Review of Educational Research

Winter 1982, Vol. 52, No. 4, Pp. 499-512

Characteristics of Media Selection Models

Robert A. Reiser and Robert M. Gagne
Florida State University

This paper identifies and evaluates the learning effectiveness of the major
features found in media selection models. The 10 different models employed as
examples are not described individually, as is done in some previous reviews.
Instead, the article focuses on the characteristics noted across models. Features
discussed include the physical forms the models take, the ways in which they
classify media, and the media selection factors they consider. Selection factors
embodied in models affect media choices. Characteristics of learners, setting,
and task are identified as factors to be given primary consideration in media
selection.

For well over half a century, educators have been concerned about choosing the
appropriate media to deliver an instructional message. E. L. Thorndike (1912) expressed
a basic viewpoint about media selection as follows:

A human being should not be wasted in doing what forty sheets of paper or two
phonographs can do. Just because personal teaching is precious and can do what
books and apparatus cannot, it should be saved for its peculiar work. (p. 167)

Concerns such as that expressed by Thorndike still exist. Schramm (1977) points out that
educators throughout the world expend a great deal of effort attempting to choose the
appropriate media to use in particular instructional situations.
How should educators go about selecting media? A number of media selection models
have been developed in an attempt to answer this question. Information concerning the
usefulness of these models, however, is limited. Although descriptions of some models
include examples of how they have been used (e.g., Briggs & Wager, 1981) or how they
have been field tested (Reiser & Gagne, Note 1), it is rare to find detailed information
about situations in which selection models have been employed. Two studies have been
identified in which media selection techniques are compared. In one instance, Braby
(1973) compared the usefulness of 10 media selection techniques. Models developed by
Briggs (1970) and the Training Analysis and Evaluation Group (1972), as well as an
intuitive approach, were judged superior to the other techniques.1 In another study,
Romiszowski (1974) found that teacher

1
Three of the media selection models reviewed in this paper are directly related to
techniques examined by Braby (1973). These include the model described by Bretz
(1971), which was evaluated in Braby's study; the model described by Branson et al.
(1975), which was based in part on the Training Analysis and Evaluation Group model
(1972); and the model described by Briggs and Wager (1981), which was based on an
earlier model developed by Briggs (1970).

499

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

trainees who used a media selection technique that he developed made better media
choices than did experienced teachers who used an intuitive approach. In view of the
limited empirical evidence regarding the relative merits of media selection models, the
question of which model to employ cannot have a simple answer. An appropriate
approach would be to identify the features incorporated in the various models, decide
which features are important, and select a model contain ing them.
This paper identifies and evaluates the relevance to learning effectiveness of the major
features found in media selection models. Unlike previous reviews (e.g., Heidt, 1978,
Levie, 1977), various models will not be described individually. Instead, the focus will be
on the characteristics noted across models. This approach invites attention to specific
features as opposed to specific models. The features reviewed here include the physical
forms the models take, the ways in which they classify media, and the media selection
factors they consider. Our primary emphasis will be on selection factors because of their
relation to the effectiveness of learning.
To identify the major features incorporated in media selection models, 10 models have
been reviewed. These include some of the most widely known models, as well as some of
the most recent ones. The models examined are those described by Anderson (1976),
Branson et al. (1975), Bretz (1971), Briggs and Wager (1981), Gagne and Briggs (1979),
Gropper (1976), Kemp (1980), Reiser and Gagne (Note 1), Romiszowski (1974), and
Tosti and Ball (1969). In subsequent sections of this paper, models will be identified only
by authors' names, omitting the date of the publication in which they are described, as
given in the previous sentence.

Display Form and Mode of Decisionmaking

Several different display forms have been used to present the 10 models. The basic
elements are generally presented in flowchart, matrix, or worksheet form. An essential
difference among these forms, other than their physical differences, is the mode of
decisionmaking each requires.
When a. flowchart is employed (e.g., Anderson; Bretz; Kemp; Reiser & Gagne;
Romiszowski), the media selection process consists of a progressive narrowing of media
choices. Questions about factors in the media selection process are posed in a specific
order, and as each question is answered, the number of candidate media decreases. When
the last question in a flowchart sequence is answered, the designer is left with a small set
of media from which to choose.
Media choices made with the use of a matrix (e.g., Branson et al.) are deferred until all
selection criteria are displayed. At that point, the designer, using such factors as the
number of criteria met and the relative importance of the various criteria, chooses from
among all the media listed within the matrix.
Media selection worksheets typically present a tabular array of media characteristics
against desired criteria. Worksheets (e.g., Briggs & Wager; Gagne" & Briggs) custom
arily require that selection decisions be deferred until all criteria have been consid ered.
Presumably, worksheets employing a progressive elimination approach to media
selection would be possible, although none has been proposed.
Regardless of their format, most of the models suggest that media should be selected
for individual instructional events (Branson et al.; Bretz; Briggs & Wager; Gagn6 &
Briggs; Gropper) or for single objectives (Romiszowski; Tosti & Ball).

500

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

Categories of Media
Several of the media selection models group media into categories. In some models
(Anderson; Branson et al.; Bretz), categories are deliberately distinguished and labeled. A
list of the various categories, and the media included in each, appears as a separate
element of these models. In other models (Kemp; Reiser & Gagn6), the categories of
media are not named, but may be inferred by noting common charac
teristics of the various media listed as appropriate for particular conditions of instruction.
Whether labeled or not, the media categories used in the various models are often
similar. Categories frequently employed include audio, print, still- and motion-visual
(with and without audio), and real objects. In addition, instructions are sometimes
referred to as a separate category of media (Bretz; Briggs & Wager; Reiser & Gagne).
Most of the proposed categories are based on the stimulus characteristics of media.
However, media may be categorized along other dimensions, such as the types of
responses they will accept and the ways in which they can adapt presentations on the
basis of learner responses (Tosti & Ball). It may also be useful to categorize media
according to their feedback capabilities (Gropper; Reiser & Gagne), and in terms of their
ability to accommodate other events of instruction (Gropper).

Physical Attributes of Media

As indicated in Table I, all the models require the designer to consider the media
attributes (i.e., physical capabilities of media) that are necessary in a given instance. For
example, a designer may have to decide whether the medium to be employed in a
particular instructional situation must be capable of depicting motion, although the
conceptual basis for this decision may not be included.
The models vary in the extent to which they stress direct consideration of media
attributes. For instance, most of the questions included in Kemp's model require the
designer to identify the media attributes that are appropriate (e.g., "audiovisual or
multi-image technique necessary?"). In contrast, the model developed by Reiser and
Gagne contains few questions concerning the media attributes necessary for a given
instructional display. Instead, the designer is expected to attend to the characteristics of
the intended learners, the instructional setting, and the learning task.

Visuals
All the models ask the designer to consider whether visual displays are necessary.
Some of the models distinguish among various types of visuals (Branson et al.; Kemp;
Tosti & Ball). For example, the Instructional Systems Development (ISD) model
(Branson et al.) requires the designer to choose between alpha-numeric, graphic,
pictorial, and three-dimensional visual displays.
Reiser and Gagne suggest that a primary reason for considering the use of visual
displays is that they can help the learner acquire and retain visual images, which support
the acquisition of intellectual skills and may also aid retention of verbal information. In
several other models, visual media are recommended to help students acquire particular
kinds of intellectual skills, such as identifying objects (Anderson; Bretz; Romiszowski)
and classifying spatial relationships (Bretz).
Visual media are also considered helpful in teaching students how to perform 501

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

5

TABLE I
Factors Prominent in Various Media Selection Models
Models
Factors

Learner, setting, & task characteristics:

Anderson Branson „
«. , Bretz
et a l
Briggs & Wager
G a n 6 & Cropper Kem
R S. Briggs rr „ Reiser & P Gagne
Romis- Tosti & zowski Ball

Instructional setting X X X X X X X X Learner characteristics X X X X X X X Categories of

learning outcomes X X X X X X X Events of instruction X X X X X X X X
Physical attributes X X X X X X X X X X Practical factors X X X X X X X X X X
Note. An X indicates that the factor was prominent in a model. A blank space indicates
that a model and its accompanying description either did not mention the factor or gave it
little attention.

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

various motor skills. Bretz points out that words alone are usually an inefficient
means of teaching a motor skill; visual demonstration is necessary. Some
models indicate that the use of visual media may need to be made contingent
on learner characteristics (Briggs & Wager; Reiser & Gagne). Reiser and Gagne
remark that visual presentations have been shown to help learners who are poor
readers.

Printed Words
The requirement for presenting printed words is recognized by all the models
reviewed. The use of printed presentations is seen as a media characteristic
appro priate for learners who are good readers. Such displays may be made
contingent on tested reading ability (Bretz), or in the case of young students, on
increasing age (Gagn6 & Briggs).
Several models suggest that audio narration, as opposed to print, should be used
with poor readers (Bretz; Briggs & Wager; Reiser & Gagne). There is some
disagree ment, however, as to whether audio narration is superior to print for
affective objectives. Bretz states that audio may be superior, but Anderson
indicates that print materials (including printed text, drawings, and
photographs) are appropriate for teaching affective objectives. Other models
indicate that if sound is not an integral part of the topic, print may be
appropriate (Anderson; Romiszowski). Branson et al. point out that print
materials can be produced at a low cost.

Sound
The capability of presenting sound is a media attribute included in several of the
models. A distinction is sometimes drawn between verbal and nonverbal sounds
(Branson et al.; Bretz; Gagne & Briggs; Kemp; Tosti & Ball). Sound media are
considered necessary to present the appropriate stimulus information if the goal
of instruction is the recall or recognition of the sounds themselves (Anderson;
Bretz; Reiser & Gagne; Romiszowski). Bretz remarks that sound media are
essential if the instructional goal is the appreciation of musical compositions,
spoken poetry, or other sounds.
It is sometimes suggested that learner characteristics be taken into account
when deciding whether sound is necessary (Bretz; Briggs & Wager). Bretz
states that because poor readers can understand spoken words more easily than
written words, sound media should be used with these learners.

Motion
Pictorial motion is a media attribute addressed by many of the models. In some,
specific distinctions about the type of motion are proposed. For example, the
ISD model (Branson et al.) asks the designer to choose among still visuals,
limited movement visuals, and full movement visuals. Somewhat similar
motion types are distinguished by Kemp.
Generally, the reasons suggested for considering the attribute of pictorial motion
pertain to the depiction of human performance and the associated idea that
learners will recognize or copy the movements shown (Anderson;
Romiszowski). While asserting that motion is often unnecessary, Bretz lists
three questions pertaining to the desirability of motion. First, is the learning
objective the performance of a

503

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

procedure, the movements of which are unfamiliar to the learner? Second, is the learner
required to learn a concept which is understandable only in terms of its "manner of
movement?" And third, must the learner comprehend changes that take place too rapidly
or too slowly to be identified, and which therefore require slow motion or fast motion
displays?

Color
Only two of the models (Anderson; Branson et al.) require designers to decide whether
the media to be selected must be capable of displaying colors. Anderson indicates that if
the color of an object is relevant to the performance of some cognitive or psychomotor
objective, a medium able to display colors should be chosen. Several studies have
indicated that the use of color in instructional materials does not result in increased
learning unless color is relevant to what is being learned (Dwyer, 1978; Heidt, 1978;
Schramm, 1972).

Real Objects
Real objects, also known as realia, have been defined as "tangible objects, real items
(as opposed to representations or models) as they are without alterations" (Association
for Educational Communications and Technology, 1977, p. 293). Most of the media
selection models reviewed require the designer to decide whether real objects should be
used in a given instructional situation. Several models indicate that real objects should be
used to teach motor skills (Anderson; Reiser & Gagne; Romiszowski). Reiser and Gagne
provide a rationale for this choice, indicating that motor skills can be learned adequately
only if they are practiced directly on real equipment or on a training device which
incorporates the physical motion charac
teristics of the equipment. Using real equipment (or a simulator) to teach students how to
perform a task is particularly important if the consequences of task error on the job are
serious.
In addition to suggesting that real objects should be used to teach motor skills,
Anderson indicates that, if practical, real objects should be used to teach cognitive skills
involving objects unfamiliar to the learner. Romiszowski makes a similar point, but
neither author includes an explanation of this prescription. It is evident that a training aid
or device that reproduces the physical appearance of the real object may serve the
purpose of familiarization.

Learner, Setting, and Task Characteristics

The extent to which the models stress consideration of learner, setting, and task
characteristics varies greatly. As indicated in Table I, while each of the 10 models
requires consideration of at least one of the four factors we have grouped under the
general heading of learner, setting, and task, only 5 of the models require direct
consideration of all these factors.
Instructional Setting
In Table I, a model is classified as being concerned with instructional setting if it poses
one or more of the following questions: In what location (e.g., school, home) is the
instruction to be delivered? Is the instruction to be presented to individuals or to a group?
If a group is to receive the instruction, what is the size of the group? Questions of this
type are included in eight of the models.

504

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

Most of the models prescribe media for various settings and group sizes, but rarely
indicate why certain media have been recommended; only occasionally is a rationale
provided. For example, one suggestion is that when instruction is delivered in an
individualized mode, narration should be presented via print media because print allows
each learner to set his or her own pace, and thus allows for more efficient use of
instructional time (Bretz).
It is sometimes maintained that individual instruction should be delivered via media
capable of providing corrective feedback, so as to provide learners with information
about the adequacy of their responses (Bretz; Reiser & Gagn6). There is not perfect
agreement, however, as to which media are capable of providing corrective feedback.
Bretz indicates that almost any medium can do so simply by providing the correct answer
so that the learner can compare it with his or her own. Reiser and Gagne, on the other
hand, indicate that corrective feedback is best provided by interactive media such as
computers or programmed texts that make possible differential feedback displaying the
correct response as well as the degree of correctness of the learner's response.
Learner Characteristics
In many media selection models, learner characteristics are considered important to the
selection decision. Many educators believe that media may be differentially effective for
different types of learners, and therefore media that are best suited for various learner
types should be identified. Although investigators have had limited success in identifying
the media most suitable for various types of learners (Bracht, 1970; Cronbach & Snow,
1977), several media selection models offer suggestions for selecting media on this basis.
Reading ability is one learner characteristic that is frequently considered in media
selection models (Branson et al.; Bretz; Briggs & Wager; Gagne* & Briggs; Reiser &
Gagne). Two of the models (Briggs & Wager; Reiser & Gagne), suggest that pictures can
facilitate learning for those who are poor readers. Another point about poor readers is that
they may benefit more from spoken words than from written words (Bretz; Briggs &
Wager). Bretz indicates that spoken words may be superior because poor readers can
more easily understand them, especially if they are spoken in the learner's own idiom. On
the other hand, it is sometimes claimed that when good readers are working in a
self-instructional mode, print materials provide greater flexibility of pacing, and therefore
may offer a learning advantage over audio presentations (Bretz). Briggs and Wager
mention one advantage of print materials as that of allowing students to easily backtrack
to review points that were missed previously. This statement is presumably subordinate to
their remark that for poor readers, spoken words are superior to written ones.
Age is another learner characteristic often mentioned in media selection models
(Briggs & Wager; Gagne & Briggs; Reiser & Gagne). Older, more experienced learners
may have developed learning strategies that enable them to manage some aspects of
instruction for themselves. These learners, in contrast to younger or less experienced
learners, might need to have fewer instructional events deliberately provided for them
(Gagn6 & Briggs; Reiser & Gagne). Reiser and Gagne point out that as more externally
generated events are omitted, there is an increased dependence on self-instructional
methods; this in turn favors the selection of media that can provide corrective feedback.

505

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

Two of the media selection models (Briggs & Wager; Gagne & Briggs) mention
Dale's (1969) "Cone of Experience" as a tool that can be used to identify the
appropriate media for a given age group. The Cone lists 12 categories of media
and experiences in a hierarchical fashion. For cognitive objectives, Dale
suggests that it is efficient to use abstract media with older learners and
concrete media and experiences with younger learners.

Categories of Learning Outcomes

Classifying objects in terms of the categories of learning outcomes they
represent is an integral step in many media selection models. The number of
categories included in the models reviewed ranged from 3 (Anderson) to 11
(Branson et al.). Most of the models included some or all of the categories of
learning described by Gagne (1977): intellectual skills, verbal information,
motor skills, attitudes, and cognitive strategies.
Several models (Branson et al.; Briggs & Wager; Gagn6 & Briggs) are similar
in their prescription of a procedure for taking several steps from categorization
of learning outcomes to selection of media. In detail, the procedure entails
 categorizing learning outcomes, planning the instructional events to be used to
teach the learning outcomes, identifying the type of stimuli necessary to present
those events, and finally, identifying the media capable of presenting those
stimuli. The categorization of learning outcomes is a key step in this process;
decisions regarding the instructional events to be emphasized or included in a
lesson are conceived to be affected by the category of the learning outcome to
be taught.
Specific sets of possible media for various types of learning outcomes are
prescribed by some formulations (Anderson; Branson et al.; Reiser & Gagne;
Romiszowski). Reiser and Gagne include a rationale for the media choices
prescribed, which are summarized in the following paragraphs.
Intellectual skills. For learners to acquire intellectual skills, they must practice
the skills by applying them to newly presented examples, and must receive
precise corrective feedback indicating the degree to which their application
efforts were correct. Hardware devices most clearly capable of providing this
type of feedback are interactive media such as computers or interactive
videodisc systems.
Verbal information. Verbal information involves the recall of general ideas and
organized knowledge, rather than the verbatim restatement of specific facts. The
feedback requirements for this type of learning outcome are less rigorous than
for intellectual skills. In effect, feedback needs to say, when appropriate, "Yes,
you have the general idea." Media that can provide feedback in a somewhat
imprecise form because they are not interactive (e.g., a text or television
presentation programmed in a linear fashion) can be used to teach verbal
information.
Motor skills. Acquiring motor skills often involves learning an "executive
subrou tine" (Fitts & Posner, 1967), a form of intellectual skills learning.
Therefore, the feedback and media requirements for motor skills are often the
same requirements for intellectual skills, especially during the early stages of
instruction. Once the basic procedure that comprises a motor skill is learned,
however, improvement in the smoothness and timing of the skill requires that
the learner practice the skill directly and receive the kinesthetic feedback
inherent in such practice. Thus, at this later stage of the learning process, real
equipment or a realistic simulator are appropriate media choices.

506

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

Cognitive strategies. To help learners acquire new cognitive strategies, that is, to help
them learn how to learn, educators often present learners with problem-solving situations.
These situations usually require the recall and application of rules; therefore, the media
requirements for cognitive strategies are often the same as the requirements for
intellectual skills. In addition, some advantage may be seen in the kinds of media (e.g.,
television) that are able to present highly realistic "problem situations."
Other possibilities lie in the employment of media with visual displays to enhance
cognitive processing strategies (Olson, 1976), or to "cultivate mental skills," as proposed
by Salomon (1979). According to Salomon, the symbolic codes made possible by various
media can function to (a) activate a mental skill, (b) short-circuit it, or (c) overtly supplant
it. Supplantation is the process he sees as leading to the learner's acquisition of new
processing strategies. The potentialities of media de
scribed by Salomon appear to offer great promise for future usage. Because they are still
in an exploratory stage, we have not explicitly included the questions they raise as factors
in media selection models. Presumably, such questions would be in addition, but not in
opposition, to those we have applied to the models described here.
Attitudes. Social learning theory (Bandura, 1969) holds that learners' attitudes can be
influenced by having them observe a respected human model making choices and
receiving reinforcement for these decisions. Therefore, attitude learning is usually well
served by the use of media that are capable of displaying the actions of the human beings
who are acting as models.
 Events of Instruction
Modern learning theory of the information-processing variety views the several
different processes of learning as occurring in a temporal sequence from the initial
reception of stimulation to the occurrence of responding (Atkinson & Shiffrin, 1968;
Bower & Hilgard, 1981; Estes, 1978). There are several steps in this sequence, each
involving a different internal process. Gagne (1977) and Gagne and Briggs (1979) call the
different external events supporting each internal process "events of instruc
tion."
Most of the media selection models indicate that the events of instruction consti tuting a
lesson should be planned before selecting the media to be used during the lesson
(Branson et al.; Briggs & Wager; Gagne & Briggs; Gropper; Reiser & Gagne;
Romiszowski; Tosti & Ball). Media are then chosen, based in part on their ability to
present the instructional events. Gagne and Briggs consider this approach to have a
sounder rational basis than one that proceeds with the selection of media before
instructional events have been planned.
Some events of instruction, like that of initially gaining attention, can be delivered
about equally well by any medium, or by an instructor. However, some instructional
events are impossible for some media, while others are inherently easy. Particularly
relevant to the differentiation of media characteristics are the events (1) informing the
learner of the objective; (2) presenting the stimulus material; (3) providing learning
guidance; (4) eliciting performance; (5) providing feedback; (6) assessing performance;
and (7) enhancing retention and transfer.
A number of media selection models provide guidelines to help designers choose
appropriate media for presenting the various events of instruction. Two models

507

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

(Briggs & Wager; Romiszowski) utilize charts indicating the degree to which partic ular
media may be appropriate for presenting different instructional events. Several models
are accompanied by discussions of particular instructional events and the media most
appropriate for presenting them. Suggestions regarding events of instruc tion are
summarized below.
Informing the learner of the objective. This instructional event is intended to provide
learners with a clear indication of what they will be expected to learn, to aid them in
maintaining their orientation to the learning task and its expected outcome. Reiser and
Gagn6 indicate that learning effectiveness is not likely to be influenced by inherent
characteristics of the media employed to inform learners of an objective. Bretz, however,
indicates that visual media, particularly those that can portray motion, may be the best
choice for informing learners about objectives in the psychomotor or intellectual skills
domains. He points out that these media can provide learners with an example of the skill
they are expected to acquire; learners can use such an example as a model for judging
their own performance during the instructional process.
In addition to other effects, the description of an objective may activate a motivational
state in learners, by making clear to them what they will be able to accomplish once they
have learned. Bretz indicates that this instructional function may best be served by those
media capable of providing a demonstration of the usefulness of the knowledge or skill
described in an objective.
Presenting the stimulus material The purpose of this event is simply to present learners
with the stimuli (the "text") of the task to be learned. For example, if learners will be
expected to state a series of facts, those facts must be communicated to them. Or, if the
learners will be required to identify items belonging to a certain class, they must be
presented with examples of items belonging to that class.
Bretz makes the point that it is often best to have this event delivered by a medium
other than a classroom instructor. In this view, the instructor's time often can be better
spent in activities requiring a greater degree of human judgment and personal interaction,
such as may be involved in diagnosing learner problems. Reiser and Gagne suggest that
instructional stimuli should be similar to the stimuli involved in the performance to be
learned. Thus, when a learning task requires the distinguishing of visual features, visual
media should be employed; if auditory features (as in foreign language learning), media
providing audio are essential.
Providing learning guidance. This instructional event is intended to suggest to learners
a mode of mental processing ("encoding") which will lead to effective storage of
whatever is to be learned. This often consists of a "meaningful organiza tion," as well as a
set of cues to aid the learner in remembering. Reiser and Gagne indicate that visual
media are often particularly useful for providing learning guidance. In particular, such
media can influence the semantic encoding process by helping learners establish mental
images, which in turn are capable of enhancing retention (Gagne & White, 1978).
Eliciting the performance. This event is an essential element in most learning theories.
By performing, learners practice what they have learned and thus set the stage for the
occurrence of reinforcement.
Eliciting performance is discussed in many of the media selection models. Several
models indicate that the responses learners will be required to make should be
categorized by type (e.g., overt, covert, motor, verbal, constructed, selected). Accord

508

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

ing to this notion, the designer should attempt to select the media that are best able to
elicit these particular types of responses (Gagne & Briggs; Gropper; Romiszowski; Tosti
& Ball). Tosti and Ball also state that the designer should consider how frequently the
responses will be required.
Gropper presents a rationale for considering response requirements when selecting
media. He advocates a behavioral approach to instruction; practice of appropriate
responses is a key element. He states that it is necessary first to identify the responses
learners should be required to make, then to choose a means of eliciting those responses,
and finally to select the media that can best be used to accomplish this end.
Providing feedback. The event of feedback consists of providing learners with an
indication of the degree to which their elicited performance has been correct. Usually, the
intention is to bring about reinforcement. Although several media selection models
discuss this instructional event, the focus of the discussion varies greatly. One model
suggests that learner characteristics should influence feedback and media selection
decisions (Gropper), another indicates that the media used to provide feedback should
depend on the type of learning outcome expected (Reiser & Gagne), while a third
suggests that almost any medium is adequate for presenting feedback (Bretz).
Gropper proposes that learner characteristics be considered when deciding the media
to be used to present feedback. He indicates that anxious learners may require feedback
that is designed to allay their anxiety. Some media may be better suited than others to
provide this type of feedback. Also, error-prone learners may benefit from media that are
able to incorporate remedial branches.
According to Reiser and Gagne, feedback requirements and the media selected to meet
them will vary, depending on the type of learning outcome the learner is expected to
attain. In the case of motor skills, kinesthetic feedback is needed for learners to improve
in smoothness and timing; this feedback is best provided by allowing the learner to
interact with real equipment or with a realistic simulator. When learners are attempting to
acquire intellectual skills, or other outcomes that include intellectual skill components, it
is desirable to provide them with "precise" corrective feedback. This type of feedback is
best arranged by such interactive media as computers, programmed texts, or interactive
TV, which make exact evaluation of learner responses possible. However, when students
are attempting to acquire verbal information, feedback need not be as precise, and the
media required do not need to have interactive properties.
In contrast, Bretz does not attempt to differentiate between various types of feedback.
He indicates that it is often important to provide learners with knowledge of results, and
that almost any medium is capable of providing learners with this kind of feedback.
Assessing performance. This event, often thought of as "testing," is intended to assess
whether learners have acquired what they were expected to learn. Bretz states that
although print media are often used to assess learner performance, in certain instances
other media may be more appropriate. Printed tests tend to be influenced by a learner's
reading ability as well as his or her mastery of the skills being tested; therefore, the
results of such tests can be misleading, especially if the learner is a poor reader. At least
in certain instances, using audiovisual media to assess the performance of poor readers
may be preferable. In Bretz's view, audiovisual

509

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

media are better able to assess learners' visual skills than are print media; in addition,
they can be used to advantage in presenting lifelike situations for assessing learner
performance.
Enhancing retention and transfer. Some instructional events are designed to increase
the likelihood that learners can recall what they have learned previously and apply it in a
variety of new situations. This purpose is accomplished by furnishing the learner with a
source of cues to be used in later recall. Events with this aim often consist of a series of
review sessions, spaced over a period of months, and involving a variety of examples.
Reiser and Gagne state that visual media can often be used effectively to enhance
retention and transfer. Visual media can help learners to acquire retainable visual images,
which are later used as sources of cues for retrieval of what has been learned.

Practical Factors
All the models indicate that practical factors such as media production costs and
hardware availability must be considered during the media selection process. Most
models propose that the consideration of these practical factors be done following a
review of learning-related factors, such as those described in preceding sections, which
typically results in the identification of a group of appropriate media.
A practical factor most frequently mentioned is the cost of media, particularly the cost
of production. Other factors often mentioned include maintenance costs, hard ware and
software availability, instructor preferences, and production time.

Conclusion
This paper has identified the major features found in a number of typical media
selection models. In some cases, a particular feature is shared by all of the models
reviewed; in other instances, features are unique to one or two. Differences among media
selection models include the physical form used to display them and the selection factors
they employ. These differences influence selection processes and decisions.
Selection of media is likely to be affected by a model's physical form of display. When
a flowchart is employed, the process involves a progressive narrowing of media choices.
However, when a matrix or worksheet is used, media choices are deferred until all
selection criteria are examined. Models that are displayed in flowchart form seem to be
easier to use. This is a particularly important consideration if many of those who will be
using the model have had minimal experience in selecting media.
Decisions about media are also influenced by the selection factors included in a model.
In several of the models reviewed, selection criteria focus on media attributes (i.e., the
physical capabilities of media), such as the ability to present sounds or depict motion.
Other models, however, focus on the characteristics of the intended learners, the
instructional setting, and the learning task. Proper identification of the media attributes
necessary in a given situation depends on consideration of the latter set of factors.
After questions about the learners, setting, and task have been addressed, the necessary
media attributes can be readily defined and the media possessing those capabilities can be
identified. The designer can then proceed to select from among

510

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

MEDIA SELECTION MODELS

those media on the basis of such practical factors as cost and availability, unimpeded by
doubts about instructional effectiveness.

Reference Note
1. Reiser, R. A., & Gagne, R. M. Selecting media for instruction. Book in preparation, 1982.

References
Anderson, R. H. Selecting and developing media for instruction. New York: Van Nostrand
Reinhold, 1976.
Association for Educational Communications and Technology. Educational technology: Defi nition
and glossary of terms. Washington, D.C.: Author, 1977.
Atkinson, R. C, & Shiffrin, R. Human memory: A proposed system and its control processes. In K. W.
Spence & J. T. Spence (Eds.), The psychology of learning and motivation: Advances in research
and theory (Vol. 2). New York: Academic Press, 1968.
Bandura, A. Principles of behavior modification. New York: Holt, Rinehart, & Winston, 1969.
Bower, G. H., & Hilgard, E. R. Theories of learning (5th ed.). Englewood Cliffs, N.J.: Prentice Hall,
1981.
Braby, R. An evaluation often techniques for choosing instructional media (TAEG Report No. 8).
Orlando, Fla.: Training Analysis and Evaluation Group, 1973.
Bracht, G. H. Experimental factors related to aptitude-treatment interactions. Review of Educational
Research, 1970, 40, 627-647.
Branson, R. K., Rayner, G. T., Cox, J. L., Furman, J. P., King, F. J., & Hannum, W. H. Inter service
procedures for instructional systems development (5 vols.) (TRADOC Pam 350- 30). Ft. Monroe,
Va.: U.S. Army Training and Doctrine Command, August 1975.
Bretz, R. The selection of appropriate communication media for instruction: A guide for designers
of Air Force technical training programs. Santa Monica, Calif.: Rand, 1971. Briggs, L. J.
Handbook of procedures for the design of instruction. Pittsburgh, Penn.: American Institutes for
Research, 1970.
Briggs, L. J., & Wager, W. W. The handbook of procedures for the design of instruction (2nd ed.).
Englewood Cliffs, N.J.: Educational Technology, 1981.
Cronbach, L. J., & Snow, R. E. Aptitudes and instructional methods. New York: Irvington, 1977. Dale,
E. A. Audiovisual methods in teaching (3rd ed.). New York: Holt, Rinehart, & Winston, 1969.
Dwyer, F. M. Strategies for improving visual learning: A handbook for the effective selection, design,
and use of visualized materials. State College, Penn.: Learning Services, 1978. Estes, W. K. Handbook
of learning and cognitive processes, Vol. 5: Human information processing. Hillsdale, N.J.: Erlbaum,
1978.
Fitts, P. M., & Posner, M. I. Human performance. Monterey, Calif.: Brooks/Cole, 1967. Gagne, R. M.
The conditions of learning (3rd ed.). New York: Holt, Rinehart, & Winston, 1977. Gagne, R. M., &
Briggs, L. J. Principles of instructional design (2nd ed.). New York: Holt, Rinehart, & Winston, 1979.
Gagne, R. M., & White, R. T. Memory structures and learning outcomes. Review of Educational
Research, 1978, 48, 187-222.
Gropper, G. L. A behavioral perspective on media selection. A V Communication Review, 1976, 24,
157-186.
Heidt, E. U. Instructional media and the individual learner. London: Kogan Page, 1978. Kemp, J.
E. Planning and producing audiovisual materials (4th ed.). New York: Harper & Row, 1980.
Levie, W. H. Models for media selection. NSPI Journal, 1977, 76(7), 4-7. Olson, D. R. Towards a
theory of instructional means. Educational Psychologist, 1976, 12, 14-35.

511

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016

REISER AND GAGNE

Romiszowski, A. J. The selection and use of instructional media. London: Kogan Page, 1974.
Salomon, G. Interaction of media, cognition, and learning. San Francisco: Jossey-Bass, 1979.
Schramm, W. What the research says. In W. Schramm (Ed.), Quality in instructional television.
Honolulu: University Press of Hawaii, 1972.
Schramm, W. Big media, little media. Beverly Hills, Calif: Sage, 1977.
Thorndike, E. L. Education. New York: MacMillan, 1912.
Tosti, D. T., & Ball, J. R. A behavioral approach to instructional design and media selection. A V
Communication Review, 1969, 17, 5-25.
Training Analysis and Evaluation Group. Staff study on cost and training effectiveness of proposed
training systems (TAEG Report No. 1). Orlando, Fla.: Training Analysis and Evaluation
Group, 1972.
AUTHORS
ROBERT A. REISER, Associate Professor, Dept. of Educational Research, Devel
opment, and Foundations, Florida State University, Tallahassee, FL 32306. Spe
cializations: Instructional design; instructional technology; mastery learning.
ROBERT M. GAGNE, Professor, Dept. of Educational Research, Development, and
Foundations, Florida State University, Tallahassee, FL 32306. Specializations: Human
learning; instructional theory.

512

Downloaded from http://rer.aera.net at PENNSYLVANIA STATE UNIV on May 8, 2016