|
Has Lead-in Lost Its Punch?
A Comparison of Prime Time Ratings Inheritance Effects
Between 1992 and 2002
The business of commercial television is the selling of audiences to
advertisers, which often translates into the buying and selling of rating
points. Television programmers have been long been aware of the capacity of
a program to "inherit" sizable ratings from the program scheduled
immediately before it. Although intervening variables, such as program
genre, lead-out, and daypart, have been shown to have some minor influence
on this phenomenon, by far the most powerful predictor of program ratings
has been the mere size of a program's lead-in audience. This was reaffirmed
empirically many times during the 1970s and 1980s. However, in recent years
conventional broadcast television, represented by "the Big Four" of ABC,
CBS, NBC and Fox, has experienced so much audience upheaval, it seems
plausible to question the potency of lead-in scheduling strategies. Given
the circumstantial evidence of plummeting ratings coinciding with
ever-increasing program competition from cable, satellite, and other
alternative media over the past decade, one might suspect that audiences
today are more discriminating and therefore, less susceptible to the tuning
inertia of inheritance effects.
The purpose of this study was to ascertain whether lead-in programming has
lost it punch in terms of influencing prime time audience ratings. A study
comparing Nielsen prime time household ratings of 1992 with 2002 was
conducted to answer this question. To date, there have been no published
studies offering this type of annual ratings comparison. In addition to
adding to the existing body of work on inheritance effects, this study
raises some provocative theoretical concerns about program scheduling
practices and audience behavior in a multi-channel environment.
Literature Review
Inheritance Effects
The overall ratings impact of lead-in programming has been confirmed
myriad times by industry and academic researchers. Beginning in 1975,
Goddhart, Ehrenberg, and Collins coined the term inheritance effects while
working on the broader issue of audience duplication among programs. They
discovered a highly predictable flow of audience between adjacent programs.
Headen, Klompmaker, and Rust (1979) proposed a more sophisticated model
introducing several independent variables including ratings, channel,
program type, daypart, and repeat viewing. Using Simmons Market Research
data, an examination of over 4,000 combinations of pairs of programs
revealed that by a substantial margin, ratings were the single best
predictor variable. A different model offered by Webster (1985) introduced
factors of audience availability, lead-in program ratings, the number of
program options, and program content. Using Arbitron ratings from one sweep
period , Webster concluded that for adjacent program pairs, lead-in ratings
and the number of program options in combination explained 80% of the
variance. A massive 22-year study of network prime time programming from
1963-1985 conducted by Tiedge and Ksobiech (1986) concluded that programs
with high-ranked lead-ins scored higher share points than those with low
ranked lead-ins. Also, fewer program options produced higher lead-in
correlations and visa versa. In 1988, the same research team using the
identical ratings data set concluded that the "pull" effects of lead-out
were minimal compared to the stronger "push" effects of lead-in programs
(Tiedge & Ksobiech, 1988). Looking at nine years of Nielsen ratings from
1976 to 1985, Walker (1988) found that the correlational relationships
among inheritance effects, lead-in, program type, and number of options
supported the earlier findings of Tiedge and Ksobiech (1986). Boemer (1987)
found in one television market high positive correlations between audience
ratings of local late evening newscasts and their respective prime time
lead-ins. Davis and Walker (1990) discovered that the most effective way to
compete in prime time against new media (cable and satellite) was to take
advantage of lead-in effects. Examining syndicated rather than network
programs, Cooper (1993) correlated the influence of several variables on
program ratings including lead-in, lead-out, number of options, program
type compatibility, network affiliation, and cable penetration. The results
from a 50-market analysis revealed that lead-in ratings completely
overwhelmed any other factor in the model. A fairly consistent conclusion
found among most but not all of these early studies that included number of
program options as a variable was that as the number of options increased,
the correlations between lead-in and lead-out programs (i.e. inheritance
effects) weakened. A more detailed examination of the definitional problems
surrounding the term program options is presented in the discussion section
of this study.
During the 1990s and beyond, inheritance effects continued to be
analyzed within the context of testing other variables. For example,
McDowell and Sutherland (2000) discovered in a single market case study of
local newscasts that audience-based brand equity of the news program could
affect the relative influence of lead-in ratings. From an advertising
perspective Napoli (2001) found that at the beginning of a new fall premier
season, the ratings of returning lead-in programs can assist network sales
departments in reducing the degree of error in forecasting the ratings for
new prime time programs. Although the above mentioned studies and several
more all included the notion of inheritance effects, none looked back to
see if this well recognized audience phenomenon has changed over time.
The Art and Science of Scheduling
Savvy television programmers will concede that the ratings performance of
many supposedly successful programs is more a matter of clever scheduling
than compelling content. When analyzing a program's ratings performance
Webster, Phalen and Lichty (2000) warn that
Some people assume the choice of a program centers upon the active
expression of a preference for a program or type of program. However, so
called structural factors have traditionally been considered important
mediators of the programs viewers choose and complicate the relationship
between viewing preference and viewing behavior". (p. 178)
Over the years these structural factors have acquired their own special
jargon as outlined by Eastman & Ferguson (2000). For example, placing a
relatively weak or unfamiliar program between two strong programs is called
"hammocking". This is a common strategy used to stimulate sampling of a new
program. Inserting a strong program between two weaker entries has been
dubbed "tent-poling" and is often associated with the notion of salvaging a
poor program line-up. Offering several adjacent programs with highly
similar content, such as an evening of sitcoms, is called "block
programming." The strategy of responding to a competitor with radically
different program content is known as "counter-programming." Eastman,
Newton, Riggs and Neal-Lunsford (1997) analyzed ways the major networks
capitalized on inheritance effects and enhanced audience flow by
positioning commercial breaks away from the natural transitions between
programs.
Because most networks and large-market stations negotiate commercial rates
with advertisers using Cost-Per-Thousand (CPM) or Cost-Per-Point
(CPP) measures of efficiency, the added ratings gained or lost from
program scheduling can amount to subsequent gains or losses in revenue
(Surmanek, J. (1996). All of the above-mentioned scheduling techniques
share a common strategic thread in that they attempt to take advantage of
the power of lead-in and although this strategy is still popular, there
have been no nonproprietary longitudinal studies to see if this strategy
has lost some of its potency. A plausible reason for alleging such a
decline is the unprecedented increase over the past decade in the number of
program options available to audiences.
Program Options Explode During the 1990s
During the 1970s and 1980s, when most of the above mentioned inheritance
studies were conducted, the media landscape remained relatively constant.
For over thirty years America was serviced by a three-network oligopoly
(Long, 1979). In 1987, Fox became a feisty competitor to the "big Three"
but did not become a significant force until the mid 1990s when the network
acquired the broadcast rights for NFL football and began to persuade
established VHF stations to switch network affiliations (Litman, 1998;
Block, 1990). Later, upstart networks WB and UPN and most recently Pax have
chipped away audiences from the larger incumbent networks. Corresponding
with this increase in broadcast networks was widespread dissemination of
remote control tuning devices, which enhance greatly the physical ease of
changing channels. Also, it should be noted that during this period, the
amount of time dedicated to watching television by the typical American
household actually increased. Despite much talk and speculation about the
potential distractions coming from internet usage, American households in
2001 watched an averaged over 53 hours of television per week, a
significant increase over the 48 hours of household watching recorded in
1990 (Nielsen Report, 2001).
The decade of the 1990s witnessed not only a dramatic increase in the
number of broadcast networks but also the number of cable/ satellite
networks. According to Nielsen Media Research, channel availability for
the typical American home (cable and non cable combined) surged from 33.2
channels in 1990 to 89.2 channels in 2001, a three hundred percent increase
(Nielsen Report 2001). In December 2002, for the first time in history,
Nielsen claimed that households watched more programming emanating from
cable networks than from broadcast networks (Romano, 2002).
During this same time span, the temptation to switch channels was
encouraged by a significant increase in advertising clutter. Ching and Lee
(2001) found that in addition to expected "surfing" among channels seeking
content, there was the deliberate avoidance of commercials, often referred
to as zapping, during prime time viewing. Given these historic shifts in
audience behavior, there is cause to take a second look at the
psychological disposition of audiences.
Audience Disposition
A primary assumption of inheritance effects studies has been that there are
significant numbers of passive or uncommitted viewers who are not motivated
to change channels (Webster, Phalen and Lichty, 2000) . The result is what
some researchers call tuning inertia, whereby the audience disposition is
to remain on the same channel unless there is a sufficient external force
that alters the mindless momentum (Cooper, 1996). This is not a new
concept. Rubin (1984) maintained that the simple act of watching
television, regardless of the specific content, can become a daily
ritualistic behavior. Some industry observers have coined the phrase "glow
and Flow", referring to the idea that programs are of secondary importance
as long as something fills the screen (Head, Spann and McGregor, 2001).
Furthermore, it is no secret that mere habit is a powerful force that often
supercedes other motivations for seeking alternative program content
(Rosenstein, 1997). None-the-less, given (a) the staggering amount of
program options available today (b) the significant drop in broadcast
network audiences and (c) the fact that people knowingly subscribe to
multi-channel services, one could postulate that audiences have come out of
their stupor and are more attentive to satisfying their viewing desires.
A second look at Inheritance Effects
The underlying rationale for this hypothesis is that subscribing to cable
or a satellite service is a deliberate act impacting people's personal
finances every month. Therefore, audiences should have a heightened
awareness of these new choices and should be motivated to investigate these
viewing options. In terms of published studies, the notion of heightened
awareness resulting in more deliberate channel changing is perplexing. For
instance Heeter (1995) found that channel changing was a sign of greater
selectivity and reevaluation of programs. However, Perse (1990) concluded
that channel changing reflected less attentive use of television.
Regardless of these inconsistent findings from early academic research,
the researchers for this study were convinced that the mass defection of
broadcast network audiences to cable over the past decade implied a certain
audience selectivity. That is, instead of passive ritualistic viewing
behavior, audiences can also take part in what Rubin (1987) calls
instrumental behavior, characterized by viewing that is planned and
attentive. With so much obvious choice, revealed through ubiquitous TV
program listings and advertising, coupled with the conscious decision to
subscribe to a multi-channel service, it seems plausible to presume that
audiences have become more discriminating and as a result, the ratings
power of lead-in programming has lost some of its punch. Given the changes
in the media market, inheritance effects should not be as strong today as
they were in the past. This leads naturally to a working research hypothesis:
H1: Inheritance effects were not as strong in 2002 as they were in 1992.
Methodology
For this study, "inheritance effects" were operationalized as the ability
of a program to retain audience ratings from the program scheduled
immediately prior. The sample frame was prime-time network ratings as
reported by Broadcasting and Cable and Electronic Media magazines
throughout 1992 and 2002.
Coders selected ABC, CBS, NBC and Fox programs where ratings were available
for the program immediately prior. Effectively, the first program was
skipped and coding started with the second program of the evening. For each
selected program, coders recorded 1.) target program share, and 2.) prior
program share. Coders attempted to use all fifty-two weeks of data.
However, ratings data were not available for seven weeks in 2002. These
weeks were skipped – making the data set for 1992 somewhat larger. Data
sets for each year were kept separate until the analysis began. At that
time, a dummy variable for year was added. Audience shares rather than
ratings were selected as the unit of analysis because shares are a function
of HUT (Homes Using Television) levels at a specific time and therefore,
offer a more standardized measure of program-to-program performance over
time.
The analysis stage presented researchers with a problem not found in
previous inheritance studies. Typically, prior studies have looked at
inheritance in context of several other variables. Given overwhelming
support for inheritance in prior studies, this study simply compares the
inheritance effect in one time period to the same effect in another time
period. Prior studies looked at the correlation (Pearson's r) between past
and target program rating. More advanced studies went on to use regression
analysis to compare inheritance to other possible predictors of audience
size. This study compared the correlation between past and current programs
in the two time periods. A regression analysis was used to test the
difference.
The basic question was whether the two regression lines were the same or if
the effect had changed. According to Gujarati (1988) there are two main
methods for comparing regressions. The Chow (1960) test is sensitive to
heteroscedasticity1 (common in rating data) and produces errors. A Park
test (as described in Gujarati, 1988) was run and confirmed the presence of
heteroskidasticity so the Chow test was not used.
The second method uses a dummy variable approach from Gujarati (1970). In
this approach, observations from both regressions (1992 and 2002) are
pooled into a single regression.
Yi = a1 + a2 Di + b1 X1 + b2 (Di Xi) + ?I
The above equation starts with the standard regression equation including a
dependant variable (Yi), the independent variable slope (b1 X1), intercept
(a1), and error term (ui). The second regression line was tested with the
additional variables a2Di for the intercept and b2(Di Xi) for the slope --
where Di was a dummy variable. In this case, "year" is entered as a dummy
variable (Di) with 1992 = 0 and 2002 = 1. The second term Di Xi is
computed by multiplying the dummy by the dependent variable. If the
measures for this second line are significant, then the two regression
lines are significantly different and the two lines can be determined from
the final equation. If the measures are not significant, then the null
hypothesis (no significant difference) can be accepted and one regression
line exists. The advantage of this method is that both regression lines can
be computed from the equation (discussed below).
There was one additional challenge to the project before continuing with
the analysis. As described in the literature review, there have been some
dramatic changes in the television market and these changes were reflected
in the data set. The average program share in 1992 was 18.0 compared to
10.9 in 2002. An ANOVA was performed to confirm that the two data sets were
significantly different (F = 2133.2, p > 0.001). In Figure One below you
can clearly see the difference between the two histograms. Not only was the
mode clearly shifted but also the curve for 2002 was more skewed than in
1992. Left uncorrected, the regression may show significance not because of
a difference in inheritance but because of other differences between the
years.
Figure One. Program Share Comparison Between 1992 and 2002
[--- ??? Graphic Goes Here ---]
In order to effectively compare the two years, the share values were
standardized for year. Standardization is the process of converting data to
the same scale by subtracting the sample mean ad dividing by the standard
deviation (Malhotra,1993). Standardization does not change the correlation
between variables but simply makes the mean equal to zero and the standard
deviation equal to one. In this study, "share" and "previous share" were
standardized by year. Once each data set (1992 and 2002), were standardized
the analysis could proceed.
Results
The data collection resulted in a very large data set. As summarized in
Table One, data was collected from 3050 programs in 1992 and 2541 in
2002. There was a shift of about seven share points difference between
1992 and 2002. Not only did overall shares drop from 18 to 10.9 but also
the maximum and minimum shares dropped about the same amount. The
correlation between target and previous program (Pearson's r for share) for
1992 was 0.618 (2-tailed significance < 0.001) and in 2002 was 0.664
(2-tailed significance < 0.001). This first level of the analysis shows a
strong and similar effect of inheritance despite a drop in overall share.
Table One. Descriptive Statistics
Year
N
Minimum
Maximum
Mean
Std. Deviation
1992
Share
3050
5
50
18.0
5.9
2002
Share
2541
1
42
10.9
4.6
A single regression analysis was sufficient to test the hypothesis. Table
Two displays the result of the regression analysis. The overall regression
equation had a reasonably strong adjusted R-square of 0.409 and the F
(1289.7) was highly significant (> 0.001). Now looking at the individual
variables in the equation, all variables in the equation were significant.
The constant (intercept) and "previous program share" were both highly
significant (> 0.001). The "dummy variable for year" and the "dummy times
previous program share" were both acceptably significant at the 0.05
level. A likely level of autocorrelation between these variables probably
reduced some of the significance. The first level of analysis was that the
regression equation supports the importance of inheritance effect and a
change in inheritance effect from 1992 to 2002. As a result, the null
hypothesis (no effect) was rejected but that is not the end of the story.
Table Two. Regression Output
Model Summary(b)
R
R Square
Adjusted R Square
Std. Error of the Estimate
.640(a)
.409
.409
.768
a Predictors: (Constant), Dummy for Year, Stand Prev. Share, Dummy X Prev.
b Dependent Variable: Standardized Share
ANOVA(b)
Sum of Squares
df
Mean Square
F
Sig.
Regression
2286.8
3
762.3
1289.7
.000(a)
Residual
3302.2
5587
0.6
Total
5589.0
5590
a Predictors: (Constant), Dummy for Year, Stand Prev. Share, Dummy X Prev.
b Dependent Variable: Standardized Share
Coefficients(a)
Unstandardized Coefficients
Standardized Coefficients
t
Sig.
B
Std. Error
Beta
(Constant)
3.816
.140
27.3
.000
Stand Prev. Share
.618
.014
.618
44.4
.000
Dummy X Prev.
.046
.021
.231
2.2
.026
Dummy for Year
-.459
.208
-.229
-2.2
.027
a Dependent Variable: Standardized Share
Figure Two graphically displays the predicted regression lines based on raw
(not standardized) data. The adjusted R-square for raw data regressions in
1992 was 0.382 and in 2002 was 0.441 (intercept and slope significant >
0.001). The lines seem similar but what was important was that the slope
of the line for 2002 was greater than 1992. This effect is exactly
opposite of what was predicted by the operational hypothesis. If the
operational hypothesis was supported, the line for 2002 would not cross the
line for 1992 and would have a more gentle slope. This means that the data
not only fails to support the null but also the operational
hypothesis. The data supports the conclusion that inheritance effect
actually increased from 1992 to 2002. While the increase was modest, it
was opposite what was predicted from the literature review.
Figure Two. Predicted Regression Lines by Year
Turning to the interpretation of the regression line summarized in Table
Two. The regression equation predicts the following.
Y = 3.82 – 0.46 Di + 0.62 Xi + 0.05 DiXi
The resulting regression equations per year would then be:
1992 Y = 3.82 + 0.62 X
2002 Y = (3.82 – 0.46) + (0.62 + 0.05) X
= 3.36 + 0.67 X
Again, the increase is statistically significant although it is
modest. What was important was the direction of the change. At the very
least, this data supports the conclusion that inheritance effect is holding
its own and quite possibly becoming more important.
Discussion
Based on the results of this study, one can conclude that despite a decade
of plummeting ratings and ever-increasing competition from other media, the
power of lead-in among the four major broadcast networks appears not to
have lost its punch. Indeed, the data suggest a modest increase in the
drawing power of lead-in programming. As revealed in the hypothesis, these
findings were not expected and open the door to theoretical speculation and
future research.
There were some obvious limitations to this study that need to be
disclosed. First, only two years were selected for comparison. A more
ambitious study might have examined each year within the decade, recording
any annual deviations from the presumed long-term stability of inheritance
effects found in this study.
A second limitation was the scope of the study in that it was restricted to
the major broadcast networks and not to cable or individual local markets.
In the same manner, it would be useful to study the magnitude of
inheritance effects on daytime "strip" programming in addition to prime
time offerings.
Another limitation was that the study dealt with simple audience ratings,
which do not reveal the complexities of audience flow between programs.
That is, a program does not necessarily "inherit" the entire audience from
its lead-in. Instead, some audiences may arrive from other channels or from
tune-in households that were not watching television at all. (For a
substantial fee, Nielsen provides such information for client subscribers).
Therefore, there is a certain "leap of faith" whereby the calculated
regression figures are intended to reflect the migration of the same
viewers from one adjacent program to another.
Assuming that regardless of the availability of dozens of new program
options, a majority of viewers are not motivated to change channels, there
remains the nagging question of explaining why this tuning inertia is still
so strong. As mentioned earlier, the researchers assume multi-channel
subscribers are sufficiently aware of non-broadcast channels. Except for
alleviating over-the-air reception problems, why else would people pay
money to watch television? This assumption is substantiated by the severe
drop in broadcast network ratings anfd a corresponding jump in cable
viewing between 1992 and 2002. Obviously many audiences have defected to
cable and other alternative media, but once they arrive at a particular
channel, perhaps the power of lead-in takes hold in the same fashion as it
does for the broadcast networks. A study of lead-in programming from a
broad sampling of cable networks would shed more light on this question.
Scrutinizing the literature review a second time may uncover some
misunderstandings, in particular the notion of available channels. For
studies conducted in the 1970s ands 1980s, program options were minimal,
including no more than a handful of channels. Many researchers defined the
number of options at a specific time as programs that were not already in
progress. For example, at 8:30 PM, if two out of four programs airing were
one hour in length and began a half hour earlier at 8:00PM, the researchers
would arbitrarily reduce the "number of options" to two. This reduced
number of "options" ,in turn, resulted in an observed increase in
inheritance effects and visa versa. This conceptualization is a far cry
from more contemporary definitions of available channels, where all
channels are accrued at a specific time, regardless of the juxtaposition of
the duration of the program.
A clue to understanding the underlying psychology of inheritance effects in
a multi-channel environment may be found in some additional data provided
by Nielsen, which reports that, while the number of available channels has
nearly tripled in recent years, there has been only a modest increase in
the number of channels actually viewed. According to Nielsen, the typical
American household today has access to almost 100 channels, yet the average
number of channels actually viewed is only 14. Looking back to 1994, which
provided less than half the channel availability of 2001, the average
number of channels viewed per household was 10. Obviously, we have a
classic example of the law of diminishing returns, where more choice does
not translate directly into more channels viewed. Some researchers, such as
Ferguson and Perse (1993), refer to this preferred subset as a viewer's
channel repertoire. This means that regardless of a three-fold increase in
the number of available channels, the major broadcast networks compete
ultimately in a much smaller arena of only a dozen or so heavily trafficked
channels. Furthermore, in terms of the proportion of channels viewed ,
there is circumstantial evidence of increased viewer loyalty. Where in
1994, audiences watched about 30 percent of available channels (10 out of
33), by 2001, audiences were watching less than sixteen percent of
available channels (14 out of 89). Admittedly, there is substantial channel
switching going on, but this frenetic activity appears to settle into a
predictably small repertoire of acceptable channels of which the "Big 4"
networks are usually included. Furthermore, within this repertoire, these
broadcast networks still manage to capture roughly half of all viewing.
In terms of measuring audience behavior, it should be noted that Nielsen
methodology attempts to discard "uncommitted" viewing by enforcing
five-minute minimum viewing thresholds before a station or network is given
average quarter-hour (AQH) viewing credit in a published report.
After pondering these data, perhaps the durability of inheritance effects
is not as surprising as the researchers initially thought. As many
restaurant owners will attest to, a huge menu does not necessarily mean
that customers will take equal advantage of all items available. Typically,
there will be a relatively small core group of meals that account for the
most of the restaurant's business. Similarly, we propose that a vast menu
of TV channel options does not produce substantial fragmentation in
viewing. Audiences may "surf", "graze" or "zap" among many channels but
ultimately, they migrate back to a familiar set of a dozen well-used
channels. Perhaps the art and science of media branding can offer some
theoretical insight into this behavior. In conclusion, the phenomenon known
as inheritance effects appears to remain a potent force in contemporary
multi-channel television and deserves continued attention and research.
References
Boemer. M. L. (1987). Correlation lead-in show ratings with local television
news ratings. Journal of Broadcasting and Electronic Media, 31, 89-94.
Block, A. B. (1990). Outfoxed: marvin davis, barry diller, Rupert Murdoch,
joan rivers and the inside story of america's fourth network. New York: St.
Martin
Ching Biu Tse, A., Lee, R. P. (2001). Zapping behavior during commercial
breaks. Journal of Advertising Research, (May/June), 25-29.
Chow, G.C., (1960), Tests of equality between sets of
coefficients in two linear regressions," Econometrica, vol. 28, no. 3.
Cooper, R. (1996). The status and future of audience duplication research:
an assessment of ratings-based theories of audience behavior. Journal of
Broadcasting and Electronic Media, 40, 96 -111.
Cooper, R. (1993). An expanded, integrated model for determining audience
exposure to television. Journal of Broadcasting and Electronic Media, 37,
401-418.
Davis, D. M.; Walker, J.R. (1990). Countering the new media. The
resurgence of share maintenance in prime time network television. Journal
of Broadcasting and Electronic Media, 34, 487-493.
Eastman, S. T., Ferguson, D. A. (2000). Broadcast/cable Programming:
strategies and practices. Belmont, CA: Wadsworth.
Eastman, S. T., Newton, G.D., Riggs, K.E. Neal-Lunsford, J. (1997).
Accelerating the flow: A transition effect in programming? Journal of
Broadcasting and Electronic Media, 41, 305-323.
Ferguson, D. A.; Perse, E. M. (1993). Media and audience influences on
channel repertoire. Journal of Broadcasting and Electronic Media, 37, 31-48.
Ferguson, D. A. (1992). Predicting I heritance effects from VRC and cable
penetration. Dowden Center Journal, 1, 28-40.
Goddhart, G. J., Ehrenberg, A. S. C, & Collins, M. A. (1975). The
television audience: Patterns of viewing. Westmead, UK: Saxon House.
Gujarati, D. M., (1970), Use of Dummy Variables in Testing for Equality
Between Sets of Coefficients, American Statistician, vol 24, no. 1, pp. 50-52.
Gujarati, D. M., (1988), Basic Econometrics (2nd edition), McGraw-Hill: New
York.
Head, S.W., Spann, T., McGregor, M.A. (2001). Broadcasting in america
(pp 336-337). Boston: Houghton Mifflin.
Headen, R. S., Klompmaker, J. E., & Rust, R. T. (1979). The duplication of
viewing law and television media schedule evaluation. Journal of Marketing
Research, 16, 333-340.
Heeter, C. (1985). Program selection with abundance of choice. A process
model. Human Communications Research, 12, 126-152.
Litman, B. R. (1998). The Economics of television networks in A.
Alexander, J. Owers, R. Carveth, Media economics: theory and practice, (pp.
131-150). Mahwah, NJ.
Long, S. L. (1979). The Development of the television network oligopoly.
New York: Arno Press.
McDowell, W. Sutherland, j. (2000). Choice vs. chance: Using brand
equity theory to explain TV audience lead-in effects. Journal of Media
Economics.
Malhotra, N.K., (1993), Marketing Research An Applied Orientation, (2nd
Edition), Prentice Hall: Upper Saddle River, NJ.
Napoli, P. M. (2001). The unpredictable audience: An exploratory analysis
of forecasting error for new prime time network television programs.
Journal of Advertising, 30, 53-60
Nielsen Report (2001). 2001 Report on Television. Nielsen Media Research.
New York: Author.
Perse, E.M (1990). Audience selectivity and involvement in the newer media
environment. Communication Research, 17, 675-697.
Romano, A. (2002, December 30). Cable's big piece of the pie.
Broadcasting and Cable. p. 8.
Rosenstein, A. W., & Grant, A. E. (1997). Reconcepualizing the role of
habit: A new model of television audience activity. Journal of Broadcasting
and Electronic Media, 41, 324-344.
Rubin, A. M. (1984). Ritualized and instrumental television viewing.
Journal of Communication, 34, 67-77.
Surmanek, J. (1996). Media planning. A practical guide. New York: NTC
Tiedge. J. T., & Ksobiech. K. J. (1986). The "lead-in" strategy for
prime-time TV: Does it Increase the audience? Journal of Communication,
36(2), 51-63.
Tiedge. J. T., & Ksobiech. K. J. (1988). The sandwich programming
strategy: A case of audience flow. Journalism Quarterly, 65, 376-383.
Walker, J. R. (1988). Inheritance effects in the new media environment.
Journal of Broadcasting and Electronic Media, 4, 391-401.
Webster, J. G. (1985). Program audience duplication: A study of
Television in heritance effects. Journal of Broadcasting and Electronic
Media, 29, 121-133.
Webster, J. G., Phalen, P. F., Lichty, L. W. (2000). Ratings Analysis:
Theory and practice. Hillsdale, NJ. Lawrence Erlbaum.
Has
Lead-in Lost Its Punch
Has Lead-in Lost Its Punch?
A Comparison of Prime Time Ratings Inheritance Effects
Between 1992 and 2002
Walter S. McDowell, Ph.D.
School of Communication
University of Miami
Coral Gables, FL 33124
[log in to unmask]
305 – 284 – 5201
and
Steven J. Dick, Ph.D.
Southern Illinois University
College of Mass Communication and Media Arts
Carbondale, IL 62901
618 – 453 – 6980
[log in to unmask]
Submitted to AEJMC Convention 2003
Media Management and Economics Division
Abstract
Has Lead-in Lost Its Punch?
A Comparison of Prime Time Ratings Inheritance Effects
Between 1992 and 2002
For decades, the single best predictor of a television program's ratings
performance has been the supposed inheritance effects derived from the
ratings of the program leading into it. Recognizing the recent dramatic
increase in the number of channels available to the typical American
household coinciding with an equally dramatic decrease in audience ratings
for the major broadcast networks, there was reason to speculate that over
the past decade "couch potato" audiences have come out of their stupor and
become more discriminating and therefore, less susceptible to this
scheduling strategy. However, an analysis of prime time ratings comparing
1992 with 2002 for ABC, CBS, NBC and Fox showed no support for this notion.
In fact, findings revealed a modest increase in inheritance effects,
suggesting that, despite the recent upheavals in the television industry,
lead-in has not lost its punch.
|
