11.2 Historical Overview

Much research on the effects of television is contradictory or inconclusive, but that doesn't make the research useless, wasteful, or futile. We need to know as much as we can about how children learn, and conscientious research of any kind can teach us, if nothing else, how to do better research (Rogers & Head, 1983, p. 170).

As Fred Rogers and Barry Head suggest, to use research on television, one needs a historical perspective. The purpose of this section is to provide that perspective. It will briefly explain the evolution of the technologies, important historical milestones, the evolution of the research, and the variety of methodological approaches used. After reading this section, you should be able to place the research in historical context and understand its significance.

11. 2. 1 Contributors to the Literature

This large body of research is the result of individuals, organizations, and fields with constituencies naturally interested in the effects of television. The disciplines that are most dedicated to reflecting on learning from television are education, communications, psychology, and sociology. Within education, the fields of educational psychology, cognitive science, and instructional technology have a continuing interest. Educational psychology and cognitive science have focused on mental processing. Instructional technology has made its greatest contributions to television research through the areas of message design, formative evaluation, and critical-viewing skills.

11.2.1.1. Organizations. Groups associated with research on television operate in diverse arenas. Government institutions, such as the National Institute of Mental Health (NIMH), the Educational Resources Clearinghouse on Information Resources (ERIC), and the Office of Research in the Department of Education have been the catalyst for many studies. Government has influenced research on television through hearings and legislation on violent programming and commercials for children. Government legislation also created the Public Broadcasting System (PBS).

Many universities have established centers or projects that pursue questions about the effects of television. These include the Family Television Research and Consultation Center at Yale University, the Center for Research on the Influence of Television on Children at the University of Kansas, the National Center for Children and Television at Princeton, and Project Zero at Harvard University. Foundations have supported research in the areas of media effects and instructional television, including the Spencer, Ford, and Carnegie Foundations. Public service organizations such as Action for Children's Television and church television awareness groups have spurred policy and research.

Research and development (R&D) organizations, such as the Southwest Educational Development Laboratory, have generated curricula on critical-viewing skills. Children's Television Workshop (CIV), the producer of Sesame Street, is an R&D organization that not only develops programming but also does research on the effects of television.

11.2.1.2. Review Articles and Books. Despite such longterm efforts, much of the literature on television lacks connection to other findings (Clark, 1983, 1994; Richey, 1986). The conceptual theory necessary to explain the relationship among variables is still evolving. Because of this, consumers of the literature are sometimes overwhelmed and unable to make decisions related to interactions in the television viewing system of programming, environment, and behavior.

Comprehensive and specialized reviews of the literature are helpful for synthesizing findings. Individual studies contribute a point of view and define variables, but it takes a review to examine each study in light of others. Fortunately, there have been many outstanding reviews of the literature. For example, Reid and MacLennan (1967) and Chu and Schramm (1968) did comprehensive reviews of learning from television that included studies on utilization. Aletha

Huston (1972) wrote a chapter for the National Society for the Study of Education (NSSE) yearbook on Early Childhood Education entitled "Mass Media and Young Children's Development" which presented a conceptual framework for studying television's effects. In 1975, the Rand Corporation published three books by George Comstock which reviewed pertinent scientific literature, key studies, and the state of research. Jerome and Dorothy Singer reviewed the implications of research for children's cognition, imagination, and emotion (Singer & Singer, 1983). In that article, they described the trend toward studying cognitive processes and formal features. By 1989, the American Psychological Association had produced a -synthesis of the literature tided Big World, Small Screen.

Other reviews have concentrated on special areas like reading skills (Williams, 1986); cognitive development (Anderson & Collins, 1988); instructional television (Cambre, 1987); and violence (Liebert & Sprafldn, 1988). Lawrence Erlbaurn Publishers offers a series of volumes edited by Dolf Zillmaim and Jennings Bryant on research and theory about television effects.

Light and Pillemer (1984) argue against the single decisive-study approach and propose reviews around a specific research question that starts by reporting the main effects, then reports special circumstances that affect outcomes, and finishes by reporting special effects on particular types of people. This integrated research strategy is especially appropriate for reviews of research on television effects.

11.2.2 Evolution and Characteristics of the Technologies

The evolution of the technologies of motion pictures and television during the latter part of the 19th century and early 20th century can be described in terms of media characteristics, delivery systems, and communication functions. It is also important to know the terminology essential to understanding research descriptions and comparisons. This terminology is given in the glossary at the end of this chapter.

11.2.2.1. Functional Characteristics. These media characteristics of film and television are primarily realism or fidelity, mass access, referability, and, in some cases, immediacy. Producers for both of these technologies wanted to make persons, places, objects, or events more realistic to the viewer or listener. The intent was to ensure that the realistic representation of the thing or event was as accurate as possible (i.e., fidelity). The ability to transmit sounds or images to general audiences, or even to present such information to large groups in theaters, greatly expanded access to realistic presentations. In the case of television, the characteristic of immediacy allowed the audience to experience the representation of the thing or event almost simultaneously with its occurrence. The notion of "being there" was a further addition to the concept of realism. As these various forms of media developed, the ability to record the representations for later reference became an important characteristic. Viewers could not only replay events previously recorded but could also refer to specific aspects or segments of the recording time and time again for study and analysis. Each of these characteristics has driven or directed the use of film or television for instructional purposes.

11.2.2.2. Delivery Systems. The State University of Iowa began the first educational television broadcasts in 1933. Educational broadcasting quickly grew, with several universities producing regular programming and commercial stations broadcasting educational materials for the general population. During the 1950s and 60s, other technical innovations emerged that expanded the flexibility and delivery of educational television. These included the development in 1956 of magnetic videotape and videotape recorders, the advent of communications satellites in 1962, and the widespread growth of cable television in the 1960s and 70s. Delivery systems encompass both transmission and storage capabilities. The various means whereby the message is sent to the intended audience differ in terms of the breadth of the population who can access the message. These means of transmission include broadcast television, communications satellite, closed-circuit television (CCTV), cable access television (CATV), and microwave relay links.

Broadcast television programming is generally produced for large-scale audiences by major networks and, with the exception of cable or microwave relay agreements, can be received free of charge by any viewer with a receiver capable of receiving the signal. Satellite communication has the capability of distributing the television signal over most of the populated globe. Closed-circuit television is produced for limited audiences and for specified educational purposes. Cable television often presents programming produced by public television organizations, public service agencies, or educational institutions for educational purposes. Today, many of the microwave relay functions have been replaced by satellite relays; however, this transmission medium is still used to distribute closed-circuit programming within prescribed areas such as school districts.

11.2.2.3. Storage Media. In the beginning, television productions were often stored in the form of kinescopes, which are rarely, if ever, used today, although some early television recording may still exist in kinescope form. Today, most video programs are stored on videotape cassette format, which is convenient and is produced in a variety of tape widths. Videotape permits a large number of replays; however, it can deteriorate after excessive use.

11.2.2.4. Communications Functions. From an instructional point of view, the most important factor in the development of any of these technologies is not the technical aspect of their development but rather the impact of the medium on the audience. Terms that relate to communications functions include instructional television (IT'V), educational television (ETV), mass media, incidental learning, and intentional leaming. Today, ITV programming is often transmitted by satellite to a school where it is either recorded and used when convenient or used immediately and interactively through a combination of computers and telecommunications. Educational television programming is typically not part Of a specific course of study and may be directed to large and diverse groups of individuals desiring general information or informal instruction.

The distinction between mass media and educational television is frequently difficult to make since most educational television programming is distributed via broadcast television, the primary mass-media mode. What differentiates mass media from educational television is the notion of intended purpose. With educational television, intentional effects are achieved through purposeful intervention to achieve educational objectives. Incidental effects, on the other hand, typically result from mass-media or entertainmentoriented programming.

11.2.3 Legislative Milestones

The history of research on television effects has been tied to important government policy actions (Wood & Wylie, 1977). In the 1930s the government declared air channels to be public property and created the Federal Communications Commission (FCC) to regulate systems such as radio. After lengthy hearings, in 1952 the Federal Communications Comniission reserved 242 television channels for noncommercial, educational broadcasting.

11.2.3.1. The 1950s and 60s. The first congressional hearings on violence and television occurred in 1952. In 1954, hearings were held to investigate the link between television and juvenile crime. When he was doing his Bobo do]] social-psychology experiments in the early 1960s, Albert Bandura published an article in Look magazine entitled "What Television Violence Can Do to Your Child." This article popularized the term "TV violence."

In 1964, Newton Minow assumed the chair of the FCC. He would prove to be a strong commissioner, remembered for his statement that television was "a vast wasteland." By 1965, advertisers had discovered that they could reach young children with advertisements for toys, candy, and cereal more cheaply and effectively on Saturday mornings than in prime time. Also in the 1960s, Congress created the Public Broadcasting System (PBS) and the Corporation for Public Broadcasting (CPB). By the end of the 1960s, the National Commission on the Causes and Prevention of Violence had issued a report stating that exposure to violence on television increased rates of physical aggression. This led to the Surgeon General's appointing a committee to study the effects television programs have on children. The decade concluded by the Supreme Court's upholding the fairness doctrine, which required stations to give equal time to political candidates.

11.2.3.2. The 1970s. The decade of the 1970s started with a ban on cigarette advertising on television, which had been initiated after the Surgeon General's report that there was a relationship between cancer and smoking. In 1972, the Surgeon General issued a report on violence that alleged that there was also a causal link between violent behavior and violence on television and in motion pictures. This first major government report on television and violence (NIMH, 1972) consisted of five volumes of reports and papers gathered through an inquiry process directed by the National Institute of Mental Health (NIMH). To prepare for this report, NIMH was empowered to solicit and fund a million dollars worth of research on the effects of television violence (Liebert & Sprafkin, 1988). By 1975, the FCC had received 25,000 complaints about violent or sexually oriented programs on television. As a consequence, in 1975 the Ford Foundation, the National Science Foundation, and the Markle Foundation cosponsored a major conference on television and human behavior. The Supreme Court ruled that the FCC could regulate hours in which "indecenf' programming could be aired.

11.2.3.3. The 1980s. In 1982, the National Institute of Mental Health confirmed the link between television and aggression and stated that "violence on television does lead to aggressive behavior by children and teenagers who watch the programs" (NIMH, 1980, p. 6); thus television was labeled a cause of aggressive behavior. In 1985, the American Psychological Association (APA) publicly concluded that violence can cause aggressive behavior and urged broadcasters to reduce violence. - As the decade ended, the FCC decided that the Fairness Doctrine was no longer necessary because there was no longer a scarcity of stations and that it was perhaps unconstitutional. Congress passed a bill to reinstitute the doctrine, but the President vetoed it. The President also vetoed legislation that would place limits on advertising during children's programs. In 1989, Congress passed the Television Violence Act granting television executives the authority to hold discussions on the issues of television violence without violating antitrust laws.

11.23A. The 1990s. This brings us to the current decade, which started with Congress's passing the Children's Television Act that requires limits on advertising and evidence that stations provide programming to meet children's needs. This is the first legislation to establish the principle that broadcasters have a social responsibility to their child audiences. The advantage of this approach is that it avoids the thorny issue of censorship. The bill became a law without presidential signature. Congress established the National Endowment for Children's Television to provide resources for production of quality children's programming as well as the Television Decoder Circuitry Act, which requires all new sets to have closed-caption capability. Over presidential veto, Congress approved the Cable Television Consumer Protection and Competition Act to regulate the cable industry. In 1993, the National Research Council of the National Academy of Sciences published a comprehensive report on the causes of violence in American society, entitled "Understanding and Preventing Violence," which addressed the role of television. The Senate Commerce Committee held hearings on television violence during which Senator Hollings complained that Congress has been holding hearings on television violence for 40 years. The idea of "V-chip" legislation to require the technology in all sets to block showing of programs rated violent was introduced at these hearings. The Telecommunications Act of 1996 required that this V-chip be installed on all new television sets. This landmark legislation had other important provisions, including one for discounted service rates for telecommunication lines into schools, especially for lines for compressed video and the Internet (Telecommunications Act of 1996). This overview of societal concerns about television documents the impetus for much research.

11.2.4 Historical Evolution of the Research

The first major research initiatives in both film and television began in the 1950s and 1960s. Research foci and variables of interest, as well as the social orientation of research, have changed considerably over the years.

Bowie (1986) reviewed research on learning from films and grouped the research into three phases:

1. Research on whether films can teach (1910-1950)
2. Research on how films teach (1940-1959)
3. Research on who learns from films (1960-1985)

Research from the last phase includes a great many experimental studies. The results of these experimental studies can be grouped in these areas: (a) use of films to teach higber-level cognitive skills, (b) effects of film viewing on individual learning, and (c) effects of film viewing on self-concept. Bowie concluded that the literature reviewed in these three areas suggests that:

• Films are effective in teaching inquiry leaming and problem solving.
• Unstructured films are more effective for teaching problem solving.
• Films are effective in teaching observation skills and attention to detail.
• Low-aptitude students tend to benefit more from films.
• Films tend to be more effective for field-independent students.
• Films can positively influence self-concept.

Research on leaming from films also served as a basis for research on instructional television.

Television research began with attention being devoted almost solely to its instructional effectiveness in formal instructional environments. The types and foci of research evolved into more varied agendas that considered not only the formal instructional implications of television but also the social, psychological, and instructional effects of broadcast television in less formal environments.

Sprafkin, Gadow, and Abelman (1992) describe the research on television as falling into three distinct chronological phases. The first of these they refer to as the "medium-orientation phase," in which television was seen as a powerful instructional tool that required research to describe its effectiveness. At this point, little attention was devoted to assessing the interaction of the media with developmental or individual differences in the viewers. The second phase that Sprafkin et a]. describe is the "child orientation phase," in which research focused more closely on the relationship of television to young viewers"individual characteristics and aptitudes. Media effects were thought to be due to a child's mental-processing characteristics, not to programming. They termed the third phase the interaction phase, in which television effects were seen as complex three-way interactions between characteristics of the medium (such as type of content), the child or viewer variables (such as age), and factors in the viewing environment (such as parents and teachers). These three phases correspond approximately to the three eras of film and television research: the period of comparative media research (see 39.5.4) during the 1950s and early 1960s (Greenhill, 1967); the media effects and individual differences research of the late 1960s through the 1970s (Anderson & Levin, 1976; Wright & Huston, 1983); and die interaction research characterized by the work of Salomon (1979, 1983) during the later 1970s through the present time.

The purpose of this section is to chronicle the evolution of these research trends and describe the nature of the research associated with each phase. In doing so, we will attempt to relate the trends to methodologies and variables.

11.2.4.1. Research Prior to 1965. Before the mid1950s, the vast majority of research was focused on the effects of instructional films, usually in controlled educational or training environments, both in formal education and in military and industrial training. This period was marked primarily by the widely quoted Instructional Film Research Program conducted at the Pennsylvania State University. This program was initiated under the auspices of the U.S. Naval Training Devices Center to study a variety of variables related to the use of instructional fihns for personnel-training purposes. One report issued through this research project summarized and evaluated over 200 film research studies from 1918 until 1950 (Hoban & Van Ormer, 1950). The major focus of the Instructional Film Research Program, however, was the conduct of an extensive series of experiments that compared instruction delivered via film with "conventional" or "face-to-face instruction." Within these comparisons, researchers also investigated the effects of various production techniques, the effect of film-based instruction on learner attitudes, and the effectiveness of various applications of instructional films (Carpenter & Greenhill, 1956; Greenhill, 1967). This series of studies represents one of the first, and certainly most extensive, attempts to evaluate thoroughly the effectiveness of the medium. The findings of these studies, however, indicated no significant differences in most cases and have been criticized for a number of methodological procedures (Greenhill, 1967).

Typical among the studies conducted in this program were those that sought to compare the relative effectiveness of motion-picture-based instruction with conventional classroom instruction. A study by VanderMeer (1949) compared ninth-grade biology students taught by: (1) sound films, (2) sound films plus study guides, and (3) standard lecture-demonstration classroom instruction. No significant differences were found across all groups on either inimediate or three-month-delayed achievement testing, although the film-only group showed a shorter completion time. This study is quite characteristic of most of these film studies in that no significant differences were found across both the experimental and control groups. Other studies focused on the relative effectiveness of instructional films for teaching performance skills and generally found no significant difference or only slight benefit from the film treatment (Greenhill, 1967). The effects of production variables were also of interest to researchers, and the relative effects of such variables as inserted questions, variants in the sound track, color versus monochrome, animation versus still pictures, and the use of attention-gaining and directing devices were all studied, albeit with few, if any, significant differences across groups.

The period between the mid-1950s and the mid-1960s was characterized by a great deal of instructional television research by a group of researchers at the Pennsylvania State University, reconstituted as the Instructional Television Research Project (Carpenter & Greenhill, 1955), as well as by other individuals (Hagerstown Board of Education, 1959; Holmes, 1959; Kumata, 1956; Niven, 1958; Schramm, 1962). These projects and summaries of research included literally hundreds of studies covering many content areas and many different age groups. In most cases, the summary reports issued by these researchers or projects provided fairly comprehensive descriptions of the general findings and conclusions. As with the film research initiatives, the television research projects focused strongly on comparative research designs and similarly resulted in "no significant differences." Few studies reported findings entirely supportive of television, and conversely few found television instruction to be less effective than conventional classroom instruction. The finding of no significant difference was seen by Greenhill (1967) as a positive result because it implied that television could be a reasonable alternative to classroom instruction and consequently, for reasons of administrative, fiscal, and logistical benefit, could be a more desirable choice of instructional method.

The comparative studies of television conducted during this time were later criticized on methodological grounds by Stickell (1963) and Greenhill (1967). Stickel] analyzed 250 comparisons and determined that only 10 were "interpretable" methodologically. Those 10 had employed random assignment of subjects, control of extraneous variables, and application of appropriate tests of significance in which the underlying assumptions of the test were met. Of the studies Stickell found to be "interpretable," none revealed significant differences.

The majority of these early comparative studies were designed to compare various forms of televised instruction to a vaguely specified standard known as "face-toface instruction ... .. conventional," or "traditional classroom instruction" (Carpenter & Greenhill, 1956; Lumsdaine, 1963). Instructional techniques and formats included (a) a single instructor teaching the same content, (b) a "live instructor" teaching a class while the same class was being televised to a remote class, (c) a number of different instructors teaching the same general lesson as the televised lesson, and (d) kinescope recordings of a lesson augmented by various, instructor-led activities. In most cases, there was little or no means of equating the instructional formats being used in terms of instructor equivalence, content congruence, or environmental similarity (Greenhill, 1967; Wilkinson, 1980; Williams, Paul & Ogilvie, 1957). Among the large number of comparative studies, there are many that simply compared the medium with some standard of live classroom instruction, while a smaller proportion made comparisons with the audio message only, comparisons of film versus kinescope, and television versus an in-studio classroom (Kumata, 1956).

As mentioned, this matter was further complicated by the fact that the vast majority of the studies, in both film and television, produced results of "no significant difference" (Greenhill, 1967; Stickell, 1963). This finding, when considered in conjunction with the general comparative nature of the research, makes it difficult to draw specific conclusions or recommendations from most of these comparative studies. Other methodological problems also plagued this early research, including: lack of equivalence of experimental groups, confounding of variables, and statistical analys s procedures that were not powerful enough to detect differences that may have been present (Greenhill, 1967).

In terms of group equivalence, two problems were apparent. First, groups were rarely pretested to determine if prerequisite knowledge was approximately equivalent. Second, little attention was given to ensuring equivalence of assignment to experimental groups. In some cases, correlative data such as IQ scores or grade point averages were used as matching variables, but because of the use of intact classes, randomization was rarely employed to assign subjects (Chu & Schramm, 1968; Stickell, 1963). Because the variables of televised instruction and conventional instruction were not clearly defined, it was almost impossible to separate other mediating variables related to production methods, technologies, viewing and teaching environments, viewer characteristics, and content organization. The result was often a serious confounding of many variables, only some of which were of interest. In terms of statistical analysis, t and F tests were used only occasionally, and analysis of covariance procedures were employed rarely because adjusting variables were infrequently assessed (Stickell, 1963). Additionally, content-related factors and objectives as well as types of learning were often not addressed or confounded (Miller, 1968).

Other more carefully defined variables continued to be investigated during this time, including: technical or production variables such as color, camera techniques, and attention-gaining and directing devices (Ellery, 1959; Harris, 1962; Kanner & Rosenstein, 1960; Schwarzwalder, 1960); pedagogical variables, such as inserted questions and presentation modes (Gropper & Lurnsdaine, 1961; Rock, Duva & Murray, 195 1); and variables in the viewing environment, such as viewing angle, group size, and distractions (Carpenter & Greenhill, 1958; Hayman, 1963; McGrane & Baron, 1959). In addition, attitudes toward televised instruction and the use of television to teach procedural skills were studied (Hardaway, Beymer & Engbretson, 1963; Pasewark, 1956).

Later studies, conducted during the 1960s and early 1970s, focused more specifically on individual variables, media characteristics, and die interaction between viewer characteristics and television effects. These studies typically employed the aptitude-treatment-interaction paradigm (see 22.3.3 to 22.2.7) described by Cronbach and Snow (1976) and were intended to explore specific effects of television on particular individuals. These designs were inherently more precise and more powerful and consequently enabled researchers to identify the effects of individual variables as well as the interaction of variables and other factors (Levie & Dickie, 1973).

During this time period, studies employed quantitative experimental methods almost, exclusively to evaluate the relative effectiveness of film and televised instruction in generally controlled environments such as laboratories, studios, classrooms, and schools. Researchers did not have the resources or research interest to investigate or describe specific effects on larger or noncontrolled populations, such as the effect of incidental learning resulting from noneducational broadcast television.

11.2.4.2. Research from 1965 On. After 1965, research focus was increasingly directed toward mass media and social effects. The formation of Children's Television Workshop (CTW) in the late 1960s directed research interest to formal features and formative evaluation (Polsky, 1974). The 1970s were also devoted to research on the relationship between televised violence and aggression. With the 1980s, a change from the behavioral to cognitive paradigm in psychology stimulated further research on mental processing (see 5.4.1 to 5.4.4) and formal features. Some research questions have persisted from the 1960s until the present, such as effect on school achievement and aggression. Research evolved from a focus on specifying variables to describing the relationships and interactions among variables. More-varied research agendas have considered not only the formal instructional implications of television but also the social, psychological, and instructional effects of broadcast television in various, less-formal environments (Comstock & Paik, 1987; Huston et al., 1992).

11.2.5 Methodological Approaches

Historically, research on television has employed four methodologies: experimental, qualitative, descriptive, and developmental (see Chapters 39 to 42). There has been a general chronological correspondence between certain methodologies and research foci, for example, between comparative studies and instructional effectiveness and between correlational studies and school achievement. For this reason, it is important to understand that research related to television has, over the years, come to address more than simply the effects of televised instruction on learning. Evolving societal demands brought about the need for different methodological approaches to study the disparate effects of television on types of viewers, on variations in viewing environments, on socialization effects, and on interaction with programming variables (Cambre, 1987). Such a broad base of research agendas has necessitated reliance on research methodologies other than those of a traditional empirical nature.

The vast majority of current television research reflects these four methodological approaches: experimental, qualitative, descriptive, and developmental. This section deals with these various research methodologies with regard to their purposes, strengths, and weaknesses as they apply to film and television research.

11.2.5.1. Experimental Methodology. Early research in television effects utilized traditional experimental designs (see Chapter 39), albeit with different levels of robustness and precision. The era of film and television research conducted during the 1940s through the mid-1960s, which has been referred to as the period of comparative research studies (see 39.5.4) generally used traditional experimental designs, such as those described by Campbell and Stanley (1963). Although many of these studies were methodologically weak in that they did not employ randomization of groups, pretests, or control groups, and have been subsequently criticized for these reasons (Greenhill, 1967; Stickell, 1963), it is important to note that there were many methodologically rigorous studies conducted during this period which continue to provide useful insights, not only into the comparative effects of television and traditional classroom instruction but also into the effects of specific variables, such as color, inserted questions, and presentation techniques (Greenhill, 1967; Reid & MacLennan, 1967). During the period of time from the mid- I 960s through the 1970s, other empirical studies were prompted by (a) better design conceptualization such as the aptitude treatment interaction paradigm, (b) more robust statistical analysis techniques, and (c) greater attention to the individual characteristics of the medium, the child, and the viewing environment. Increasingly, research moved from the laboratory or classroom to the home and social environment. Two types of experimental studies that compare variables are common in research on television: laboratory and field experiments. The former has advantages when comparing theories, testing hypotheses, and measuring effects; the latter is suited to checking the results of laboratory experiments in real-life settings (Comstock, 1980). An example of a laboratory experiment would be three treatments (i.e., violent first segment, violent last segment, and nonviolent segment) given to three randomly assigned groups who are given written instruments assessing recall. An example of a field experiment would be randomly assigning children to watch specific television shows at home and then administering attitude surveys and comprehension measures.

The major advantage of the laboratory experiment is that random assignment of subjects to specific treatment conditions can control for the effect of other variables. The disadvantage is that there is no certainty that the setting is realistic. The major disadvantage of the field experiment is that it produces little consistent evidence because control of variables is less rigorous. Nevertheless, one can be more confident in how realistic the findings are with a field experiment; however, realism and validity are gained at the expense of control of variables and the possibility of drawing causal conclusions. Laboratory research, on the other hand, generally allows one to draw cause-effect conclusions about interactions.

11.2.5.2. Qualitative Methodology. Qualitative research methodology (see Chapter 40) includes approaches that typically use nonexperimental methods, such as edmography or case studies, to investigate important variables that are not easily manipulated or controlled and which emphasize the use of multiple methods for collecting, recording, and analyzing data (Seels & Richey, 1994). Although case histories have been used frequently in television research, ethnographic studies are becoming more common. The trend towards qualitative research emerged after new research questions began to be asked about the mediating effect of the home context for television viewing (Leichter et al., 1985). Often with qualitative research, the purpose is hypothesis generating rather than hypothesis testing. Unlike survey methodology, qualitative research cannot present a broad picture because it concentrates on single subjects or groups, although longitudinal studies can describe how groups, or individuals change over time. There is no attempt at representative sampling as in survey research. Examples of case studies abound in literature on early ITV and ETV projects. Ethnographic studies have been conducted by photographing or videotaping the home environment, which mediates television viewing (Allen, 1965; Lewis, 1993). An example of a recent ethnographic study on leaming from television is the Ghostwriter study conducted by CTW (Children's Television Workshop' October 1994). Ghostwriter is an after-school literacy program that encompasses a mix of media including television and utilizes outreach programs with community organizations. Edmographic techniques were used to gather data on wide variations in observed phenomena in disparate settings. For example, case studies were done at Boys' and Girls' Clubs in Los Angeles and Indianapolis and at Bethune Family Learning Circle in Baltimore.

11.2.5.3. Descriptive Methodology. Studies in this category (see Chapter 41) include survey research such as demographic, cross-cultural, and longitudinal, in addition to content and meta-analyses. The common denominator among such studies is the use of survey techniques for the purpose of reporting characteristics of populations or samples.

Survey research uses samples of group populations to study sociological and psychological variables. To do this, data can be collected by personal or telephone interview, questionnaires, panels, and structured observation. Demographic research uses facts and figures collected by others, such as the census bureau or television information offices. Cross-cultural studies based on surveys use factual data about groups to draw generalizations.

There are many longitudinal and cross-sectional studies in the body of literature on learning from television Sometimes these are based on qualitative research, sometimes on quantitative research, and sometimes on both. The longitudinal method can reveal links between earlier and later behavior and changes in individuals over time, but the changes may be the result of many factors not just developmental maturation. Cross-sectional studies can demonstrate age differences in behavior by observing people of different ages at one point in time. They provide information about change over time in cohort groups but not change in individuals. A sequential method combines the cross-sectional and longitudinal approaches by observing different groups on multiple occasions. Obviously, the more variables are controlled in each of these methods, the more reliably results can be interpreted. If there is not sufficient control of variables, the results from a cross-sectional study can conflict with the results of a longitudinal study. The longitudinal method is more extensively used, perhaps because it is easier and less expensive.

Parallel longitudinal studies in Australia, Finland, Israel, Poland, and the United States (Heusman & Eron, 1986, cited in Huston et al., 1992) revealed a pattern of involvement with violence related to amount of television viewing. The amount of violence viewed at age 8 predicted aggression at age 18 and serious criminal behavior at age 30. Because this was a relational study, however, it could not be determined whether more violence was viewed because of the viewer's personality or whether violent programming affected the viewer through desensitization or some other mechanism (Eron, 1982; Huesmann, Eron, Lefkowitz & Walder, 1984, cited in Huston et al., 1992). Milesky, Kessler, Stripp, and Reubens (1982) conducted a similar study and concluded that other research did not support the hypothesis. On the other hand, methodology experts who examined other studies supported the hypothesis on violence and aggression (Cook, Kendzencky & Thomas, 1983, cited in Huston et al., 1992).

Content analyses are used to determine variables such as (1) the number of violent, anti- or prosocial incidents in a program; (2) characteristics of roles given ethnic groups, gender, age, or occupations portrayed; and (3) values presented on television, such as in commercials. Meta-analyses, which use statistical techniques for synthesis of the literature, and integrated research studies, which use comprehensive surveys and graphic comparison of the literature, are used to draw conclusions from multiple studies on a research question.

11.2.5.4. Developmental MeChodology. Formative evaluation as a research methodology (see Chapter 42) developed in response to a need for procedures to systematically try out and revise materials during a product development process (Cambre, 1987). It is one of the major contributions of television research. According to Flagg (1990), "The goal of formative evaluation is to inform the decision-making process during the design, production, and implementation stages of an educational program with the purpose of improving the program" (p. 241). The techniques used in formative evaluation of television programs are important areas of competency for instructional technologists. Formative evaluation studies pose research questions rather than hypotheses, and techniques employed range from oral reports and videotaping reactions to short questionnaires. Evaluation models incorporate phases, such as pre- and postproduction, in the research process.

An example of formative evaluation studies on television is the AIT report on the development of a lesson in the form of a program entitled "Taxes Influence Behavior" (Agency for Instructional Television, 1984). Students were questioned about attention to the program, interest, story believability, character perceptions, storyline comprehension, and program objectives. Teachers were asked about the program's appeal, curriculum fit, objectives, and utilization. Revisions and recommendations for teachers were based on the data collected.

It was Children's Television Workshop (CTW) that pioneered techniques for formative and summative evaluation (Flagg, 1990). After specifying message design variables and then investigating the effect of these variables on psychological phenomena such as attention, CTW developed techniques for investigating relationships formatively, so that designs could be changed, and summatively, so that effects on behavior could be reported. In doing so, CTW forever put to rest the assumption that one style of television is best for all young children (Lesser, 1974) and the assumption that television was not an interactive enough medium to teach intellectual skills to young children.

Periodic bibliographies issued by CTW document not only the research done there but also research related to CTW productions. Sammur (1990) developed a "Selected Bibliography of Research on Programming at the Children's Television Workshop" that annotated 36 formative, summative, and theoretical research studies on the four educational children's television series produced by CTW. The CTW research program reflects the systematic application of design, development, and evaluation procedures that is necessitated by the expense of producing for sophisticated educational technologies.

11.2.6 Summary

Film research during the 1950s contributed an identification of variables, especially variables related to message design. However, much of this research was methodologically flawed. Therefore, today it is useful primarily for the model it set for television research and the variables it identified. During the 1960s, television research emphasized comparative studies and frequently focused on message design variables.

The 1970s were a period of transition in that there was a move from ITV to ETV research, and a move from comparative studies to the study of specific variables and effects via the aptitude treatment interaction paradigm. There was also a methodological shift to qualitative, descriptive, and developmental studies in addition to traditional empirical studies.

During the 1980s and 90s variables began to be categorized.into a viewing system consisting of programming,

environment, and behavior, all of which interrelated. We turn now from a chronological consideration of the historical context of film and television research to findings in the major areas of interest to researchers.