance in
conducting the literature search for this chapter. We also wish to thank
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

George Alliger, Irv Goldstein, John Mathieu, Eduardo Salas, and Tom Taber
for their helpful comments and suggestions on an earlier draft of this chapter.
by University of Connecticut on 02/24/13. For personal use only.

Literature Cited

Ackerman. P. L. 1988. Determinants of in­ and its Implications. New York: Freeman.
dividual differences during skill acquisition: 2nd ed
cognitive abilities and information process­ Anderson, J. R. 1987. Skill acquistion: com­
ing. J. Exp . Psychol.: Gen. 117:288-318 pilation of work-method problem solutions.
Ackerman, P. L. 1989. Within-task in­ Psychol. Rev. 94: 1 92-2 1 0
tercorrelations of skilled performance: im­ Arvey, R. D . , Cole, D . A. 1989. Evaluating
plications for predicting individual differ­ change due to training. See Goldstein 1989,
ences? (A comment on Henry and Hulin, pp. 89-118
1987). J. Appl. Psychol. 74:360-64 Baldwin, T. T. 1991. Effects of alternative
Ackerman, P. L . , Kyllonen, P. C. 1991. modeling strategies on outcomes of in­
Trainee characteristics. In Training for Per­ terpersonal skills training. J. Appl. Psychol.
formance: Principles of Applied Human In press
Learning, ed. J. E. Morrision, 193-230. Baldwin, T. T., Ford, J. K. 1988. Transfer of
West Sussex, England: John Wiley & Sons training: a review and directions for future
Ackerman, P. L., Sternberg , R. J . , Glaser, research. Pers. Psychol. 41:63-105
R . , ed. 1989. Learning and Individual Dif­ Baldwin, T. T., Magjuka , R. 1. 1991. Organi­
ferences: Advances in Theory and Re­ zational training and signals of importance:
search. New York: W. H. Freeman effects of pre-training perceptions on inten­
Adams, J. A. 1987. Historical review and tions to transfer. Hum. Res. Dev. 2(1):25-36
appraisal of research on the learning, reten­ Baldwin, T. T. , Magjuka, R. J . , Loher, B . T.
tion , and transfer of human motor skills. 1991. The perils of participation: effects of
Psychol. Bull. 101:41-74 choice on training on trainee motivation and
Ajzen, I. 1985. From intentions to actions: a learning. Pers. Psychol. 44: 51-66
theory of planned behavior. See Kuh1 & Bandura, A. 1986. Social Foundations of
Beckman 1985, pp. 11-39 Thought and Action. Englewood Cliffs , NJ:
Alderfer, C. P . , Tucker, R. C., Alderfer, C. Prentice-Hall
J., Tucker, L. M. 1991. The race relations Barrett, G. V., Caldwell, M. S., Alexander,
competence workshop: an intergroup educa­ R. A. 1989. The predictive stability of abil­
tional procedure. Hum. Relal. In press ity requirements for task performance: a
Alliger, G. M . , Janak, E. A. 1989. Kirkpat­ critical reanalysis. Hum. Perform. 2:167-81
rick's levels of training criteria: thirty years Barrick, M. R . , Mount, M. K. 1991. The big
later. Pers. Psychol. 42:331-42 five personality dimensions and job per­
Alliger, G. M., Serbell, C. V., Vadas, J. E. formance: a meta-analysis. Pers. Psychol.
1989. Computer-based simulation for soft 44:1-26
skills training. J. Interact. Instr. Dev. 4:8- Basadur, M . S., Wakabayashi, M., Graen, G .
15 B. 1990. Individual problem solving styles
Alluisi, E. A. 1991. The development of tech­ and attitudes toward divergent thinking be­
nology for collective training: SIMNET, a fore and after training. Creal. Res. J. 3:22-
case history. Hum. Factor. In press 32
Anderson, J. R. 1985. Cognitive Psychology Bass, B . M . 1990. Handbook of Leadership:
 TRAINING & DEVELOPMENT 435

Theory, Research and Managerial Im­ Cascio, W . F. 1 989. Using utility analysis to
plications. New York: Free Press. 3rd ed assess training outcomes. See Goldstein
Black, J. S. 1988. Work role transitions: a 1 989, pp. 63-88
study of American expatriate managers in Casner- Lotto , J. and associates. 1989. Suc­
Japan. J. Int. Bus. Stud. 19:277-94 cessful Training Strategies. San Francisco:
Black, J. S . , Mendenhall , M . 1 990. Cross­ Jossey-B ass
cultural training effectiveness: a review and Cassidy-Schmitt, M. C . , Newby, T. J. 1 986.
a theoretical framework for future research. Metacognition: relevance to instructional
Acad. Manage . Rev. 1 5: 1 1 3-36 design. J. Instr. Dev. 9:20-33
Bouffard-Bouchard, T. 1 990. Influence of Clark, R. C. 1 988. Metacognition and human
self-efficacy on performance in a cognitive performance improvement. Peiform. Im­
task. 1. Soc. Psychol. 1 30: 353-63 prove. Q. 1 :33-45
Bowers, C. A . , Morgan, B . B . , Salas, E . Clark, K. E . , Clark, M. B . 1 990. Measures of
1 99 1 . The assessment o f aircrew coordina­ Leadership. West Orange, NJ: Leadership
tion demand for helicopter flight require­ Library of America
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

ments. Paper presented at the Int. Symp. Cohen, D. J. 1 990. What motivates trainees .
Aviation Psycho!. Train. Dev. 1. Nov . :9 1-93
Brinkerhoff, R. O. 1 989. Evaluating Training Cook, T . D . , Campbell, D . , Peracchio, L.
by University of Connecticut on 02/24/13. For personal use only.

Programs in Business and Industry. San 1 99 1 . Quasi-experiments. See Dunnette &

Francisco: Jossey-Bass Hough 1 99 1 , pp. 49 1-576
Burke, M. J . , Day, R. R. 1 986. A cumulative Cooke, N . M . , Schvaneveldt, R. W . , 1 988.
study of the effectiveness of managerial Effects of computer programming experi­
training. J. Appl. Psychul. 7 1 :232-45 ence on network representations of abstract
Butterfield, E. C . , Nelson, G. D. 1 989. programming concepts. Int. J. Man-Mach.
Theory and practice of teaching for transfer. Stud. 29:533-50
Educ. Technol. Res. Dev. 37:5-38 Cranny, C. J . , Doherty , M. E. 1 988. Im­
Campbell, J . P. 1 97 1 . Personnel training and portance ratings in job analysis: notes on the
development. Annu. Rev. Psychol. 22:565- misinterpretation of factor an alysi s . J. Appl.
602 Psychol. 73:320-22
Campbell, J. P. 1 988. Training design for per­ Cronbach, L. J . , Snow, R. E. 1977. Aptitudes
formance improvement. In Productivity in and Instructional Methods: A Handbookfor
Organizations, ed. J . P. Campbell, R . J . Research on Interactions. New York:
Campbell, and Associates, p p . 1 77-2 1 6 . Irvington
San Francisco: Jossey-Bass DeBloois, M. L. 1 988. Use and Effectiveness
Campion , M. A . , Campion, J. E. 1 987 . of Videodisc Training. Falls Church, VA:
Evaluation of an interview skills training Future Systems Inc.
program in a natural field setting. Pers. Psy­ Dechant, K . 1 990. Knowing how to learn: the
chol. 40:675-9 1 "neglected" management ability . 1. Man­
Cannon-Bowers, J . , Prince, C . , Salas, E . , age. Dev. 9:40-49
Owens, J . , Morgan, B . B . , Gonos, G . H . Deitchman , S. J. 1 990. Further explorations in
1 989. Determining aircrew coordination estimating the mil itary value of training.
training effectiveness. Presented at the 1 1 th IDA Paper P-23 ! 7 . Alexandria, VA: Inst.
Interserv.lIndust. Conf. , Fort Worth , TX Defense Anal.
Cannon-Bowers, J. A. , Salas, E . , Converse, Derry, S. J . , Murphy, D. A . 1986. Designing
S. A. 1 990. Cognitive psychology and team systems that train learning ability: from
training: training shared mental models of theory to practice. Rev. Educ . Res. 56: 1-39
complex systems. Hum. Factor. Soc. Bull. Drucker, E. H . , Campshure, D . A . 1 990. An
33( 1 2): 1 -4 analysis of tank platoon operations and their
Caplan , L. J . , Schooler, C. 1 990. Problem­ simulation on simulation networking (SIM­
solving by reference to rules or previous NET). ARI Res. Prod. 90-22. Alexandria,
episodes: the effects of organized training, VA: US Army Res. Inst. Behav. Soc. Sci.
analogical models, and subsequent com­ Dunnette, M. D . , Hough, L. M . , eds. 1 99 1 .
plexity of experience. Mem. Cognit. 1 8 : Handbook of Industrial and Organizational
2 1 5-27 Psychology, Vol. 1 . Palo Alto, CA: Con­
Carnvevale, A. P . , Gainer, L. J . , Meltzer, A . sulting Psychologists Press. 2nd ed.
S . 1990a. Workplace Basics. San Francisco: Dweck, C . S. 1 986. Motivational processes
Jossey-Bass affecting learning. Am. Psychol. 4 1 : 1 040-
Carnevale, A. P . , Gainer, L . J . , Villet, J . 48
1 990b. Training i n America: The Organiza­ Eberts, R. E . , Brock, J . F. 1987. Computer­
tion and Strategic Role of Training. San assisted and computer-managed instruction.
Francisco: Jossey-Bass In Handbook of Human Factors, ed. G .
Carnevale, A. P . , Schulz, E. R. 1 990. Evalua­ Salvendy, p p . 976- 1 0 1 1 . New York:
tion practices. Train. Dev. J. 44:S-23-S-29 Wiley-Interscience
 436 TANNENBAUM & YUKL

Eden, D . 1990. Pygmalion in Management. Franz, T. M . , Prince, C . , Cannon-Bowers, J .

Lexington, MA: Lexington Books A . , Salas, E. 1 990. The identification of
Elliott, E. S . , Dweck, C. S. 1988. Goals: an aircrew coordination skills. Proc. Dept. De­
approach to motivation and achievement. J. fense Symp . 1 2 :92-96
Pers. Soc. Psychol. 54:5- 1 2 Frayne, C. A . , Latham, G. P. 1987. The
Faria, A . J . 1 989. Business gaming: current application of social learning theory to em­
usage levels. 1. Manage. Dev. 8:59-65 ployee self management of attendal)ce. J.
Feldman. D. 1 988. Managing Careers in Appl. Psychol. 72:387-92
Organizations. Glenview, IL: Scott Fores­ Gagne, R. M . , Glaser, R. 1987. Foundations
man in learning research . In Instructional Tech­
Feldman, D. C. 1989. Socialization, resocial­ niques: Foundations, ed. R. Glaser, pp. 49-
ization, and training: reframing the research 83. Hillsdale, NJ: ErIbaum
agenda. See Goldstein 1989, pp. 376- Gist, M. E. 1989. The influence of training
416 method on self-efficacy and idea generation
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

Fendrich, D . W. , Healy, A . F. , Meiskey, L. . among managers. Pers. Psychol. 42:787-

Crutcher, R. J . , Little , W . , Bourne, L. E . 805
1988. Skill maintenance: literature review Gist, M. E . , Bavetta, A. G . , Stevens, C. K.
and theoretical analysis. Air Force Hum.
by University of Connecticut on 02/24/13. For personal use only.

1990. Transfer training method: its in­

Res. Lab. Tech. Rep. 87-73. Brooks, AFB, fluence on skill generalization, skill repeti­
TX: Air Force Systems Command tion, and performance level. Pers. Psychol.
Finney, M . , Von Glinow, M. A. 1 988. In­ 43:50 1-23
tegrating academic and organizatiunal Gist, M. E . , Schwoerer, C . , Rosen, B. 1989.
approaches to developing the international Effects of alternative training methods on
manager. J. Manage. Dev. 7 : 1 6-27 self-efficacy and performance in computer
Fleishman, E. A . , Mumford, M . D. 1989a. software training. 1. Appl. Psychol . 74:
Abilities as causes of individual differences 884--9 1
in skill acquistion. Hum. Perform. 2:201- Glickman, A . S . , Zimmer, S . , Montero, R .
23 c . , Guerette, P . J . , Campbell , e t al. 1987.
Fleishman, E. A . , Mumford, M . D . 1989b. The evolution of teamwork skills: an empir­
Individual attributes and training perfor­ ical assessment with implications for train­
mance. See Goldstein 1989, pp. 1 83- ing. Tech Rep. NTSC 87-0 1 6 . Arlington,
225 V A: Office of Naval Research
Fleishman, E. A . , Zacaro, S. J. 1 99 1 . Toward Goldstein, A. P . , Sorcher, M. 1974. Chang­
a taxonomic classification of team perfor­ ing Supervisor Behavior. NY: Pergamon
mance functions: initial considerations, sub­ Goldstein, I. L. 1980. Training in work orga­
sequent evaluations, and current formula­ nizations. Annu. Rev. Psychol. 3 1 :229-
tions. See Swezey & Salas. In press 72
Fleming, R. K . , Sulzer-Azaroff, B . 1 990. Goldstein, I. L . , ed. 1989. Training and De­
Peer management: effects un staff teaching velopment in Organizations. San Francisco:
performance. Presented at the 15th Annu. Jossey-Bass
Convent. Assoc. Behav. Anal . , Nashville Goldstein, I. L. 1 99 1 . Training in organiza­
Fletcher, J. D. 1 98 8. Intelligent training sys­ tions. In Handbook of Industrial/Organiza­
tems in the military. In Defense Applica­ tional Psychology. ed. M . D. Dunnette, L .
tions ofArtifiCial Intelligence: Progress and M . Hough, Vol. 2. 2nd e d . In press
Prospects, ed. S . J. Andriole, G . W . Hop­ Goldstein, I. L. , Braverman, E. P. , Goldstein,
ple, pp. 33-59. Lexington, MA: Lexington H . W. 1 99 1 . The use of needs assessment in
Books training systems design. In ASPAIBNA
Fletcher, J. D. 1990. Effectiveness and cost of Handbook of Human Resource Manage­
interactivc videodiscs in defense training ment, Vol. 5. Developing Human Re­
and education. Alexandria, VA: Inst. De­ sources, ed. K. Wexley, pp. 5-35-5-75.
fense Anal. Washington, DC:BNA Books
Ford, J . K. 1 990. Understanding training Goldstein, I. L . , Gessner, M. J. 1 988. Train­
transfer: the water remains murky. Hum. ing and development in work organizations.
Resour. Dev. Q. 1 :225-29 In International Review of Industrial and
Ford, J. K . , Noe, R. A. 1987. Self-assessed Organizational Psychology. ed. C . L.
training needs: the effects of attitudes Cuuper, I. Robertson, pp. 43-72 . Lundon:
toward training, managerial level, and func­ Wiley
tion. Pers. Psychol. 40:39-53 Goldstein, I. L . , Gilliam, P. 1 990. Training
Ford, J. K . , Quinones, M. , Sego, D . , Speer, system issues in the year 2000. Am. Psy­
J. 1 99 1 . Factors affecting the opportunity to chol. 45 : 1 34-43
use trained skills on the job. Presented at the Goodman, P. S . , Miller, S. M. 1990. Design­
6th Annu. Conf. Soc. Indust. Org . Psy­ ing effective training through the technical
chol. , St. Louis life cycle. Natl. Prod. Rev . 9 : 1 69-77
 TRAINING & DEVELOPMENT 437

Gordon, J. 1 988. Who is being trained to do Jacobs, J. W . , Prince, C . , Hays, R. T . , Salas,

what? Train. Mag. Oct : 5 1 -60 E. 1 990. A meta-analysis of the flight sim­
Green, A . J . K . , Gilhooly, K . J . 1 990. In­ ulator training research. NAVTRASY­
dividual differences and effective learning SCEN TR-89-006. Orlando, FL: Naval
procedures: the case of statistical comput­ Training Systems Center
ing. Int. 1. Man-Mach . Stud. 33:97-1 1 9 Jacobs, R. L. , Baum, M . 1 987. Simulation
Griffin , R . W . , Cashin , W. E . 1989. The and games in training and development.
lecture and discussion method for manage­ Simul. Games 1 8 :385-94
ment education: pros and cons. 1. Manage. Jonas, H. S. III, Fry, R. E . , Srivasta, S. 1990.
Dev. 8:25-32 The office of the CEO: understanding the
Hackman, J. R. 1 989. Groups That Work (And executive experience. Executive 4: 36-47
Those That Don't). San Francisco: Jossey­ Jones, J. W . , Steffy, B. D . , Bray, D. W. eds.
Bass 1 99 1 . Applying Psychology in Business.
Hamel, C. J . , Kotick, D . , Layton, M. 1 989. Lexington, MA: Lexington Books
Microcomputer system integration for air Kahneman, D. 1 973. Attention and Effort. En-
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

control training Spec Rep. SR89-0 I . Orlan­ glewood Cliffs , NJ: Prentice Hall
do, FL: Naval Training Systems Center Kanfer, F. H. , Hagerman, S. 1 987. A model
Helmreich, R. L . , Wilhelm, J. A. 1 989. When of self regulation. In Motivation, Intention,
by University of Connecticut on 02/24/13. For personal use only.

training boomerangs: negative outcomes and Volition, ed. F. Halish, J. Kuhl, pp.
associated with cockpit resource manage­ 297-307. New York: Springer-Verlag
ment programs . In Proceedings of the Fifth Kanfer, R. 1 99 1 . Motivation theory and in­
Symposium on Aviation Psychology, ed. R. dustrial and organizational psychology. See
S. Jensen, pp. 692-97. Columbus, OH: Dunnette & Hough 1 99 1 , pp. 75-170
Ohio State Univ. Kanfer, R . , Ackerman, P. L . 1 989. Motiva­
Helmreich , R. L . , Wilhelm, J . A . , Gregorich, tion and cognitive abilities: an integrative/
S. E . , Chidester, T. R. 1 990. Preliminary aptitude-treatment interaction approach to
results from the evaluation of cockpit re­ skill acquisition. 1. Appl. Psychol. Monogr.
source management training: performance 74:657-90
ratings of flightcrews. Aviation, Space, En­ Kaplan, R. E. , Kofodimos, J. R . , Drath, W.
viron. Med. 6 1 :576-79 H. 1 987. Development at the top: a review
Henry, R. A . , Hulin, C. I. 1 987. Stability of and a prospect. In Research on Organiza­
skilled performance across time: some tional Change and Development. ed. W.
generalizations and limitations on utilities. Pasmore, R. W. Woodman, pp. 229-73.
1. Appl. Psychol. 72:457-62 Greenwich, CT: JAI Press
Hicks, W. D . , Klimoski, R. J. 1 987. Entry Kelleher, D . , Finestone, P. , Lowry, A. 1 986.
into training programs and its effects on Managerial learning: first notes from an un­
training outcomes: a field experiment. studied frontier. Group Org. Stud. 1 1 : 1 69-
Acad. Manage. 1. 30:542-52 202
Hill, T . , Smith, N. D . , Mann, M. F. 1 987. Keys, B . 1 987 . Total enterprise business
Role of efficacy expectations in predicting games. Simul. Games 1 8:225-4 1
the decision to use advanced technologies: Keys, B . , Wolfe, J. 1988. Management
the case of computers. J. Appl. Psychol. education and development: current issues
72:307- 1 3 and emerging trends. 1. Manage. 1 4:205-
Hinrichs, J . R. 1 976. Personnel training. In 29
Handbook of Industrial and Organizational Keys, B . , Wolfe, J. 1 990. The role of man­
Psychology, ed. M. D. Dunnette, pp. 829- agement games and simulation in education
60. Chicago: Rand-McNally and research. 1. Manage. 16:307-36
Hogan, P. M . , Hakel, M. D . , Decker, P. J . Kinzie, M. B. 1 990. Requirements and bene­
1 986. Effects o f trainee-generated versus fits of effective interactive instruction:
trainer-provided rule codes on generaliza­ learner control , self-regulation and continu­
tion in behavior-modeling training. J. Appl. ing motivation. Educ. Tech . Res. Dev.
Psycho!. 7 1 :469-73 38: 1-2 1
Howard, G. S . , Dailey, P. R. 1 979. Re­ Kirkpatrick, D. L. 1 976. Evaluation of train­
sponse-shift bias: a source of contamination ing. In Training and Development Hand­
of self-report measures. J. Appl. Psychol. book, ed. R. L. Craig, pp. 1 8 - 1 - 1 8-27.
Meas. 64: 144-50 New York: McGraw-Hili. 2nd ed
Howell, W . c . , Cooke, N . J. 1989. Training Kraiger, K . 1 988. Implications of research on
the human information processor: a review expert/novice differences for training
of cognitive models. See Goldstein 1989, assessment and design. Paper presented at
pp. 1 2 1-82 the Annu. Meet. Soc. Indust. Drg . Psycho! .
Hsu, E . 1 989. Role-event gaming and simula­ Kram, K. E. 1 985. Mentoring and Work: De­
tion in management education: a conceptual velopment Relationships in Organizations.
framework. Simul. Games 20:409-38 Greenview, l L : Scott Foresman
 438 TANNENBAUM & YUKL

Kraut, A. I . , Pedigo, P. R . , McKenna, D. D . , Helping Employees Make Tough Career

Dunnette, M. D . 1 989. The role o f the man­ Decisions. San Francisco: Jossey-Bass
ager: what ' s really important in different Luthans, F . , Kreitner, R. 1 985. Organization­
managerial jobs. Acad. Manage. Exec. al Behavior Modification and Beyond, Illi­
3:286-93 nois: Scott, Foresman & Co
Kuhl, J. 1 985. Volitional mediators of cogni­ Manz, C . , Sims, H . 1 989. Superleadership:
tion-behavior consistency: self-regulatory Leading Others to Lead Themselves. New
processes and action versus state orienta­ York: Simon & Schuster
tion. See Kuhl & Beckman 1 985, pp. 1 1-39 Marx, R. D. 1 982. Relapse prevention for
Kuhl, J . , Beckman, J. eds. 1 985. Action Con­ managerial training: a model for mainte­
trol: From Cognition to Behavior. Berlin: nance of behavior change. Acad. Manage.
Springer-Verlag Rev. 7:433-41
Kyllonen, P. c . , Shute, V. J. 1 989. A taxon­ Marx, R. D . , Hamilton, E. E. 1 99 1 . Beyond
omy of learning skills. See Ackerman et al skill building: a multiple perspective view
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

1 989, pp. 1 1 7-63 of personnel training. Issues Trends Bus.

Laker, D. R. 1 990. Dual dimensionality of Econ. In press
training transfer. Hum. Resour. Dev. Q. Marx, R. D . , Karren, R. K. 1988. The effects
1 :209-23 of relapse prevention training and in­
by University of Connecticut on 02/24/13. For personal use only.

Lance, C. E " Mayfield, D, L . , Gould, R, B " teractive followup on positive transfer of

Lynskey, M . C . 1 99 1 . Global versus de­ training. Paper presented at the Nat!. Acad.
composed estimates of cross-job retraining Manage. Meet.
time. Hum. Perform. Vo!. 4 : 7 1 -88 Marx, R, D . , Karren, R. J. 1990. The effects
Lassiter, D. L. , Vaughn, J. S . , Smaltz, V. E., of relapse prcvcntion and post-training fol­
Morgan, B . B . 1 990. A comparison of two lowup on time management behavior. Pre­
types of train in g interventions on team com­ sented at the Annu� Meet� Acad. Manage � ,
munication performance. Presented at the San Francisco
Meet. Hum. Factors Soc. Marsick, V. 1 990. Experience-based learning:
Latham, G . P. 1 988. Human resource training executive learning outside the classroom. J.
and development. Annu. Rev. Psychol. 39: Manage. Dev. 9:50-60
545-82 Mathieu, J. E . , Leonard, R. L. 1 987. Apply­
Latham, G. P. 1989 . Behavioral approaches to ing utility concepts to a training program in
the training and learning process. See Gold­ supervisory skills: a time based approach.
stein 1 989, pp. 256-95 Acad. Manage. J. 30:31 6-35
Latham, G. P . , Frayne, C. A. 1 989. Self­ Mathieu , J. E . , Tannenbaum, S. I . , Salas, E .
management training for increasing job at­ 1 990. A causal model of individual and
tendance: a follow-up and replication. J. situational influences on training effective­
Appl. Psycho!. 74:4 1 1 - 1 6 ness measures. Presented at the 5th Annu.
Lepsinger, R . , Mullen, T. P . , Stumpf, S . A . , Conf. Soc. Indust. Org. Psycho!. , Miami
Wall, S . J . 1 9 8 8 . Large scale management Maxwell , S . E. , Cole, D. A . , Arvey , R. A . ,
simulations: a training technology for Salas, E . 1 99 1 . A comparison o f methods
assessing and developing strategic manage­ for increasing power in randomized be­
ment skills. In The Practice of Management tween-subjects designs. Psychol. Bull. In
Development, ed. S . Mailick, S . Hober­ press
man, S . J . Wall, pp. 1 73-83. New York: Mayer, S. J . , Russell, J. S. 1987. Behavior
Praeger modeling training in organizations: con­
Levine, E, L . , Baker, C. V. 1 99 1 . Team task cerns and conclusions. J. Manage . 1 3:21-
analysis: a procedural guide and test of a 40
methodology. Paper presented at Annu. McCall, M. W. Jr. , Lombardo , M. M. 1982.
Meet. Soc. Indust. Org. Psycho! . Using simulation for leadership and man­
Lippert, R, C, 1 989. Expert systems: tutors, agement research: through the Looking
tools and tutees. J. Computer-Based Instr. Glass. Manage. Sci� 28:533-49
16:1 1-19 McCall, M. W. Jr. , Lombardo, M. M . , Morri­
London, M . 1 987. Employee development in son, A. M. 1 988. The Lessons of Experi­
a downsizing organization . J. Bus. Psychol. ence. Lexington, MA: Lexington Books
2:60-73 McCauley, C. D . , Lombardo, M. M . , Usher,
London, M . 1 989. Managing the Training En­ C. J. 1989. Diagnosing management de­
terprise. San Francisco: Jossey-Bass velopment needs: an instrumcnt based on
London, M . , Bassman, E. 1 989. Retraining how managers develop. J. Manage. 1 5 :
midcareer workers for the future work­ 389-403
place. See Goldstein 1 989, pp. 333-75 McGehee, W . , Thayer, P. W. 1 96 1 . Training
London, M . , Mone, E. M. 1 987. Career Man­ in Business and Industry. New York: Wiley
agement and Survival in the Workplace: Mills, G. E . , Pace, R. W. 1 989. What effects
 TRAINING & DEVELOPMENT 439

do practice and video feedback have on the Pellum, M . W., Teachout, M. S. 1 990. A
development of interpersonal communica­ longitudinal evaluation of training effective­
tion skills. J. Bus. Commun. 26: 1 59-76 ness using multiple levels of information.
Modrick, J. A. 1 986. Team performance and Presented at the Annu. Meet. Soc. Indust.
training. In Human Productivity Enhance­ Org. Psycho!.
ment. Vo!' 1 : Training and Human Factors Pentland, B . T. 1 989. The leaming curve and
in Systems Design, ed. J. Zeidner, pp. 1 30- the forgetting curve: the importance of time
66. New York: Praeger and timing in the implementation of tech­
Morgan, B . B . , Salas, E. 1 988. A research nological innovations . Presented at the 49th
agenda for team training and performance: Annu. Meet. Acad. Manage . , Washington,
issues, alternatives, and solutions. Proc. /n­ DC
terserv .!/ndust. Train. Sys. ConI 10:560- Peters, L. H . , O'Connor, E. J . , Eulberg, J .
65 R . , Watson, T. W. 1 988. A n examination
Morrison, J. E . , Holding, D. H. 1 990. De­ of situational constraints in Air Force work
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

signing and gunnery training strategy. ARI settings. Hum. Perform . 1 : 1 33-44
Tech. Rep. 899. Alexandria, VA: Army Pintrich, P. R . , Cross, D. R . , Kozma, R. B . ,
Res. Inst. Behav. Soc. Sci. McKeachie, W. J . 1 986. Instructional psy­
Mumford, M . D . , Weeks, J . L . , Harding, F. chology. Annu. Rev. Psychol. 3 7 : 6 1 1-5 1
by University of Connecticut on 02/24/13. For personal use only.

D . , Fleishman, E. A. 1 987. Measuring Porter, L. W . , McKibbin, L. 1 989. Future of

occupational difficulty: a construct valida­ Management Education and Development.
tion against training criteria. J. Appl. Psy­ New York: McGraw-Hill
chol. 72:578-87 Pursell, E. D . , Russell, J. S. 1 99 1 . Employee
Mumford, M. D . , Weeks, J. L . , Harding, F. development. In ASPAIBNA Handbook of
D . , Fleishman, E. A. 1 988. Relations be­ Human source Management. Vo!. 5 . De­
tween student characteristics, course con­
tent, and training outcomes: an integrative
modeling effort. J. Appl. Psychol. 72:578-
J
veloping uman Resources, ed. K. Wexley,
pp. 5-7 . 5- 1 1 9 . Washington , DC: BNA
Books I I
87 Ragins, Ii. �. 1 989. Barriers t o mentoring: the
Murphy, K. R. 1989. Is the relationship be­ female manager's dilemma. Hum. Relat.
tween cognitive ability and job performance 42: 1-22 \
stable over time? Hum. Perform. 2 : 1 83-200 Reilly, R. R . , Israelski, E. W. 1 988. Develop­
Noe, R. A. 1 9R6. Trainees' attributes: ne­ ment and validation of minicourses in the
glected influences on training effectiveness. telecommunication industry. J. Appl. Psy­
Acad. Manage. Rev. 1 1 :736-49 chol. 73:72 1-26
Noe, R. A. 1 988a. An investigation of the Revans, R. W. 1 982. The Origin and Growth
determinants of successful assigned mentor­ of Action Learning. Hunt, England: Chat­
ing relationships. Pers. Psychol. 4 1 :457-79 well-Bratt, Bickley
Noe, R. A. 1 988b. Women and mentoring: a Robeltson, I. T. 1 990. Behavior modelling: its
review and research agenda. Acad. Man­ record and potential in training and develop­
age. Rev. 1 3:65-78 ment. Br. J. Manage. 1 : 1 1 7-25
Noe, R. A. 1 99 1 . Mentoring relationships for Robertson, I. T . , Downs, S. 1 989. Work sam­
employee development. See Jones et al ple tests of trainability: a meta-analysis. J.
1 99 1 , pp. 475-82 Appl. Psychol. 74:402- 1 0
Noe, R. A . , Schmitt, N. 1 986. The influence Ronen, S . 1 989. Training the international
of trainee attitudes on training effectiveness: assignee. See Goldstein 1989, pp. 4 1 7-53
test of a mode!. Pers. Psycho!. 39:497-523 Rosow, J. M . , Zager, R. 1 988. Training-The
Nordhaug , O. 1989. Reward functions of per­ Competitive Edge. San Francisco: Jossey­
sonnel training. Hum. Relat. 42:373-88 Bass
Oser, R . , McCallum, G . A . , Salas, E . , Mor­ Rouillier, J . Z. , Goldstein, I . L. 1 99 1 . Deter­
gan, B . B . 1 989. Toward a definition of minants of the climate for transfer of train­
teamwork: an analysis of critical team be­ ing. Presented at the Meet. Soc. Indust.
haviors. Tech Rep. TR-89-004. Orlando, Org. Psycho!.
FL: Naval Training Systems Center Ruderman, M. N . , Ohlott, P. J . , McCauley,
Osigweh, C. A. B. 1 989. Casing the case C. D. 1 990. Assessing opportunities for
approach in management development. J. leadership development. See Clark & Clark
Manage. Dev. 8:41-57 1 990, pp. 547-62
Ostroff, c., Ford , J . K . 1989. Assessing train­ Russell , J. S . , Wexley, K. N. , Hunter, J . E .
ing needs: critical levels of analysis. See 1 984. Questioning the effectiveness o f be­
Goldstein 1 989, pp. 25-62 havior modeling training in an industrial
Parry, S. B . , Reich, L. R. 1 984. An uneasy setting. Pers. Psychol. 37:465-8 1
look at behavioral modeling. Train. Dev. J. Ryder, J. M . , Redding, R. E . , & Beckschi, P.
Mar.:57-62 F. 1987. Training development for complex
 440 TANNENBAUM & YUKL

cogmtIve tasks. Proc. 31st Annu. Meet. and developing the older worker: im­
Hum. Factors Soc. 2 : 1 26 1-65 plications for human resource management.
Saari, L. M . , Johnson, T. R . , Mclaughlin, S. In 14 Steps to Managing an Aging Work­
D . , Zimmerle, D. M. 1 988. A survey of jorce, ed. H. Dennis. Lexington, NY
management training and education prac­ Sterns, H . L . , Doverspike, D. 1 989, Aging
tices in U. S. companies. Pers. Psychol. and the training and learning process. See
4 1 :73 1-43 Goldstein 1989, pp. 299-332
Sackett, P. R . , Zedeck, S . , Fogli, L. 1 988. Streufert, S . , Pogash, R . , Piasecki, M. 1 988.
Relations between measures of typical and Simulation-based assessment of managerial
maximum job performance. J. Appl. Psy­ competence: reliability and validity. Pers.
chol. 73:482-86 Psychol. 4 1 :537-57
Salas, E . , Dickinson, T. L . , Converse, S. A . , Stumpf, S. A . , Dunbar, R. L. M. 1 990. Using
Tannenbaum, S. 1. 1 99 1 a. Toward an un­ behavioral simulations in teaching strategy
derstanding of team performance and train­ implementation. Organ. Behav. Teach.
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

ing. In Teams: Their Training and Perform­ Rev.

ance, ed. R . W. Swezey & E. Salas. New Sundstrom, E . , DeMeuse, K. P . , Futrell, D .
Jersey: Ablex. In press 1 990. Work teams: applications and
Salas, E . , Dickinson , T. L . , Tannenbaum, S . effectiveness. Am. Psychol. 45 : 1 20-33
by University of Connecticut on 02/24/13. For personal use only.

1 . , Converse, S . A . 1 99 1 b. A meta-analysis Swanson, H. L. 1990. Influence of metacogni­

of team performance and training. Tech. tive knowledge and aptitude problem solv­
Rep. Orlando, FL. Naval Training System ing. J. Educ. Psychol. 82:306-14
Center. In press Swezey, R. W . , Salas, E . , eds. 1 99 1 a. Teams:
Savage, R. E . , Williges, B. H . , Williges, R. Their Training and Performance. Norwood,
C. 1 982. Empirical prediction models for NJ: Ablex. In press
training group assignment. Hum. Factor. Swezey, R. W . , Salas, E. 1 99 1 b. Guidelines
24:41 7-26 for use in team training development. See
Schneider, B . , Konz, A. 1 989. Strategic job Swezey & Salas 1 99 1 a. In press
analysis. Hum. Resour. Manage. 28:51-63 Tannenbaum, S. 1 . , Beard, R. L . , Salas, E .
Schoorman, F. D . , Schneider, B. 1988, 1 99 1 a. Team building and its influence on
Facilitating Work Effectiveness. Lexington, team effectiveness: an examination of con­
MA: Lexington Books ceptual and empirical developments In
Schuler, R. S . , Jackson, S. E. 1 987. Organi­ Issues, Theory, and Research in Industrial!
zational strategy and organizational level as Organizational Psychology, ed. K. Kelley.
determinants of human resource manage­ Amsterdam: Elsevier. In press
ment practices. Hum. Resour. Plan. 1 0: Tannenbaum, S. I . , Mathieu, 1. E . , Salas , E . ,
1 23-4 1 Cannon-Bowers, J . A . 1991b. Meeting
Snow, R. E. 1989. Cognitive-conative apti­ trainees' expectations: the influence of
tude interactions in learning, In Abilities, training fulfillment on the development of
Motivation, and Methodology: The Minne· commitment, self-efficacy, and motivation.
sota Symposium on Learning and Individual 1. Appl. Psychol. In press
Differences, ed. R . Kanfer, P. L. Acker­ Taylor, M. S . , Locke , E. A. , Lee, c . , Gist,
man, R. Cudeck, pp. 435-74. Hillsdale, M. E. 1 984. Type A behavior and faculty
NJ: Erlbaum research productivity: What are the mech­
Snow, R , E . , Lohman, D. F. 1 984. Toward a anisms? Organ. Behav. Hum . Perform.
theory of cognitive aptitude for learning 34:402-1 8
from instruction . 1. Educ. Psychol. 76:347- Tetrault, L. A . , Schrieshiem, C . A . , Neider,
76 L. L. 1 988. Leadership training interven­
Sonnenfeld, J . A . , Peiperl, M. A. 1 988. Staff­ tions: a review. Orgall. Dev. J. Fal1:77-83
ing policy as a strategic response: a typolo­ Thayer, P. W. 1 989. A historical perspective
gy of career systems. Acad. Manage. Rev, on training. See Goldstein 1 989, pp. 457-
1 3 :588-600 68
Sparrow, J. A. 1 989. The measurement of job Thornton, G. C. III, Cleveland, J, N. 1 990.
profile similarity for the prediction of trans­ Developing managerial talent through sim­
fer of learning: a research note. J. Occup , ulation. Am. Psychot: '45 : 1 90-99
Psychol. 62:337-4 1 Thorpe , J. A. 1 987. The new technology of
Sprangers, M . , Hoogstraten , J. 1 988. On de­ large scale simulation networking: im­
lay and reassessment of retrospective plications for mastering the art of nonfight­
ratings. J. Exp . Educ. 56: 1 48-53 ing. ProC . 9th Interserv.llndust. Train. Syst.
Sprangers, M . , Hoogstraten , J. 1 989. Pretest­ Conf.
ing effects in retrospective pretest-posttest Trimble, S. K . , Decker, P. J . , Nathan, B. R .
designs. J Appl. Psychol, 74:265-72 1 989, Effect o f positive and negative mod­
Stems, H. L . , Doverspike , D. 1 988. Training els on learning in behavior modeling train-
 TRAINING & DEVELOPMENT 441

ing. Presented at the 4th Annu. Conf. Soc. ing. Presented at the 6th Annu. Conf. Soc.
Indus!. Org. Psychol . , Boston Indus!. Org. Psychol . , S!. Louis
Tubbs, M. E . , Ekeberg, S. E. 1 99 1 . The role Wlodkowski, R. J. 1985 . Enhancing Adult
of intentions in work motivation: im­ Motivation to Learn. San Francisco: Jossey­
plications for goal-setting theory and re­ Bass
search. Acad. Manage. Rev. 1 6: 1 80-99 Wolfe, J. 1 990. The evaluation of computer­
Tung, R. 1 988. The New Expatriates: Manag­ based business games. Methodology, find­
ing Human Resources Abroad. Cambridge, ings and future needs. In ABSEL Guide to
MA: Harper & Row Experiential Learning and Simulation Gam­
Van Velsor, E . , Ruderman, M . , Phillips, A. ing. ed. J. W. Gentry. New York: Nichols
D. 1989. The lessons of the Looking Glass. Publishing
Leader. Org. Dev. J. 10:27-3 1 Wolfe, J . , Roberts, C. R. 1 986. The external
Von Glinow, M. A . , Milliman, J. 1990. De­ validity of a business management game: a
veloping strategic international human re­ five year longitudinal study Simul. Games
Annu. Rev. Psychol. 1992.43:399-441. Downloaded from www.annualreviews.org

source management: prescriptions for multi­ 1 7:45-59

national company success. J. Manage. Woods, S . , Tannenbaum, S. I. 1 990. Evaluat­
Issues 2:9 1-104 ing supervisory training: a case study. Pre­
Weiss, H. M . 1 99 1 . Learning theory and in­ sented at the Conf. Soc . Indust. Org. Psy­
by University of Connecticut on 02/24/13. For personal use only.

dustrial and organizational psychology . See cho! .

Dunnette & Hough 1 99 1 , pp. 1 7 1-222 Yuki. G. 1 989. Leadership in Organizations.
Wexley, K. N. 1 984. Personnel training. Englewood Cliffs, NJ: Prentice-Hall
Annu. Rev. Psychol. 35:5 1 9-5 1 Yuki, G . , Lepsinger, R. 1 99 1 . An integrative
Wexley, K. N. , Baldwin, T. T. 1986. taxonomy of managerial behavior: implica­
Posttraining strategies for facilitating posi­ tions for improving managerial effective­
tive transfer: an empirical exploration. ness. See Jones et al 1 99 1 , pp. 563-72
Acad. Manage. J. 29:503-20 Yuki, G . , Wall, S . , Lepsinger, R. 1 990. Pre­
Wexley, K. N . , Latham , G. 1 99 1 . Developing liminary report on validation of the man­
and Training Human Resources in Organ­ agerial practices survey. See Clark & Clark
izations. Glenview, IL: Scott-Foresman. 1 990, pp. 223-37
2nd cd Zeitz, C. M . , Spoehr, K. T. 1989. Knowledge
Willbur, J. 1 987. Does mentoring breed suc­ organization and the acquisition of pro­
cess? Train. Dev. J. Nov . : 38-41 cedural expertise. Appl. Cogn. Psychol.
Williams, T. c . , Thayer, P. W . , Pond, S. B. 3 :3 1 3-36
1 99 1 . Test of a model of motivational in­ 'Ley, M. G. 1 988. A mentor for all reasons.
fluences on reactions to training and learn- Pers. J. Jan. :46-5 1