DEVELOPING AN INTEGRATED THEORY OF RECALL OF NEWS STORIES
 
 
by
 
 
 
 
 
 
 
 
Margaret H. DeFleur and Melvin L. DeFleur
 
 
 
 
 
 
 
 
 
 
 
 
 
 
                                        College of Communication
                                        Boston University
                                        640 Commonwealth Avenue
                                        Boston, MA 02215
                                        email Margaret DeFleur: [log in to unmask]
 
 
 
 
 
 
 
 
ABSTRACT
 
 
 
 
 
        This paper has two objectives: First, an axiomatic theory of news recall
is derived from studies of psychological attention sets, principles of
perception, studies of folk-tale recall and theories of memory storage.
Its seven propositions predict the general nature of  patterns of recall
among individuals who attend to and retell a typical spot news story. The
second objective is to check these predictions against data obtained from
a large-scale news recall experiment. The results indicate that the theory
has predictive value.
 
 
 
 
 
 
 
 
 
 
DEVELOPING AN INTEGRATED THEORY OF RECALL OF NEWS STORIES
 
BACKGROUND
 
        Limited Recall: A General Consensus from Prior Studies
 
        Assumptions of Audience Deficiency vs Efficiency
 
 
OBJECTIVE ONE: DEVELOPMENT OF A FORMAL THEORY
 
        Attention: Anticipatory Sets Prior to Exposure to News Stories
 
                Level of Personal Relevance
                Standardized Scripts and Narration Schemata
 
        Perception: Selectivity and Patterning
 
                Selecting and Winnowing Details
                Gestalt Principles of Perception
 
        Memory: Contrasting Approaches in Psychological Traditions
 
                The Ebbinghaus Tradition
                The Bartlett Tradition
                Contemporary Theories
                Schema Construction and Encoding in Memory
 
        Retelling: An Axiomatic Theory of Recall of Spot News Stories
 
 
OBJECTIVE TWO: COMPARING THE THEORY TO DATA FROM THE EXPERIMENT
 
        Methodology
 
        Findings
 
                Limited Recall of Details
                Selective Recall of Details
                Evidence of Assimilation
                Modal Configurations of Story Recall
 
DISCUSSION AND CONCLUSIONS
 
        REFERENCES
 DEVELOPING AN INTEGRATED THEORY OF RECALL OF NEWS STORIES
 
        This paper has two objectives: The first is to inductively derive a formal
theory of news recall from several major psychological traditions. The second is
to check the theory's predictions against data obtained in a large-scale
experiment. The intellectual foundation for the development of the theory comes
from a broad range of classic and contemporary studies of psychological
attention sets, Gestalt principles of perception, Bartlett's studies of social
factors in folk-tale recall and contemporary theories of memory storage. The
seven-proposition axiomatic theory that is formulated predicts the general
nature of the memory limitations and the patterns of  recall that will be
obtained when individuals attend to, remember, and retell a typical spot news
story.
        The second major objective is to check empirical data obtained from a
large-scale news recall experiment against the predictions made by the theory.
To do this, each of  480 subjects was exposed, in a controlled individual
session, to one of three news stories that was presented by one of four media
(newspaper, television, radio and personal computer). Each of the three stories
was of similar length and each focused on an event that is commonly reported in
local news--a house fire with no loss of life, a car accident in which a person
drowned, and a murder of members of a local family. Immediately after exposure
in the sessions, subjects were asked to recount as many details of the story as
they could. Their versions of the story were recorded on tape for later review.
The analysis of the data, and the conclusions reached, focus on three issues
concerning the ways in which these stories were recalled by members of a
well-motivated, attentive and intelligent audience:  To what degree were the
facts recalled accurately? What kinds of details were retained or forgotten?
And, finally, into what patterns were the details organized for recall and
recovery? As will be shown, the findings offer support for the theory.
BACKGROUND
        Within the growing body of evidence on recall of news stories several lines of
inquiry can be identified:  Many of the studies that have been published over
the last half century have been concerned with comparisons between media. That
is, they contrast the amount of information retained by subjects exposed to,
say, newspapers, as compared to radio or television.[1] Another focus has been
to identify demographic characteristics among an audience that make a
difference. Thus, quantitative patterns of recollection of news stories have
been compared among people of different ages, genders, educational attainment,
income and ethnic backgrounds.[2]  A number of recent studies have attempted to
relate news recall to specific concepts and issues drawn from psychological
studies of memory.[3] Still another branch of news recall studies have assessed
the role of production factors, such as story formats--inverted pyramid vs
narrative used by newspapers, talking heads vs voice-over-tape in television,
and so on--in determining which stories will be remembered better.[4]
Limited Recall: A General Consensus from Prior Studies
        What all of these communication research traditions have in common is a
conclusion that people remember relatively little after exposure to news
stories. Some differences have been found from one medium to the next--with
recall from print media being somewhat better than from broadcast news. However,
the general situation is that people just do not recall very much, even when
tested immediately after exposure to a news story. For example, in a survey
study, adult viewers of TV news could recall the topic of slightly more than one
story from an average of 20 typically presented in a broadcast. In another, more
than half of those interviewed could not recall the content of even a single
story presented in a newscast containing 19 items.[5] Even when people are
contacted ahead of time and agree to pay close attention, recall is still very
low.[6]  Thus, a substantial literature indicates that recall of news of any
type is very limited, regardless of media, demographics or production variables.
Assumptions of Audience Deficiency
        For journalists, this consensus can be frustrating. It raises the uncomfortable
question as to whether they are doing their job right. Some assign blame to
journalists on the grounds that the limited time and space they devote to any
story precludes effective presentation of details with a consequent loss of
audience recall.[7]  However, threading through much of the literature on the
limited degree to which people recall news is an implicit assumption that there
is something wrong with the audience--some shortcoming that they ought to try to
overcome. For many researchers, the gap between what people claim and what they
do poses a clear inconsistency:
     Although people say that [news is significant] in their lives,
     behavioral research concerning the use of different mass media and
     also cognitive research concerning information uptake from the media
     togther provide evidence that often contradicts that personal claim.[8]
 
Various explanations have been advanced to try and explain this inconsistency.
These include assuming that audiences have limited intellectual levels, that
they fail to concentrate adequately or that it is due to the preponderance of
older people attending to the news with memories failing with age.
        Rather than dismiss this inconsistency as a deficiency of the audience, or due
to poor encoding on the part of journalists, it may be more appropriate to view
it as an indicator of efficiency on the part of people attending to the flow of
news. That is, it is entirely likely that people attend to news employing the
same efficient habits of perception and memory that they use when focusing their
attention on any complex aspect of their social environment. For example, just
as people use stereotypes for responding "efficiently" to individual
representatives of various social categories (minorities, women, the aged,
etc.), they may deliberately exclude from the task of remembering any news
content that has little meaning in their lives. In any case, limited remembering
is a factor that must be addressed in any comprehensive theory of news recall.
 
OBJECTIVE ONE: DEVELOPMENT OF A FORMAL THEORY
        What is needed is a comprehensive theory that addresses the basis of such
limited recall by integrating as much as possible of what is known on all phases
of the experience of a member of an audience who pays attention to a news story
and then commits it to memory for later recall. Such an integration requires
that concepts, theories and conclusions be drawn from a broad spectrum of
sources. To illustrate, four such phases that such a theory must encompass can
be identified: The first is a pre-perception phase, in which the individual's
various psychological and cultural "sets" determine the amount of attention that
will be paid to a story, and pre-indicate to what degree it will be seen as
relevant to his or her life-concerns. Obviously, these sets are an important
factor in the individual's motivation to learn and accurately recall the content
of the story. A second is the perceptual phase, in which the essential ideas in
the story being presented by a medium are selected and understood as a
meaningful pattern of concepts and relationships that the person recognizes and
can readily interpret. In the third, the encoding and retention phase, the
person organizes into memory what seem to be the main ideas of the story,
winnowing out what appear to be unnecessary or extraneous details, and storing
the resulting coherent configuration in working memory.  Finally, in a recall
phase, the stored configuration, which for a number of reasons may differ from
that originally stored, is remembered, or retold if required.
        To accomplish such an integration it is necessary to draw upon a complex and
extensive literature in both psychology and communication research that has
accumulated over many decades. This presents a Jovian task. Equally demanding is
the need to formulate such a theory according to definitions advocated in
virtually all books and treatises on communication research (and general social
science) methodology. Such discussions typically define in abstract terms what
are essentially deductive or axiomatic theories set forth in systematic
propositions. For example, Reinard explains that a theory can be defined as a
body of interrelated principles that explain or predict (whatever is under
study). He notes that such theories have three characteristics, including an
abstract calculus (a logical method by which conclusions are drawn from
premises), theoretical constructs (that define whatever the theory is all
about), and rules of correspondence (that relate the theory to phenomena that
can be observed in the real world).[9] Unfortunately, specific examples of
actual theories formulated according to such an axiomatic model are difficult to
find. Indeed, no systematic theory could be identified in any literature that
attempted to integrate into a single set of concepts and deductive propositions
a calculus, theoretical constructs or rules of correspondence related to
remembering news. Indeed, no theory of any kind could be identified that
integrated anticipatory sets, perceptual organization and patterns of memory
storage and recall. Nevertheless, that specifically is the first objective of
the present paper.
        Fortunately, in a general sense (not specifically related to news), all of
these issues have been under theoretical development by psychologists for a very
long a time--in some cases for more than a century. These extensive writings
offer rich sources for attempting to develop an integrated theory of news
recall. Therefore, the present section will review briefly several classic
traditions of conceptualization and research in psychology that appear to be
relevant as foundations for developing a series of formal propositions that can
explain how people remember a news story. Specifically, the discussion below
will address (1) anticipatory sets that influence an individual's attention to a
story,  (2) how he or she goes about perceiving the information provided, (3)
how the person organizes it into a meaningful pattern consistent with his or her
personal and cultural expectations, and (4) how it is committed to memory in a
form that the person can recall, often with a significant but predictable level
of inaccuracy.
        Therefore, the discussions that follow present a brief review of relevant
theoretical traditions in the psychological study of psychological sets, the
process of perception, the organization of memory encoding and the nature of
recall. Next, based on that review, a set of formal propositions are developed
that integrate selected psychological traditions into an axiomatic  theory of
news recall. The theory deals specifically with relatively simple spot news
stories. Finally, the second objective of the paper is to present data from a
relevant experiment to determine whether its findings are consistent with the
theory.
Attention: Anticipatory Sets Prior to Exposure to a Story
        The concept of a set comes from the earliest days of empirical psychological
research.[10]  The most celebrated studies of psychological sets were those of
Wilhelm Wundt and his students. They were carried out in the first psychological
laboratory, established in Leipzig in 1879. These were simple experiments
showing that when subjects expected an event to happen, they were able to react
to it much faster than when it was not expected. In other words, they were
psychologically "set" to respond in a particular way. Today, the concept of set
has a similar meaning. A person approaching an experience--such as reading,
hearing or viewing a news story-- brings to the task a number of learned
expectations or psychological sets.
        Level of Personal Relevance. One well-understood category of sets pertains to
the person's cognitive make-up. The individual will have an existing structure
of preferences, interests and other predispositions that will play important
parts in determining how much attention is paid to a particular news story about
a specific topic. This is the basis of much selectivity in attention. It plays
an obvious role in the high levels of interest in sports stories on the part of
many males, attention to financial stories by business persons, preoccupation
with fashion news shown by others, and so on.
        Standardized Scripts, Frames and Narrative Schemata. Less obvious are
culturally- defined sets that bring people to anticipate what they will
encounter in an often-experienced situation. Important here are what
psychologists and others call scripts, a concept that has been used in research
on television.[11] A script is a general set of anticipations about what will be
encountered in a situation that is familiar. Examples are going to a restaurant,
checking in at the airport, attending a cocktail party, and so on. The learned
script is not only a product of having experienced such a situation before, but
it also allows the person to anticipate what activities will be required, and
who will do what to whom with what consequences. An almost identical idea is a
frame. A more recent concept, this refers to an anticipated structure of a
stereotyped situation.[12] Scripts and frames play a part in attending to a news
story--a familiar experience to most people. It comes as no surprise to them,
therefore, that the story will present information about something that happened
to somebody, with a number of consequences.
        A related idea is the narration schema (sometimes called a "narrative
script").[13] We learn from childhood that all "stories" follow a certain
pattern and we come to expect it. A setting is introduced, along with a set of
characters. Things happen to the characters or actors as events unfold. Those
events have consequences--sometimes positive, sometimes negative--but some sort
of resolution is reached by the end of the story. Spot news stories are often
structured in this manner: The house catches on fire, the firemen arrive, the
spokesperson is interviewed, the fire is extinguished, the plight of the
residents is discussed  and the cause of the fire is identified.
        Whether conceived of as scripts, frames or narrative schemata, it is clear that
people encountering any story, including one presenting news, will be "set" by
past experience to anticipate a somewhat stereotyped structure in what they will
encounter. Thus, sets play a significant part in focusing attention and
influencing the organization of perception.
Perception: Selectivity and Patterning
        The study of human perception, an important stage in the experience of a member
of an audience attending to and understanding a news story, has been studied
intensively by psychologists since the beginnings of the twentieth century. A
massive literature, based on an impressive accumulation of experiments and other
types of empirical research, has revealed significant insights into the way in
which physical and verbal stimuli are selectively apprehended by the senses and
transformed into a configuration of meanings that can be stored in memory.[14]
Of particular interest in the present context is attention to and perception of
verbal material, such as a narrative, a story or a news report, and the way it
is organized into a meaningful configuration that can be stored in memory.
Specifically, some of the principles of Gestalt psychology can be of help in
understanding the perception of news. While Gestalt principles are most
frequently associated with the perception of physical stimuli, they can be
applied to the ways in which members of an audience interpret verbal accounts of
human social actions and events as well.[15]
        Selecting and Winnowing Details. As is well understood, the process of
perception is highly selective. Some details will stand out in the perceiver's
focus of attention because, as noted earlier,  they are personally relevant to
his or her interests. Others will be perceived very clearly because they are
central or unusual. Still others, interpreted as non-essential in understanding
the events being observed, are likely to be ignored.  In any case, it is the
influence of selectivity in what audience members apprehend with their senses in
a stimulus situation that becomes the over-riding consideration when people
construct patterned interpretations of the meaning of events that they witness.
That initial perceptual pattern will be continue to be the organizing factor as
they assemble those meanings into a form that can be stored in their memory and
recalled later. It is precisely this perspective, drawn directly from basic
Gestalt psychology, that provides the theoretical source for the perceptual
stage in the theory of recall of news being developed.
        Gestalt Principles of Perception. After anticipatory sets have influenced the
level of attention given a news story, the next stage of cognitive processing
takes place when a member of an audience undertakes to identify the meanings of
the words selected to make up the story. No formulation stands out more clearly
in the annals of psychology than the principles of perception advanced by the
Gestalt School. Gestalt psychology began in the early 1920s in Germany in Carl
Stumpf's Berlin Psychological Laboratory. Its founders were Max Wertheimer, Karl
Koffka, Wolfgang Kohler--names very familiar to the student of classic theories
of psychology. Each left Germany to migrate to the United States where they
continued to develop principles of perception currently identified with the
school.[16]  Systematic statements of the Gestalt theories were set forth in
now-classic books by Wolfgang Kohler and by Kurt Koffka.[17]
        Initially, Gestalt theory represented a rebellion against the investigational
procedures and assumptions prevailing in American psychology during the late
19th and first two decades of the 20th century. Structuralism and associationism
were the contemporary paradigms within which psychologists sought to identify
the fundamental nature of human experience. Psychological investigations of the
time were aimed at identifying the elements of consciousness that could explain
human behavior. The search for these elements was conducted by using the method
of introspection. Gestalt theorists rejected this approach and insisted that the
psychological behavior of human beings should be based on empirical observation
of everyday experience. Moreover, they conducted numerous experiments that
showed the importance of patterns in perception.
     "We perceive, they held, according to patterns, and the process
     that patterns our sensations takes precedence over the individual
     elements. Thus, a melody will be perceived as one thing whether its
     elements are in the key of C or the key of G--two quite different
     sets of tones.[18]
 
        The insistence of the Gestalt psychologists on empirical observation--as
opposed to the self-examination approach of the introspectionists--became a
major foundation for conducting social psychological experiments in laboratory
settings. Gestalt theorists such as Kurt Lewin led the way in developing an
experimental approach and field theory perspectives in the study of social
processes and phenomena.[19]  By the end of the 1930s, a number major social
psychological concepts, such as leadership, source credibility in communication,
social conformity and attitude change were being studied in controlled
situations with objective methods of observation.
        An important feature of the Gestalt school for present purposes is that it was
founded on the principle that processes like human perception and learning
cannot be explained by identifying the elementary sensations and associations
involved in immediate experience. That is, people do not apprehend a stimulus
situation as an unorganized set of independent elements (sounds, colors, smells
or tactile sensations) apprehended with their senses. They perceive the meaning
of what they observe in terms of patterns of inter-related elements of
experience. In observing an auto accident, for example, eye witnesses do indeed
hear sounds, see colors, and so on, but their overriding interpretation is that
of an organized configuration of experiences--a sequence of events in a
recognizable time order, leading to a set of outcomes and consequences logically
related to prior events.  If asked to recall what they saw and heard, including
a news story, it will be this pattern around which they will organize what they
will be able to remember.
Memory: Contrasting Approaches in Psychological Traditions
        Investigations into the functioning of human memory, using a scientific
approach, began more than a century ago.[20]  Patterns of recall of different
types of messages presented by various media have been under study by
psychologists for many decades. Kellerman has reviewed 233 books and articles
published between the 1950s and 1985 on this general topic. She concluded that
little consensus had been reached regarding the influence of and type of media
on memory.[21]  In 1997, psychologists Stein, Ornstein, Tversky and Brainard,
citing more than 1,250 published works, concluded that memory for such material
"is both accurate and inaccurate, depending on the conditions under which
information is encoded and retrieved."[22] In short, extensive investigations of
memory for prose, folk tales, or everyday events, by massive numbers of
psychological investigators have not produced consensus regarding theories that
can explain how people recall news stories. Theories of memory functioning and
storage abound, with almost as many models and explanations as there are
investigators. It is a subject of intense investigation and lively debate.[23]
        Nevertheless, memory is a key factor. Regardless of what or how people
initially perceive a physical stimulus, an event or a story, its meanings vanish
if they are not committed to memory. The experience must be organized in such a
way that it can be committed to the person's "mental filing system" in a form
that can be recovered if needed. Thus, cognitive organization of a perceived
event for the purpose of encoding it in memory is the third major phase that
takes place in the experience of recalling a news story. To be able to
accomplish this phase of cognitive processing of news, the story's elements of
meanings must be interlinked and retained in such a way that they seem logical
and consistent with the expectations of the perceiver. Theories of memory
organization and storage, therefore, may provide explanations of how this phase
is accomplished.
        Attempts to conceptualize human memory go back at least as far as Plato, who
used a spatial metaphor based on a cage full of different types of birds to
describe its structure. However, the scientific study of human memory by
psychologists did not begin until late in the nineteenth century.[24] Since that
time, it has been characterized by two major but very different methodological
traditions. The most prominent stems from the classic experiment of Hermann
Ebbinghaus. Published in 1885, it was from the data obtained in this
investigation that he produced his well-known "curve of forgetting."[25] The
second tradition, much less used in the past but now receiving much more
attention, began with the 1932 studies of the recall of folk-tales by Frederick
C. Bartlett. It was in these studies that the concept of "schema" was first
used. Also from Bartlett's studies came the initial ideas for the concepts
leveling, sharpening and assimilation inherent to Allport and Postman's
embedding theory of recall in rumor transmission.[26]
        The Ebbinghaus Tradition. The experimental approach used by Ebbinghaus
dominated the investigation of human memory by psychologists for well over one
hundred years. Ebbinghaus used himself as a subject, memorizing stimulus
material and testing himself after various periods of time. He was  able to show
a characteristic pattern of forgetting--rapid at first and slowing over
time--that describes how human beings lose the ability to recall stimulus
material that they have memorized. Even today, such investigations often have
two major features that were central to Ebbinghaus methodology.  That is, they
are almost always conducted using tightly controlled experimental laboratory
procedures, and they make use of stimulus material deliberately designed to
limit cultural influences. To illustrate, Ebbinghaus stated that he had hit upon
a way "of penetrating more deeply into memory processes:"
     Out of the simple consonants of the alphabet and our eleven
     vowels and dipthongs, all possible syllables of a sort were
     constructed, a vowel sound being placed between two consonants. These
     syllables, about 2300 in number, were mixed together and then drawn
     out by chance and used to construct series of different lengths,
     several of which each time formed the material for the test.[27]
 
Essentially, Ebbinghaus set about to memorize various combinations of  "nonsense
syllables," and to test his recall of those combinations after various periods
of time had elapsed. He relied on repetition and rote memory to acquire the
stimulus content initially. And even though he had a sample of only one
(himself), his initial study used a complex and tightly controlled design in
which numerous conditions and variations were observed.
        The Reaction Against Ebbinghaus. Today, more than a century later, a great body
of publications--literally hundreds of books and thousands of journal articles
based in large part on the Ebbinghaus experimental approach--have been published
on one aspect of human memory or another.[28]  Such memory experiments usually
meet high standards of experimental and measurement precision. Multiple subjects
are used and studied within statistically sophisticated research designs.
However, the stimulus material that subjects are asked to memorize still often
consists of culturally neutral content--something like a series of numbers, word
fragments, abstract shapes, color panels, simple word-pair associations, or some
other content that is not embedded within a meaningful context. There are many
exceptions to this generalization, but overall, the influence of the Ebbinghaus
approach to the psychological study of human memory is still very evident in
these types of experiments.
        In recent times there has been a reaction against the Ebbinghaus tradition.
There is little doubt that the thousands of studies of memory, its correlates
and conditions, that the approach has yielded have enriched our understanding of
this complex process of cognitive functioning. At the same time, this huge body
of theory and findings does not appear to have provided explanations that have
immediate and direct relevance for much of what people attend to, learn and then
recall in their daily lives. Because of this limitation, at least some
psychologists have advocated abandoning the Ebbinghaus tradition entirely. In
1978, for example, Ulric Neisser advanced what was than a new and radical
position:
     . . . he dismissed the work of the past 100 years as largely
     worthless. . .Neisser believes that the important questions about
     memory are those that arise out of everyday experience. We ought, he
     claimed, to be finding out how memory works in the natural context of
     daily life at school, in the home or at work. . . .The traditional
     laboratory experiments, according to Neisser, have failed to study
     all the most interesting and significant problems and have shed no
     light on them.[29]
 
        Because of growing skepticism about the generalizability of laboratory studies
based on culturally meaningless material, there is today a strong movement
within psychology to try to  study memory in everyday situations. Studies have
appeared in recent years on topics such as eyewitness testimony, recall of
emotional events, such as sexual abuse, salient medical experiences and a
variety of other past complex events.[30]  What has become more relevant to
studies of memory for narrative content, such as that in a typical news story,
is the early work of Bartlett.
        The Bartlett Tradition. In the early 1930s, Sir Frederick C. Bartlett conducted
memory experiments making use of  a very different experimental approach. He
asked British experimental subjects to recall an organized story that they read
through twice. Thus, he avoided the use of meaningless material.[31] The story
was "The War of the Ghosts," drawn from a folk-tale of the Kwakiutl--a Native
American group in the Pacific Northwest--and it described a seal-hunting episode
that turned into a battle. The story sometimes posed difficulties for subjects
trying to recall the account accurately because the story contained cultural
concepts and activities with which they were not familiar--seal-hunting, ghosts
and travel by canoe. However, that very difficulty led to new insights. It
allowed the subjects to "modernize" their versions of the stories by using
culturally familiar language that was incorporated into the story, sometimes
altering its meaning.
        In overview, what Bartlett found was that his subjects were able to recall
personally constructed general impressions of what they had heard. Their
versions were shorter and were organized around what appeared to the subjects as
salient details. Their accounts often contained concepts and ways of expressing
ideas from their own culture assimilated into their descriptions. Thus, recall
of this type of material was based on what Bartlett referred to as personal
schemas, woven into an account that "made sense" within the experience and
culture of  the person asked to repeat the story.
        The "method of serial transmission," used in other studies by Bartlett,
provided the basis for the well-known studies of rumors in wartime by Allport
and Postman.[32] Those researchers developed an embedding theory to describe
changes in an experimental story (rumor) as it was told, remembered and retold
in stages from one person to the next. As did Bartlett, they found a
characteristic pattern in what their subjects could recall at each stage: They
shortened the story--which was referred to as leveling; they organized the story
around salient details--a process they termed sharpening; and the subjects
brought in ideas and expressions from their own culture that were not in the
original--which was what Bartlet had called assimilation.
        The Bartlett tradition had little influence on the psychological investigation
of memory until relatively recent times. Today, few psychologists use material
foreign to the culture of experimental subjects, but his procedures of asking
them to recall meaningful content (as opposed to numbers, nonsense syllables,
etc.) and the methodology of analyzing their personally reconstructed accounts
are being more actively used in the study of memory as it functions in everyday
life.  Obviously, this approach--rather than the Ebbinhaus procedures--is more
relevant to the study of how and to what degree people can remember news
stories.
        Contemporary Theories. In the brief space of a journal article, it is
impossible to summarize the major theoretical conceptualizations of human memory
functioning that are currently being debated and investigated by psychologists
who follow either of these traditions. In the last two decades there has been a
virtual explosion of interest and research on memory and memory functions.
Literally thousands of experiments and other investigations of a long list of
issues and topics have been published. The present discussion can make no
pretense of reviewing all possible alternative contemporary branches and
theories of memory functioning. In 1991, an International Conference on Memory
was held at Lancaster, England. The 179 papers presented at that conference
provided an overall view of current research on this important topic. Most of
the central traditions within memory research were represented. Selected reports
from that conference provide insights into the directions that such research and
theory development are now taking.[33]
        To provide an idea of the scope of current research on memory, investigations
have recently been conducted into the following, which represent only a partial
list: autobiographical memory; conceptual memory, contextual memory,
dual-process influences on memory, episodic memory, eidetic memory, emotional
memory, everyday memory, false memories, flashbulb memory, generic recollective
memory, hierarchically organized memory, imagery in memory, linguistic memory,
long-term vs short-term memory; nonverbal memory, narrative recall, permanent
memory, procedural memory, recollective memory, repressed memories,
retrospective memory, schemas in memory, scripts in memory, sensory memory,
sensory-motor memory (as in driving), spatial memory, temporal memory, visual
memory and working memory,
        Among these many alternatives, an important theoretical direction that may be
relevant to the study of news recall focuses on working memory. As Eysenk
describes it, "The working memory system consists of a modality-free central
executive; an articulatory loop; and a visio-spatial scratch pad."[34] What this
means in simpler terms is that we retain experience in our consciousness
(central executive) for at least a limited period; we are able to articulate in
words what we have stored there; and both visual and spatial elements play a
part in what we recall. Those features of memory do appear relevant and
applicable to the recall of news stories, and they will be incorporated into the
news recall theory under development. However, as will also be explained,
learning, storing and recalling news stories appears to have additional elements
that do not appear to be present in current theories of working memory.
        Schema Construction and Encoding in Memory. Schema memory structures provide a
contemporary approach to understanding how human beings organize and encode into
memory news stories to which they attend. As was indicated, the term came
originally from the work of Bartlett, but in recent decades it has been
developed in sophisticated ways by a body of researchers and theorists to
indicate how people organize and store experiences. A schema is a kind of
"mental organization" used to remember ideas and events. The term can be defined
very simply as a personally organized structure of perceived and remembered
experience. Schemata provide the mental structure in which we commit what we
perceive to memory. Included in that process is media content--such as a story
in the evening news, or one read in the morning paper. Since each person has a
unique cognitive structure (of needs, beliefs, attitudes, values and so on)
developed from a lifetime of prior experience, it is not surprising that each
will encode a personally unique pattern of the details and relationships
perceived in a news story. That encoded pattern will be stored in working memory
as a schema. Thus, one person's schema for a particular news story may or may
not be consistent with that of another. As Harris puts it:
     The way we comprehend a program we watch on TV is through a
     constant interaction of the content of the program and the knowledge
     already in our minds. The mind thinks in response to what we see and
     those thoughts become an important part of the constructive process
     of comprehension.[35]
 
Noted earlier was the concept of a narration schema.[36]  As was explained,
narration schemata for stories include identification of one or more actors, a
set of events that those actors have experienced, and a set of consequences
brought about by those events. These provide convenient organizing categories
that any reader, listener or viewer can use in encoding and storing the details
of the Gestalt of meanings developed during the process of perception.
 
Retelling: An Axiomatic Theory of Recall of Spot News Stories
        From all of the theoretical sources discussed in previous sections, a
three-stage axiomatic theory of recall for spot news stories can be inductively
formulated. Consisting of six basic assumptions or postulates, and a predictive
proposition,  it incorporates selected fundamental concepts and principles from
psychological research and theory regarding attention, perception, memory
storage and recall of narrative or story-like content. The theory is stated in
such a way that its seventh or predictive proposition follows as a logical
consequence of the prior postulates.
        As noted earlier, the axiomatic format is the theory structure that is widely
used in almost all scientific disciplines. While not yet used routinely in
communication research, in books on methods it has now become widely advocated a
model for theory-development.[37] As defined by philosopher of science Ernest
Nagel, an axiomatic theory structure consists of a calculus--some  logical
system of reasoning that weaves the propositions together. The propositions
include a set of concepts and are stated in as a set of postulates (that if
taken as true) predict the behavior of the phenomena at issue in such a way that
it can be checked against reality.[38]
        The theory as set forth in Figure 1 below attempts to meet those criteria. In
some physical sciences the "abstract calculus" defining the basic theory may
consist of mathematical derivations. However, that need not be the case,
particularly in the social sciences (including communication). As Abraham Kaplan
points out, such a calculus can consist of one "whose sentences contain names
for the 'theoretical entities' being dealt with and provide horizontal
specifications for the meanings of these names."[39] Thus, the calculus in such
a theoretical structure is one of verbal intuitive logic. It indicates that if
the basic assumptions (postulates) of the theory can be regarded as correct,
then the final or derived proposition predicts a situation that should exist
Figure 1
A THEORY OF SPOT NEWS STORY RECALL
Stage One: Anticipatory Sets Prior to Perception
   1. A reader, listener or viewer brings to any story, including a news
   report, a number of
                personal anticipatory sets learned from prior experience, including a number
of culturally-
                defined scripts or narration schemata that provide the expected structure of
such an                 account.
 
   2. Those personal sets and culturally-defined expectations influence
   both the level of attention  given to a story and its anticipated
   structure, which includes identification of a setting
                in which one or more actors experience a set of events that bring about a set
of
                consequences.
 
   Stage Two: The Process of Perceptual Organization
 
   3. Perception is a process of organizing a unique Gestalt of meanings
   for a situation, such as
                a news story, that has the attention of the individual--a comprehensible
pattern of                      interpretations consistent with personal sets and cultural
expectations.
 
   4. The process of perceptual organization of a news story is highly
   selective, focusing on
                those details to include in the Gestalt that seem logically needed for
consistency,                    personally salient, or otherwise central, colorful or dramatic;
others that appear                      unimportant will be winnowed out.
 
Stage Three: Patterns of Storage and Recall
 
   5. The resulting Gestalt of the story will be prepared for storage in
   memory as a
                personally organized schema of linked images, concepts, relationships and
other elements
                of meaning that the person has constructed during the process of perceptual
organization.
 
   6. The stored schema will be encoded into working memory to include a
   number of linked
                details perceived to be central to the story's setting, principal actors,
major events and                consequences, others that appear to be dramatic or colorful
and still others that the               individual may incorporate from other schemata.
 
        7. Therefore: When called upon to recall the story, it will be restated in what
the person              believes is a logical configuration of central and salient,
(possibly plus some dramatic)
                details providing a leveled, sharpened and assimilated version of its content.
 
 in reality--which in turn provides a guide for empirical observation and
research. The formal theory as stated in Figure 1 follows this pattern. If its
first six postulated assumptions are taken "as if" true, and if its logic
(calculus) is consistent, then recall patterns of news stories should show the
characteristics predicted in the derived seventh ("therefore") proposition. To
determine if that prediction has merit, relevant empirical data must be
examined.
 
OBJECTIVE TWO: COMPARING THE THEORY TO DATA FROM THE EXPERIMENT
                The research report that follows addresses the prediction stated in
proposition seven above. The report consists of a systematic examination of the
verbal accounts of 480 subjects who were asked to read, listen to or view and
then retell one of three spot news stories. The question is, were those subjects
able to recall a logical sequence of salient and other details, and was there
evidence of leveling, sharpening and assimilation in their versions of the
content of these news reports? If that is indeed the case, evidence is provided
in support of the theory. If such evidence in not found, then the theory cannot
be regarded as a valid explanation of the manner in which members of this
audience perceived, organized, remembered and recalled these particular news
stories. If reasonably compelling evidence is found, it does not "prove" the
theory in the sense that statistical probability considerations lead an
investigator to "accept" a more sharply focused and specific null hypothesis.
Confirming evidence from one study, at best, indicates that the theory's
prediction is "consistent with" the data at hand, which implies at least some
"support." Further research on other stories with other types of audiences would
add to that support. Only a substantial accumulation of evidence indicating that
the theory's predictions support its predictions will provide a high level of
confidence in its validity.
  Methodology
                A large-scale experiment was conducted in which one of three local news
stories was individually presented to 480 subjects.[40] In addition, four media
were used for the presentations: newspaper, computer, television, and radio. The
3 x 4 factorial design held constant a number of audience and situational
variables that can influence what, and  how much, people remember from exposure
to a typical news story.  See Figure 2 for the content of the three news stories
used.
                Each subject in the experiment was presented with only one story in either the
newspaper, computer, television or radio format. Each was told that the purpose
of the experiment was to see how much, and what type of information, he or she
would remember after carefully reading, viewing or listening to a news story one
time. Subjects were told that, immediately after exposure,  they would be asked
to repeat all of the details that they could remember into a tape recorder.
        The subjects who participated in the experiments were students taking
introductory courses in media studies or mass communication at one of three
universities in the U.S. These included a very large Midwestern state
university, a middle-sized northeastern private university and a small private
institution with religious affiliation. Obviously, this selection does not
result in a sample that is representative of American college students. However,
using three different schools did avoid the limitation of reporting results from
only a single type of institution.
        The 480 subjects were randomly assigned to each of the twelve experimental
conditions  (40 subjects in each cell). Exposure took place in a quiet room with
no one present except an experimenter. The procedures and controls of the
experiment were designed to limit the influence of extraneous individual
variation, to reduce attention distractions within the environment, and to
enhance motivation to recall the details of the story.
                Several factors were considered in the development of the three news stories
used in the experiment. These included prior knowledge, proximity, and
importance. Since prior knowledge is often cited as a factor influencing recall,
this was controlled by using accounts with no history. Proximity is an accepted
"news value" that influences interest and recall. This was controlled by using
stories that were identified as originating in the local community. Finally,
story importance is a factor that has been shown to create audience interest and
result in higher levels of recall. This was controlled by using stories with
three different levels of newsworthiness. One was a story of a residential fire
caused by arson that resulted in property damage. The second story described a
car accident, in which the car plunged into a river and caused the driver to
drown. The third, at the highest level of newsworthiness, concerned an argument
in which a man shot and killed his brother and other members of the brother's
family. The news stories were designed to be very similar in the number of
words,  the types of details, and all other features (See Figure 2). These
stories were reviewed by a former editor of a major newspaper, a former
broadcast news director and a media scholar. All agreed  with the newsworthiness
ranking of the three stories.
        In the newspaper version, subjects were given the story in a typical column
format on one printed page. They were asked to read each sentence at their
normal pace, but only once. To ensure compliance, each subject was observed by
the experimenter during the presentation. Similar requirements were imposed for
the computer version in which the story was contained on one screen and required
no manipulation by the subjects. For the stories presented by television,
subjects were seated in front of the screen and the taped story was played only
one time. Each story was presented in a "word-story" or "talking head" format in
which a professional anchor, unknown to the subjects, presented a story with no
accompanying video or graphic display.
Figure 2
 
SPOT NEWS STORIES USED IN THE EXPERIMENT
 
              Fire Story                            Drowning Story                      Murder Story
 
 
        A city employee dismissed from his job last night was arrested in connection
with a fire that destroyed his boss' $120,000 home at 25 Oak St. this morning.
        Warren T. James, 33, 35 Elysian Dr., was arrested at his home Tuesday evening
after his former boss, Terry Arlington, director of the Parks and Recreation
Department, told police she received three threatening telephone calls Monday
from James, said police.
        James, who had been fired for poor work performance and frequent ab-sences, had
been Arlington's assistant for two years. He was reprimanded twice prior to his
dismissal, said Arlington.
      Arlington, who lives alone, was at work when the fire broke out. Neighbors
called the fire department when they saw flames coming out of an attic window at
Arlington's 25 Oak St. home, said Fire Chief Ed Wilson.
   Twelve firefighters and three fire trucks responded to the fire which
apparently started in the kitchen at about 10 a.m.
    Firefighters battled the blaze for about two hours. The fire, which burned
for almost half an hour before firefighters arrived, destroyed most of the home,
including Arlington's collection of 19th century children's books, Wilson said.
He declared the home a total loss.
  Two empty gasoline cans in Arlington's dining room led fire-fighters to
suspect arson, said Wilson.
        A neighbor told police she recalled seeing a blond, slender man, fitting James'
description, leaving Arlington's house shortly before the fire broke out.
   James is currently at the county mental health center undergoing psychiatric
testing, police said.
 
 
 
 
 
        A woman whose car plunged into the Ontario River Thursday afternoon may have
suffered a heart attack shortly before her car went off the road.
        Norma Bernard, 65, 12 Adams St., was pulled from her car by a police rescue
squad and then rushed to St. Joseph's Hospital where she was pronounced dead on
arrival, said Emergency Services Director, Dr. Jill Griffin.
        Police said they at first thought Bernard, who had been in the partially
submerged car for about 45 minutes, might have been alive but suffering from
hypothermia.
        The tracks across two lanes of River Street initially led police to believe
Bernard's car may have skidded off of the twisting, rain-slicked stretch of
River Street located two miles east of downtown, said investigating officer
Karen Thomas.
   However, a preliminary autopsy revealed Bernard may have undergone a heart
attack, said Griffin.
   Bernard had been on her way to what she thought was lunch at the Golden
Buddha restaurant with her husband George, said George Bernard.
        In reality, however, she was on her way to a surprise retirement party
thrown by her husband and  co-workers from the Internal Revenue Service. Bernard
had been an accountant for the IRS for 30 years, her husband said.
   In the past 10 years, two other individuals also died when their cars went
into the river along the same two miles of River Street, said Thomas.
        No one from the highway department would comment on why there are no guard
rails along the section of road where the accidents occurred.
 
 
 
 
 
        A local restaurant owner was arrested on Interstate 90 Thursday night in
connection with the murders of four family members and the attempted murder of a
fifth family member.
        Jeremy Hanna, 45, 124 Forest St., was charged with the murder of his brother
Thomas Hanna and three members of his brother's family. Hanna and his brother
were co-owners of the Little Middle East restaurant, 62 Ontario St.
        Police found the bodies of Thomas Hanna, 42; his wife, Kathy, 40; and their
children, Rebecca, 14; and Jeffrey, 6, at their 75 Atlas Road home.
   A third child, Joshua, 8, is in critical condition at Park General Hospital
after being shot in the chest, said Harrison Boxer, an emergency room
supervisor.
        Neighbors called police after hearing several rounds of gunfire following a
heated argument at the Hanna home at about 5:30 p.m. Several neighbors said they
saw the suspect, a frequent visitor to his brother's home, run out of the house
and drive off in his van.
        A waitress at the Little Middle East restaurant told police that half an hour
before the shooting occurred, Jeremy Hanna quarreled loudly on the telephone
with his brother, then ran out of the restaurant.
        Police said Hanna became enraged after he discovered $5,000 missing from the
restaurant's bank account. He assumed his brother Thomas took the money because
his brother was a month late on his home's mortgage payment, said police.
        Hanna was being held Friday in the county jail in lieu of $500,000 bail, police
said.
 
 
 
 
 
 
The radio version consisted of the television sound track on audio tape, played
only once.
        After exposure, each subject's tape-recorded account of the story was
transcribed. These accounts were then examined to assess how many (and which) of
the details of the story were recalled by the subject, and with what level of
accuracy. To accomplish this, each detail in each of the stories was identified
and listed. For example, the fire story contained such details as "A city
employee was arrested," and "He was arrested in connection with a fire" and "The
fire had destroyed his boss's home." The fire story contained a total of 48
separate details; the drowning story contained 37, and the murder story was
composed of 48 details. Next, each detail recalled by a subject was scored on a
four-point level of accuracy scale, where 0 = "no facts of detail correct;" 1 =
"at least some facts essentially correct;" 2 = "most facts essentially correct;"
and 3 = "all facts essentially correct." For example, a subject who recalled all
48 details of the fire story very accurately could have earned up to 3 x 48 =
144 points.
        Two analysts were used to assess the level of accuracy of the details recalled
by each subject. The analysts worked independently when scoring each subject's
transcription, and the scores were correlated to provide an index of reliability
(r = .927). The two detail accuracy scores for each subject, determined by the
two analysts, were then averaged for each subject to obtain a final measure of
unaided recall of details.
Findings
        The results from the experiment offer evidence for assessing the news recall
theory in terms of the adequacy of its logical prediction (as stated in
proposition seven). To make this assessment, four issues can be examined: First,
in an overall sense, how well did these 480 subjects remember the story to which
they were exposed? That is, was a pattern of leveling (loss of details) in
evidence? Second, which of these details did they recall best as central or
salient, and which tended to be forgotten? In other words, did selective recall
and sharpening take place? Third, were extraneous ideas and meanings
incorporated in these configurations? In other words, was there evidence of
assimilation? Fourth, and finally, into what Gestalts of meanings did they
organize their versions of the story? That is, did they encode it in memory as a
logical configuration of situations, actors, events and consequences?
        Limited Recall of Details. Addressing the leveling issue, the overall results
are consistent with a consensus reached among many previous studies. In this
experiment, even though these subjects were intelligent, urged to perform well,
queried immediately after exposure, and tested in a controlled attention
situation, the majority could recall very little. The following figures
represent how well at least half of the subjects could recall the details of the
individual stories, even at the lowest level of accuracy: On average, only 12 of
the 48 details in the fire story, that is 25 %, could be recalled even at the
lowest level by at least half of the subjects. For the 37 details of the
drowning story, the comparable figures were 10 details, or 27%, recalled by at
least half of the subjects. Only 9 of the 48 details of the murder story, 19
percent, were recalled by at least half of the subjects, again at even the
lowest accuracy level.
        There were also differences by medium. For example, on average, newspaper
presentations were remembered best, but with only 31.02% of details recalled at
even the lowest level of accuracy. This was followed by computer presentations
with a similar mean recall of 29.70%. Stories presented by television resulted
in a recall of 27.50% of details. Finally, radio-presented stories resulted in a
recall of only 25.88% of details. The statistical significance of these
differences along with comparisons by story and gender have been reported
elsewhere. [41]
        By any measure or comparison, then, and regardless of the abilities and
motivation of the audience or circumstances of exposure, leveling was very much
in evidence. This finding is consistent with the prediction made in proposition
seven in the theory.
        Selective Recall of Details.  Addressing the second issue, the results of the
analysis indicate clearly that sharpening did take place. In each story, only a
few details achieved a mean accuracy score of  at least 1.00 or greater (on the
0 to 3 scale).  In the fire story, for example, fourteen details were recalled
with mean scores ranging from 1.048 to 1.810. This included such details as "a
city employee was arrested," "his name is Warren James," and "fire destroyed his
boss's home." These fourteen details were also recalled by the largest number of
subjects (from 42% to 72.2%). In the drowning story, eleven details stand out as
having been recalled better than others (mean scores of 1.042 to 1.772). In the
murder story, fourteen details were recalled with  mean scores ranging from
1.027 to 1.993. These details were recalled at those levels by as many as 83% of
the subjects.
        Conversely, a large number of details were selectively dropped from the stories
recalled by the subjects. These were details receiving mean accuracy scores of
.500 or lower--which means that almost none of the facts were recalled
correctly. In the fire story, a total of fourteen details received such low
scores. In fact, these were completely forgotten by as many as 92% of the
subjects. This category included such details as "the name of the Fire Chief is
Ed Wilson," "flames came out of the attic window," and "the fire started at
about 10 a.m." In the drowning story, ten details received mean scores of .500
or lower and were completely forgotten by as many as 95% of the subjects.
Finally, in the murder story, twelve details achieved the lowest accuracy of
recall scores and were not remembered by as many as 97% of the subjects.
        There is little doubt, then, that patterns off encoding and recall were highly
selective. A limited list of details was recalled at least somewhat accurately
by relatively large numbers of the subjects. However, many other details were
ignored or forgotten by the majority. As predicted by the theory, then,
sharpening appears to be a clear pattern in the way these subjects encoded and
recalled these spot news stories.
        Evidence of Assimilation. Content analyses of  25 randomly selected accounts of
each story were conducted to identify examples of assimilation--modifications of
the basic accounts reported by the subjects due to their psychological sets or
use of culturally familiar language. This analysis was not intended as a full
quantitative audit of every change that took place. It was a relatively simple
qualitative analysis, designed to see if such modifications of the language,
details or meanings contained in the stories were present. It was not difficult
to identify a number of examples of assimilation. The following illustrate what
happened:
        In the fire story one subject reported that James, the alleged arsonist, was an
"assistant to the mayor" (rather than to the Director of the Parks and
Recreation department). The fire chief's name became Arlington (actually the
name of James' boss). Another account indicated that James was "accused of
arson" (in the story the police only "suspected arson"). In a change of gender,
Ms. Arlington ( James' boss) became "Mr. Arlington." Finally, one subject
indicated that "James had spent time in a mental hospital." (The story reported
only that he was at the county mental center "undergoing psychiatric testing.")
        In the drowning story, one subject indicated that Norma Bernard's "car went
into a lake" (rather than the river).  Still another reported that the car "went
into Lake Ontario" (rather than the Ontario River). One subject reported that
"the lady died as she went into Ontario Lake." One account claimed that she
"died of hypothermia" (rather than a heart attack). Even more colorfully, one
account indicated that the car "skidded off a bridge that had no guard rails."
As to guard rails, one subject reported that the "police would not comment on
their absence." (Highway department officials were the ones who refused to
comment).
        In recounting the murder story, a number of examples were found. One subject
reported that "four bodies were found around Interstate 90" (the assumed
perpetrator was arrested on that highway, but the bodies were found in their
home).  One subject relocated the restaurant, indicating that it "is off
Interstate 90" (as opposed to 62 Ontario Street). In another change, the $5,000
"was missing from the safe" (as opposed to the restaurant's bank account). In
one case, the loud quarrel between Hanna and his brother took place "in the
restaurant" (rather than via telephone). In another recalled story, "Hanna was
arrested on four counts of murder" (a phrase from the subculture of TV police
dramas and local news). Several accounts indicated that neighbors heard "blasts"
(not rounds) of gunfire."
        While these are not gross distortions, they offer clear examples of words,
ideas and expressions that were assimilated into the story from psychological
sets or culturally familiar language. Given the fact that all of the subjects in
the experiment were successful college students in communication courses, and
under the pressures for accuracy described earlier, one might anticipate that
there would be relatively few such modifications. That does not appear to be the
case, and the patterns of assimilation predicted by the theory were very
definitely a part of the news recall process in this experiment.
        Modal Patterns of Recall.  The theory predicts that the leveled, sharpened and
assimilated information recalled from the news story by the subjects will be
restated as a logical configuration of central and salient (plus some colorful
but minor) details. There is no simple way to check this prediction fully, short
of displaying each recorded account. Obviously, every one of the 480 versions
cannot be reproduced here. However, whether or not the theory predicted what
happened can be answered in a preliminary way by re-constructing the "modal"
version of each news story from the best remembered details recalled by the
greatest number of subjects (mean accuracy of recall scores of 1.00 or greater).
These modal patterns (see Figure 3) show not only what was best recalled but how
it tended to be organized by the subjects. Moreover, these modal versions can be
compared to the original news story.
Figure 3
 
MODAL STORIES
(Constructed from details best remembered by subjects)
 
          Fire Story                        Drowning Story                               Murder Story
 
 
   Warren James, a city employee, was arrested at his home Tuesday evening in
connection with a fire that destroyed his boss's home, valued at $120,000.
James, whose address is 35 Elysian Dr., had been dismissed from his job last
night. A blond and slender man fitting James' description was observed leaving
the scene. The boss, Terry Arlington, lost her collection of 19th century
children's books in the fire. Twelve firefighters responded to the blaze.
   A woman, whose car left the road and plunged into the Ontario River, may have
suffered a heart attack. Norma Bernard, 65, of 12 Adams Street, worked for the
Internal Revenue Service and was on her way to a surprise retirement party.
During the past 10 years, two other persons have died along the same section of
the road where there are no guard rails. At first, police thought she was alive,
but suffering from hypothermia.
   Jeremy Hanna, 45, a local restaurant owner, was arrested  Thursday night  on
Interstate 90 in Connection with the murders of four family members. These
included his brother, Thomas Hanna, eight-year- old Joshua, and six-year-old
Jeffery. A third child is in critical condition. Police said Hanna became
enraged when he discovered $5,000 missing from the restaurant account and
assumed his brother took the money.
 
 
 
        While there were many variations from one subject to another, most individual
accounts had considerable similarity to these modal stories. Overall they were
well-organized, leveled and sharpened versions of the original, with a number of
examples of assimilation. The subjects' stories efficiently identified and
displayed the details that make up the "core" of the story. In addition, as
noted, many minor details were dropped.
        An important question is whether there were qualitative differences in the
kinds of details that were incorporated into these accounts. The theory suggests
that those discussing settings, actors, events and consequences will influence
the core details in the structure of the recalled story. One way of
understanding the qualitative characteristics of details remembered from the
stories is to organize those best recalled into categories based on the kind of
information expressed in each.
        The qualitative characteristics of the best-recalled details were classified by
two coders working independently, each of whom was asked to sort all of the
details in the original stories into one of six categories. The categories are:
(1) Major details about central actors;
 (2) Core/essential details of story events; (3) Minor details about central
actors; (4) Minor/supporting details of story events; (5) unusual or dramatic
details; and (6) Details about minor actors. The coders then compared their work
and reached consensus on the placement of any details with which they disagreed.
Finally, mean accuracy of recall scores for each category were computed and are
shown in Table 1.
Table 1
 
CATEGORIES OF DETAILS REMEMBERED
(All Stories and Media Combined; N=480)
 
                   Mean Accuracy Score                          Detail Category
                        1.055                           Major details about central actors
                        0.999                           Core/essential details of story events
                        0.713                           Minor details about central actors
                        0.641                           Minor/supporting details of story
                        0.529                           Unusual/dramatic details
                        0.452                           Details about minor actors
 
        Table 1 indicates, in a simple descriptive way, that when all stories and media
are combined, subjects remember the major details about central actors best,
followed closely by the core or essential details of story events. Minor details
about central actors, or minor details of story events followed next. Unusual or
dramatic details, such as the "19th century collection of children's books" in
the fire story or the fact that main actor in the murder story was arrested on
"Interstate 90," often stand out in the recollections of the subjects, but not
ahead of the major details about central actors or the core details of the
story. Details about minor characters are least remembered. The same pattern
prevailed in an independent analysis of each story separately.
 
DISCUSSION AND CONCLUSIONS
        Developing a theory to explain how audiences will recall spot news stories from
a large and complex base of established concepts, principles and generalizations
developed in psychology over more than a century is a task with many potential
pitfalls. Controversies and arguments can be generated over which schools of
psychology are appropriate, which theories or bodies of findings provide the
most adequate base for explaining processes of attention, perception, memory and
recall, or even if such an approach has any merit at all. The form of the theory
may also invite debate. Inductively formulating such a theory in axiomatic form
from such a complex intellectual base, with "as if" postulates, a common-logic
calculus, and a complex prediction of what would be found if the theory is
correct, has not been common in communication research. In this approach, the
theory is modeled directly on formats used in physical science and have been
clearly set forth by numerous philosophers of science. Even so, such theory
structures seem very different from what is normally found in communication
scholarship and may invite criticism from some. On the other hand, this approach
to theory-building has been increasingly advocated in communication research
methods textbooks for some time, even though few actual examples have been
proposed and assessed with empirical data.
        In spite of all these potential risks of criticism, it may be time to attempt
this type of theory construction more widely in the study of the relationship
between news and audiences. This systematic approach has served the physical,
biological and even social sciences well. The present paper begins the task of
developing just such a systematic theory. No claim is made that it is the only
way to explain the kinds of news-related behavior that it predicts. Furthermore,
it is clearly understood that further research may force modification, or
possibly even total abandonment, of this particular formulation. However, that
is precisely the central goal of academic research--to test predictions made by
theories to see if they correspond to reality. There is little reason to
conclude that increasing attention to formal axiomatic theories in the field of
mass communication will limit our understanding of its processes and influences
on audiences. Moreover, a significant advantage is that such systematic theories
can offer clear guides to research that is needed to assess their merits. If
extensively tested and refined as a result of such research, such a formulation
may be of practical use in understanding how people attend to, cognitively
process, remember and recall stories they encounter in the news. It is possible
that understanding these processes better will at some point enable journalists
to prepare news stories that people will remember more fully and accurately.
        In the present situation, the overall conclusion from the data-analysis is that
support was found for the theory. In controlled experimental settings, 480
subjects attended to one of three local spot news stories presented by one of
four media. Immediately afterward, they were asked to recount the content of
that news story. Content analyses of their reports show that the subjects in the
experiment did show the patterns of recall that are predicted by the theory.
When called upon to relate what they had read, seen or heard, these subjects, by
and large, produced an efficient and logical configuration of central and
salient details providing a leveled and sharpened version of the story with
clear evidence of assimilation. While in an overall sense, what they remembered
was limited--as has been the case in many studies of news recall--their versions
contained details about the major actors and the central events that define the
core of the story as well as some colorful or dramatic details. In short, the
theory seemed to predict with some accuracy the ways in which these intelligent
and motivated subjects attended to, learned and recalled the spot news stories
under study.
 
REFERENCES:
[1] Examples are: John Stauffer, Richard Frost, and William Rybolt,
     "Recall and Learning from          Broadcast News: Is Print Better?" Journal of
     Broadcasting, 25:253-262 (1981); D. C.; Williams, J. Paul and J. C.
     Ogilvie, "Mass Media, Learning and Retention," Canadian Journal of
     Psychology, 11: 157-163 (1957); and Karen Brown, "Comparisons of Factual
     recall from Film and print Stimuli," Journalism Quarterly, 55: (350-353
     (Summer, 1978).
[2]  Barrie Gunter, "News Awareness and Retention Across the Audience,"
     Chapter 4 in Poor          Reception: Misunderstanding and Forgetting Broadcast
     News (Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers, 1987), pp.
     83-109. See also: P. J. Tichenor, G. A. Donahue and C. N. Olien, "Mass
     Media Flow and Differential Growth of Knowledge," Public Opinion Quarterly
     , 34: 158-170 (1970); and J. P. Robinson, "World Affairs Information and
     Mass Media Exposure," Journalism Quarterly, 44: 2-40 (Spring, 1957).
[3]  See, for example, the extensive review of relevant literature on such
issues in: Prabu David,         "The Role of Imagery in Recall of Deviant News,"
Journalism and Mass Communication       Quarterly 73: 804-820 (Winter, 1996).
[4]  For a review of relevant literature to 1982, see: Colin Berry, Barrie
     Gunter and Brian Clifford,         "Research on Television News," Bulletin of the
     British Psychological Society, 35: 301- 304 (1982); Barrie Gunter, Colin
     Berry and Brian Clifford, "Remembering Broadcast News: The Implications of
     Experimental Research for Production Techniques," Human Learning 1: 12-38
     (1982); A summary of studies to 1991 can be found in Hans-Bernd Brosius,
     "Format Effects on Comprehension of Television news," Journalism Quarterly
     , 68: 396-401 (1991). Also: Dona Hayes and Melvin L. DeFleur, "The
     Influence of Production Formats on Audience Recall and Interest in
     Televised Local News Stories," Paper presented at the Annual Meetings of
     the Association for Education in Journalism and Mass Communication,
     August, 1992.
[5]  W. Russell Neuman, "Patterns of Recall Among Television News
     Viewers," Public Opinion   Quarterly, 40: 115-123 (1976).
[6]  Elihu Katz, Hanna Adoni and Pnina Parness, "Remembering the News: What the
Picture Adds    to Recall," Journalism Quarterly, 54: 231-239 (1987).
[7]  Barry Gunther, op. cit., p. xi.
[8]  Gunther, ibid, p. x.
[9]  John C. Reinard, Introduction to Communication Research, 2nd ed. (Boston:
McGraw Hill,    1998), pp. 48-50.
[10]  James McK. Cattell, "The Time Taken Up By Cerebral Operations," doctoral
dissertation,   1886. Cited in Henry E. Garrett, Great Experiments in Psychology
(New York: Appleton-    Century-Crofts, Inc., 1951) p. 313.
[11]  For a review of this research, see: Richard J. Harris, A Cognitive
Psychology of Mass              Communication, 2nd ed. (Hillsdale, N.J.: Lawrence Erlbaum
Associates, Publishers,         1994), pp. 26-29.
[12] M. Minsky, "A Framework for Representing Knowledge," in P. H. Winston, ed.,
The
        Psychology of Computer Vision (New York: McGraw Hill, 1974), p. 211.
[13]  Nancy L. Stein, Peter A. Ornstein, Barbara Tversky and Charles Brainard,
Memory for      Everyday and Emotional Events (Mahwah, NJ: Lawrence Erlbaum
Associates,
        Publishers, 1997), p. 167 See also: Charles F. Cofer, ed.,  The Structure of
Human   Memory (San Francisco: W. H. Freeman and Company, 1976), pp. 105-106.
[14]  John R. Anderson, Cognitive Psychology: and Its Implications, 2nd
     ed., (New York: W. H.      Freeman and Company, 1985), See especially:
     Chapter 3, "Perception and Attention," pp. 36-72 and Chapter 4,
     "Perception-Based Knowledge Representations," pp. 73-101; Also: Michael
     Eysenck, A Handbook of Cognitive Psychology (Hillsdale, NJ: Lawrence
     Erlbaum Associates, Publishers, 1987), pp. 23-47.
[15]  Gordon W. Allport and Leo Postman, "The Basic Psychology of Rumor,"
Transactions of the     New York Academy of Science, Series II, 1945, VIII, p. 68.
[16]  Kurt Koffka, "Gestalt Psychology," Psychological Bulletin, 19, April,
1922, pp. 111-130
 [17]  Wolfgang Kohler, Gestalt Psychology (New York: Liveright, 1929) and Kurt
Koffka,                 Principles of Gestalt Psychology ((New York: Harcourt, Brace, 1935).
[18]  James A. Schellenberg, Masters of Social Psychology (New York: Oxford
University Press,       1978), p. 64.
[19]  See Kurt Lewin, "Field Theory in Social Psychology," in Morton Deutsch and
Robert M.       Krauss, Theories in Social Psychology (New York: Basic Books, Inc.,
1965, pp. 37-76.
[20]  Systematic investigations of human memory began in Germany in
     1884-5. See: Hermann       Ebbinghaus, Uber das Gedachnis: Untersuchungen zur
     Experimentellen Psychologie  (Leipzig: Dunker and Humbolt, 1885) p. 27.
     Translated in 1913 as  Memory: A Contribution to Experimental Psychology
     (New York: Dover Press, 1964
[21]  Kathy Kellerman, "Memory Process in Media Effects," Communication
Research, 12:83-131     (January, 1985).
[22]  Nancy L. Stein, et. al. op. cit., p. 2.
[23]  See: Henry L. Roediger, III and Fergus I. M. Craik, Varieties of Memory
and Conciousness        (Hillsdale, JJ: Lawrence Erlbaum Associates, Publishers,
1989).
[24]  Edward Hamilton, Plato: The Collected Dialogues (New York: Bollingen
Foundation, 1961),      p. 904.
[25]  Hermann Ebbinghaus, op. cit. p. 27.
[26]  Frederick C. Bartlett, Remembering (Cambridge: Cambridge University Press,
1932).
[27]  Cited in Wayne Dennis, "Hermann Ebbinghaus: On Memory," Readings in the
History of      Psychology (New York: Appleton, Century, Crofts, 1967), p. 304.
[28]  See: William K. Estes, "One Hundred Years of Memory Theory," in David
Gorfein and     Robert R. Hoffman, (Eds) Memory and Laearning: The Ebbinghaus
Centennial              Conference (Hillsdale, NJ: Lawrence Earlbaum Associates,
Publishers, 1987), pp. 11- 33.
[29]  Gillian Cohen, Memory in the Real World (Hillsdale, NJ: Lawrence
     Erlbaum Associates,        Publishers, 1989), p. 2; See also: Ulric Neisser,
     "Memory: What Are the Important Questions?" in M. M. Gruneberg, P. E.
     Morris, and R. N. Sykes, (Eds.) Practical Aspects of Memory (London:
     Academic Press, Inc., 1978).
[30]  An anthology of such studies can be found int  Nancy L. Stein, et. al.
[31]  Bartlett, op cit.
 [32]  Gordon W. Allport and Leo Postman, The Psychology of Rumor (New York:
Henry Holt and  Company, 1947).
[33]  Alan F. Collins, Susan E. Gathercole, Martin Conway, and Peter E. Morris,
Theories of     Memory  (Hillsdale, NJ: Lawrence Erlbaum and Associates,
Publishers, 1993).
[34]  Eysenk, M. W. A Handbook of Cognitive  Psychology (Hillsdale, N.J.:
Lawrence Erlbaun        Associates, Publishers, 1987), p. 93.
[35]  Richard Jackson Harris, A Cognitive Psychology of Mass Communication
     (Hillsdale, N.J.:          Erlbaum Associates, Publishers, 1994), p. 25; see
     also: B. Hoijer, "Television-Evoked Thoughts and Their Relation to
     Comprehension," Communication Research 16: 179-203 (1989).
[36]  Charles F. Cofer, ed., The Structure of Human Memory (San Francisco: W. H.
Freeman and     Company, 1976), pp. 105-106.
[37]  Melvin L. DeFleur, Lucinda Davenport, Mary Cronin and Margaret DeFleur,
"Audience       Recall of News Stories Presented by Newspaper, Computer, Television
and Radio,"
        Journalism Quarterly 69: 1010-1022 (Winter, 1992).
[38]  Ernest Nagel, The Structure of Science: Problems in the Logic of
Scientific Explanation (New     York: Harcourt Brace and World, Inc., 1961), p. 90.
[39]  Abraham Kaplan, The Conduct of Inquiry: Methodology for Behavioral Science
(San            Francisco: Chandler Publishing Company, 1964), p. 299.
[40]  See: Melvin L. DeFleur, et. al. op. cit. The present article describes a
theoretical analysis    based in part on those data.
[41] See: DeFleur, et. al., ibid., pp. 1010-1022.