Perceptual Discrepancies in the Time Duration and Number of Television Commercials

Peter H. Webb, University of Illinois
ABSTRACT - In recent years, increasing pressure has been applied to television networks to control both the quality and quantity of advertising. The industry has denied major increases in commercial time in the last 25 years. This research shows that people's perceptions of the amount of advertising on television are substantially different than the actual amounts, and also differ as a function of how the commercials are scheduled within programs.
[ to cite ]:
Peter H. Webb (1979) ,"Perceptual Discrepancies in the Time Duration and Number of Television Commercials", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 85-89.

Advances in Consumer Research Volume 6, 1979      Pages 85-89


Peter H. Webb, University of Illinois


In recent years, increasing pressure has been applied to television networks to control both the quality and quantity of advertising. The industry has denied major increases in commercial time in the last 25 years. This research shows that people's perceptions of the amount of advertising on television are substantially different than the actual amounts, and also differ as a function of how the commercials are scheduled within programs.


In recent years, debate about the contents of television has steadily increased. With television reaching deeper and deeper into the lives of more and more Americans (the set in the average television household, which includes 96% of all households, is now on more than six hours a day), a growing number of public interest groups, government agencies, program sponsors, and viewers themselves are expressing concern over the quality of this mass medium. Much of this concern focuses on aspects of programming, such as the amount of sex and violence reaching younger audiences or the limited a mount of public service programming available to all types of audiences. However, non-program material, including commercials, promotional announcements, public service announcements, billboards, station breaks, and credits, has also received its share of attention. For example, cigarette commercials were banned from the air waves in 1971, the amount of advertising on Saturday morning children's programming was limited in 1974, and the Federal Trade Commission is currently contemplating further restrictions on advertising to children.

The sheer volume of non-program material is a problem of concern for several groups. "Clutter", as labeled by broadcasting practitioners, is a problem for advertisers, due to potential decreases in effectiveness of commercial messages in a crowded environment. It is a problem for TV broadcasters, with increasing operating costs on the one hand and pressures from advertising agencies and government to reduce revenue generating non-program material on the other hand. Finally, it is a problem for viewers, who presumably turn on the set for the programming it offers, but all too often face frequent and lengthy program interruptions for commercials, promotional announcements and other non-program elements.

Although concern that viewers are being swamped with increasing amounts of advertising on television (advertising accounts for more than 80% of all non-program material) has resulted in threats of involuntary regulation from government agencies and threats of boycotts from advertisers and advertising agencies, the television industry has steadfastly maintained that the amount of advertising has not increased appreciably in the past 25 years. This raises several issues.

The first issue concerns how one defines the "amount" of advertising. Does this mean the amount of time de voted to advertising, the number of commercial interruptions, or the total number of commercials? In 1952, the Code Authority of the National Association of Broadcasters, a self-regulating agency of the industry, set the maximum allowable range for advertising from seven minutes per hour for prime time network television to nine and one half minutes for non-prime time local television. Comparable figures in 1971 were ten minutes and sixteen minutes, increases of 43% and 68% respectively. [1978 allowances range from nine and a half to six teen minutes.]

With respect to commercial interruptions, restrictions on the maximum allowable number of interruptions by the Code Authority (a complex rule differing by time of day and type of program), has led to subtle changes in programming practices including the widespread use of "lead-ins" and "trailers", set apart from the main body of the program. The result is probably more interruptions than ever, although no longitudinal data is available.

Finally, although data do not extend as far back as the early days of television, results of an extensive study published in 1972 by the American Association of Advertising Agencies indicated a 50% increase in the number of commercials (and a 33% increase in the number of commercial minutes) over the eight year period from 1963 to 1971. Certainly one of the major contributions to the increase in the number of commercials has been the shift from the sixty second to the thirty second commercial.

Another important issue concerning the amount of advertising on television is viewer perceptions. Even if the actual amount of time devoted to advertising is neither "large" (at least relative to other media--another defense of current practices put forth by the industry) nor increasing substantially, what if viewers perceive this to be the case? Is this not a cause for concern? The viewer appears to have received little attention in this matter. Even in Steiner's 1960 land mark study, The People Look at Television (published in 1963), the one section of his lengthy questionnaire having to do with advertising focused almost entirely on content, with little attention on quantity. How ever, in an open-ended question concerning the irritating aspects of advertising, constant interruptions and frequent repetitions of the same commercials were common responses. A careful replication of this study ten years later showed similar, and in some cases even stronger results (Bower, 1973). An additional item in the replication study not included in Steiner's study revealed that 70% of the respondents felt television has too many commercials. Beyond these relatively sparse data, little exists to indicate exactly how people perceive the amount of television advertising (or other elements of non-program material).

As part of a larger study of clutter on television, data were gathered concerning just such perceptions. After exposure to videotaped television programming, respondents were asked a series of questions about the commercials including how many there were and how much time was devoted to them. The little that could be found in the literature to base any hypotheses on is presented in the next section. Following this, the specific studies are described and the results are presented.


One aspect of cognition that has received little attention in advertising research is perception of time duration. This is particularly relevant to the problem of clutter since the amount of time devoted to non-program material is one way of operationally defining clutter. As noted earlier, practitioners frequently argue over whether or not clutter is actually increasing, but rarely present evidence of viewers' perceptions. Burke Marketing Research, Inc. (1972) did report an increase over a two-year period in people's perceptions of the total amount of television advertising. In extensive telephone interviews, Burke found that 67% of a nation wide sample of 1750 respondents perceived more advertising on television in 1972 than in 1970. In addition, 50% of the respondents felt that there was more advertising than programs were worth, while just 3% felt the opposite. No precise data on how much advertising had increased were ascertained.

The only other research relating directly to time perceptions of advertising was a 1966 study of "piggyback" commercials by Martilla and Thompson. Piggybacking refers to the practice of allocating a commercial time slot to advertisements for two or more unrelated products. In this study, test commercials included 60 and 30 second versions of five commercials. For the 30 second version, a second 30 second commercial was "piggybacked'' either in front of or behind the test commercial. Commercials were embedded in one of two 30 minute programs, which respondents watched in a laboratory setting.

When questioned immediately following the presentations, on the average the 60 second test commercial was judged to be 67 seconds long (a control 60 second commercial in a different position in the program was estimated on the average as 64 seconds long). However, the two 30 second commercials appearing in the same position as the 60 received a combined estimate of 83 seconds-a 38% overestimate of the actual time compared with a 12% overestimate for the 60 second version.

Although this finding confirmed Martilla and Thompson's hypothesis that time perception of advertisements increases as the number of advertisements increases even if the actual total time remains constant, no explanation was offered for this difference. What factors might relate to consumer misperceptions of time?

Unfortunately, most behavioral research involving time treats it as a fixed parameter and not a dependent variable. But there are some studies in which time estimates have been measured as a function of varying physical situations or task characteristics (Arvedson, 1974). Two such studies suggest several dimensions relevant to clutter which may affect estimates of time duration.

Ornstein (1968) had subjects study a set of drawings of varying complexity. He then asked subjects to estimate how long they had been given to study the drawings. He found that the more complex the drawing the longer the time interval was perceived to be (in actuality, the interval was 30 seconds in every case). This study was replicated by Hogan (1975), using slightly different stimuli, but varying the same dimension. The pictures were presented for just fifteen seconds in his experiment. Unlike Ornstein, Hogan found a curvilinear relationship between complexity and perceived time duration. The lowest estimates were given for moderately complex stimuli. Higher estimates were made for the simpler and the more complex drawings.

Although these two studies show seemingly different results (it may be that Ornstein did not include stimuli simple enough to show the curvilinear relationship found by Hogan), they do suggest that a systematic relationship does exist between stimulus complexity and perceived duration of time. To the extent that continuity is related to complexity, a string of unrelated non-program elements is likely to be judged more complex than an integrated program segment. This may cause an overestimate of the time devoted to non-program material. Thus from a time duration standpoint, people may judge clutter to be worse than it actually is. This contention is supported by Martilla and Thompson's (1966) finding that two 30 second commercials were perceived as longer than a single 60 second commercial.

It should be noted that at the time this study was done in 1966, 30 second commercials were not nearly as dominant as they are today. It may be that the overestimate in duration for the two 30's was due to a variance between actual scheduling and viewers' expectations of scheduling. In fact, although there is no direct evidence, it may be speculated that viewer expectation lies at the heart of many reactions to clutter, including time perception. The overall hypothesis would be that the closer the match between actual and anticipated characteristics of an advertisement or a group of advertisements, the more accurate would be the perceived time duration. Variance from expectation of any kind may lead to distorted estimates of time perception.

Both Ornstein's and Hogan's studies indicate distortions are likely to be positive--that overestimates of time duration are far more common than underestimates. Thus variation in time perception may occur not only as a function of complexity, but also as a function of other variables such as relevance, affect, and novelty (both absolute and relative to surrounding material). For example, an advertisement seen as more or less relevant than expected would be perceived as longer than one whose degree of relevance matched expectations. The same phenomenon would occur for advertisements seen as more pleasant, more unpleasant, or about as pleasant as expected. Finally, changes in scheduling patterns for non-program material would create divergences from expectations that could alter perceptions of duration of elements comprising the schedules.


In the first study of television clutter, subjects were brought to a central location to view closed-circuit programs. The subjects were told that they were participating in a study of humor and violence on television. Participants saw two program segments (each approximately ten minutes in length) taped directly off the air at an earlier date: one from a prime time situation comedy, the other from a prime time crime drama series. Each segment included 30 seconds of lead-in from the beginning of the show and was edited in such a way that it could "stand by itself" as representing a complete story.

Placed between the two program segments was a simulated station break that contained various elements of clutter comprising the experimental treatments. Thus the videotaped elements employed in this study were edited and combined in such a way as to realistically represent the transition from one television program to another. Subjects saw either zero, four, or eight commercials, credits for the first program, and a station identification. All commercials were 30 seconds in length. Specifically, the two tapes representing "low" clutter conditions contained four commercials in two groups of two each separated by the program credits (30 seconds) and a five second station identification. The two tapes representing "high" clutter conditions contained eight commercials (including two public service announcements) in two groups of four each, again separated by the program credits and station ID. A fifth tape contained the credits and station ID, but no commercials at all.

A total of 192 subjects viewed the program, both men and women aged 18 to 55, from predominantly middle income families in suburban San Francisco. Each of the five tapes was viewed by eight groups of five subjects (eight of the forty groups were missing one subject). Subjects viewed the tapes in a room set up to look like a living room. They were instructed to act as they would watching TV at home, including talking with each other if they wished, getting coffee and cookies (placed at the back of the room), or browsing through magazines placed on a coffee table in front of the TV monitor.

Following the presentations, subjects filled out a questionnaire that included a series of questions about the programs (designed to be consistent with the cover story and to eliminate short term memory traces), questions about the contents of commercials, and questions about the number and duration of commercials. The latter questions were simply, "How many commercials were there?" and "How much time was devoted to commercials?"




Table 1 shows estimates of number and time devoted to commercials for high and low clutter conditions compared with actual amounts. These figures are raw estimates. Originally, it was thought that these estimates should be "normalized" by perceived length or presentations to eliminate errors in estimation due only to poor sense of time duration. However, 46 percent of the subjects estimated the presentations to be 30 minutes long. Since this was probably due to expectation rather than actual perception, the adjusted measure of commercial duration was dropped. As hypothesized, the time devoted to commercials was significantly overestimated in all conditions. In low clutter (four commercials), 81% of the subjects overestimated commercial time, compared with 65% in high clutter (eight commercials). The average amounts of the overestimates were 2.65 minutes for low clutter (t = 8.33, p < .001) and 1.75 minutes for high clutter (t = 4.96, p < .001). Even when no commercials were aired, the average subject perceived that the presentation contained almost a minute and a half of commercials. This is probably attributable to prior experience with television programming. Even when commercials cannot be recalled, people assume they were pre sent since this is the case with all programming except on public television.

The influence of prior experience is also evident from the degree of ever-estimation in high and low clutter conditions. When the amount of commercial time is in creased by a factor of two, the estimated amount of time increases by Just 26%. In the absence of prior experience, this percentage would probably be much greater.

With respect to number of commercials, estimates were surprisingly accurate. Subjects in the low clutter conditions (four commercials) perceived an average of 4.15 commercials. Subjects in the high clutter conditions (eight commercials) perceived an average of 7.21 commercials. In the baseline condition (zero commercials), subjects perceived an average of 1.05 (the only significant difference from actuality: t = 2.55, p < .01). Thus when few or no commercials were aired, subjects over-estimated the number, but when a large number of commercials were aired, subjects under-estimated the number. Once again, prior experience appears to have guided perceptions. Actual station breaks, such as the one simulated in this study, usually contain fewer than eight, but more than two or three commercials. It is interesting to note that deviations in perceived number of commercials are much smaller than those in amount of time devoted to commercials.

It is also interesting to note that subjects perceived the average commercial to be significantly longer than 30 seconds. Dividing their estimates of time devoted to commercials by estimates of the number of commercials shows an average perceived commercial duration of 49 seconds in high clutter, 67 seconds in low clutter, and 83 seconds in the baseline condition. This, of course, is an indirect estimate of commercial length, and might differ from explicit estimates of length which were not included in the questionnaire.


The second study of clutter focused on alternative scheduling patterns of commercials within programs. Once again, subjects were brought to a central location to view closed-circuit television programs. With a few minor alterations, the methodology employed was identical to that of the first study. Having been told they were participating in a study of television ratings, subjects viewed one of two half-hour network programs: a prime time situation comedy or a daytime serial. Twelve 30 second commercials were edited into the pro grams in one of three different schedules: Six breaks containing two commercials each (6 X 2), three breaks containing four commercials each (3 X 4); or two breaks containing six commercials each (2 X 6).

The 6 X 2 condition is typical of commercial scheduling patterns actually used for daytime serials. The 3 X 4 condition is a typical interruption pattern for prime time programming, although in this instance, the number of commercials was greater than would ordinarily be the case in prime time. This was due to the desire to hold the number of commercials and the time devoted to them constant in order to focus on differences due only to the scheduling patterns or the programs. The 2 X 6 condition is atypical of any current practices, but is interesting in view of previous findings that people would prefer a "bunching" of commercials into fewer interruptions (Burke, 1972).

A total of 189 subjects viewed the programs. As in the first study, the subjects were aged 18 to 55 from mostly middle income families in suburban San Francisco. In this study, subjects were all women. Nobody participated in both studies. Each of the six tapes (two programs; three commercial schedules) was viewed by eight groups of four subjects (three of the 48 groups were missing one subject). In all other respects, the study was conducted in exactly the same manner as the first study.


Table 2 shows estimates of number and time devoted to commercials for the different programming and commercial scheduling conditions. [As was the case in study one, a high percentage of subjects estimated the presentation to be 30 minutes long (48%). This was probably due more to expectation than actual perception. As a result, time perception data for commercials were not weighted by presentation length estimates as originally intended.] It is interesting to conditions, the amount of time devoted to commercials was significantly overestimated (by as much as 83%). However, the perceived number of commercials was substantially underestimated. In fact, only 25 percent of the subjects perceived as many as twelve commercials--the actual number. This was the same pattern found in the first study. The amount of time devoted to commercials was consistently overestimated as hypothesized. However, the number of commercials was underestimated in the high clutter conditions (7.21, compared with eight in actuality). It can be seen from Table 2 that as the actual number of commercials further increases, the degree of underestimation also increases (23%, com pared with 10% in the first study).



Based on the time duration findings of Ornstein and Hogan discussed earlier, it was hypothesized for this study that the perceived time devoted to commercials would increase as the number of interruptions increased. This hypothesis is based on the notion that increasing the complexity of stimuli within a fixed duration of time leads to increased estimates of that duration. It was here postulated that more interruptions produce a more complex stimulus "package" (program plus commercials). In addition, more interruptions are likely to be more annoying which also figures to increase the estimate of time duration.

From Table 2, the hypothesis is supported (from a one way analysis of variance, F = 4.05, 2 and 184 d.f., p < .02). The three interruption conditions were perceived as containing 5% more minutes of commercials than the two interruption conditions and the six interruption conditions 22% more. The estimated number of commercials does not reflect this pattern. The three break conditions show the minimum estimate, while the two break conditions actually show the maximum estimate, despite the minimum time estimate for these conditions. This may have occurred as a result of the two break conditions containing six consecutive commercials during each break, an unusually large number without at least a station break in between them.

It was also hypothesized that perceptions of commercial amount and duration would be higher for the prime time program, than for the daytime program. This was based on the fact that the number and duration of commercials included in the presentations was normal for daytime programming, but abnormally high for prime time programming. It can be seen from the table that the hypo thesis is supported, although the differences are very small and not significant.


These studies of television clutter, with specific focus on perceptions of amount of advertising, show that wide variations exist between actuality and perceptions. The first of the two studies showed significant overestimates in the amount of time devoted to commercials. The second study showed significant variations in commercial time estimates as a function of the number of commercial breaks and how the breaks are positioned within programs.

These findings present potential problems, both for advertisers and for the television industry. Voluntary (NAB TV code subscription) or even federal regulation of the amount of non-program material on television based solely on actuality may not properly reflect viewer reaction. Thus, for example, as commercials get shorter, or as the number of commercial interruptions increases, even if the actual amount of advertising re mains within established guidelines, viewers perceive such changes as representing substantial increases in the amount of advertising on television. This may produce negative feelings toward commercials and even specific brands involved. Alternatively, it may result in actual decreases in television audiences. Evidence that this may already be occurring comes from Nielsen ratings indicating an "unexplained" drop in several day time audiences in late 1977 (Advertising Age, 11/28/77).

Of a more positive nature, these studies suggest that negative effects on viewers or non-program material may be reduced by paying more attention to the viewer's perceptions of specific practices. Thus, for example, bunching of commercials into fewer interruptions may decrease the intrusiveness of the interruptions, resulting in more favorable reactions or even larger audiences. It should be cautioned, however, that such changes should be based on aspects of commercial response other than simply perceptions of number and duration. Traditional measures of response such as re call, attitudes and purchase intentions may reveal an inverse relationship between time perception as dealt with in these studies and commercial effectiveness.


"Another TV Mystery: Case of the Missing Young Women." Advertising Age, 48 (November 28, 1977): 6.

Lennart A. Arvedson, "Deadlines and Organizational Behavior: A Laboratory Investigation of the Effects on Individual Task Performance." Unpublished doctoral dissertation, Stamford University, 1974.

Robert T. Bower, Television and the Public. New York: Holt, Rinehart and Winston, 1973.

Burke Marketing Research, Inc. "Viewer Attitudes Toward Commercial Clutter on Television and Media Buying Implications.'' Presentation to the 18th ARF Conference, November 14, 1972.

H. Wayne Hogan, "Time Perception and Stimulus Preference as a Function of Stimulus Complexity." Journal of Personality and Social Psychology, 31 (January, 1975): 32-35.

John A. Martilla and Donald L. Thompson, "The Perceived Effects of Piggyback Television Commercials." Journal of Marketing Research, 3 (November, 1966): 365-71.

Robert E. Ornstein, "On the Experience of Duration." Unpublished doctoral dissertation, Stanford University, 1968.

Gary B. Steiner, The People Look at Television. New York: Alfred A. Knopf, 1963.