Content-Type: text/html

"57 Channels and Nothin' On"1:
 Electronic Monitoring of Television Remote Control Device Usage
in the Home Environment
 
 
Barbara K. Kaye, Ph.D.
 
and
 
Barry S. Sapolsky, Ph.D.
 
 
Submitted to the Mass Communication and Society Division of the
Association for Education in Journalism and Mass Communication,
1995.
 
 
 
 
 
 
 
Barbara K. Kaye is an Assistant Professor in the Department of Radio/TV at
 
           Southern Illinois University at Carbondale.  Barry S. Sapolsky is
Professor and
 
               Chair of the Department of Communication at Florida State
University.
 
 
 
Barbara K. Kaye                         Barry S. Sapolsky
1056 Communication Bldg.                        356 Diffenbaugh Building
Department of Radio/TV                  Department of Communication
Southern Illinois University                    Florida State University
Carbondale, Il 62901-6609               Tallahassee, FL 32306-2064
618-453-6068                            904-644-8774
<a href="/cgi-bin/wa?LOGON=A3%3Dind9603A%26L%3DAEJMC%26E%3D0%26P%3D243530%26B%3D--%26T%3Dtext%252Fplain" >[log in to unmask]</a>                           <a href="/cgi-bin/wa?LOGON=A3%3Dind9603A%26L%3DAEJMC%26E%3D0%26P%3D243530%26B%3D--%26T%3Dtext%252Fplain" >[log in to unmask]</a>
 
 
 
Running Head: 57 Channels
 
1 Bruce Springsteen, Human Touch, 1992
 
        57 Channels
"57 Channels and Nothin' On":
 Electronic Monitoring of Television Remote Control Device Usage
in the Home Environment
 
 
        New technologies have given rise to new media and new uses of existing media.
The remote
 control device (RCD) is just one of many new media technologies, such as the
VCR and
 
          cable television that has become immensely popular with viewing
audiences.  The RCD has
 
            allowed television viewers greater opportunity for selective
exposure, selective attention
 and control over their viewing habits.
        Of all U.S. households, 80% now have an RCD which either operates the
television set
 
          directly or through a VCR (Bellamy, 1993).  Further, it is estimated
that an RCD will be
 
            found in nearly 100% of all television households by the year 2000
(Klopfenstein, 1993).
 
            The diffusion of RCDs has had a great impact on television viewing
habits and program
 
          selection which is not only of academic interest, but of great concern
to the television
 
            and advertising industries.
        The purpose of this study was to electronically monitor viewers' television
remote
 
        control device behaviors in a naturalistic environment.  Actual RCD
activity based on
 
          electronic records of in-home use was examined in light of the theory
of selective exp
 
          osure.  Selective exposure research indicates that viewers choose to
watch programs that
 
            are consistent with their values and beliefs, thus avoiding
inconsistent information
 
         (Griffin, 1991). According to Bellamy & Walker (1990), selective
exposure is significantly
 related to reported RCD use and is a motivation for using an RCD.
         The RCD was originally designed to make television viewing easier and more
convenient.
 
            However, consumers have "reinvented" new uses for the RCD, leading
to changes in viewing
 
            behavior and selective exposure (Walker & Bellamy, 1992).  With the
availability of RCDs
 
            selectivity has become an almost effortless activity.  Viewers
merely have to push a
 
         button to be selectively exposed to whatever message will gratify their
needs. Studies
 
           have shown that RCDs along with cable television have changed
television viewing habits
 
            and program selection. RCDs facilitate the ease of channel changing,
leading to greater
 
            awareness of available programming and more program choices (Heeter,
1985; Heeter &
 
        Greenberg, 1985a; Selnow, 1989; Walker & Bellamy, 1989).
        Additionally, studies show that RCDs have influenced the television industry's
 
    programming strategies by reducing inheritance effects (audience carryover
from one
 
        program to the next on the same channel) and, along with cable
television, contribute to
 
            the reduction of network audience share (Heeter & Greenberg, 1985a;
Selnow, 1989; Walker,
 
            1988; Webster, 1986; Webster & Wakshlag, 1983).  When there were
only three major networks
 to choose from, viewers could easily become aware of program alternatives.
With the
 
          widespread availability of 60+ channel cable television, viewers'
choices have become
 
          greatly expanded and thus the process of selecting which programs to
watch is now more
 
           complex.  Cable television has fragmented the mass audience,
resulting in lost shares for
 
            the major broadcast networks. Channel loyalty has decreased among
broadcast channels as
 
            these lost shares appear to spread themselves widely across the
various options available
 
            on cable systems (Webster, 1986).  Alternatively, it can be argued
that with a large
 
         number of cable channels from which to choose, viewers may simplify
their choice processes
 by identifying favorites, thus increasing channel loyalty for both broadcast
and cable
 
            networks (Heeter & Greenberg, 1985a).
 
Measuring RCD Use
        Questionnaires and Telephone Surveys. The RCD allows for such rapid channel
changing that
 obtaining valid and reliable measures of grazing activity is a challenge.  To
date, the
 
            measurement of RCD use has been largely limited to self-report,
usually in the form of a
 
            questionnaire or telephone survey that asks respondents to recall
their viewing behaviors
 
            pertaining to RCD use (Ainslie, 1989; Bellamy & Walker, 1990;
Copeland, 1989; Copeland &
 
            Schweitzer, 1993; Ferguson, 1991a; Ferguson, 1991b; Ferguson, 1992a;
Perse & Ferguson,
 
           1993; Heeter, 1985; Heeter & Greenberg, 1985b; Perse, 1990; Umphrey &
Albarran, 1993;
 
          Walker & Bellamy, 1989;  Walker, Bellamy & Traudt, 1993; Wenner &
Dennehy, 1993; Yorke &
 
            Kitchen, 1985).  In these studies viewers were asked questions such
as how often they
 
          switched channels, why they switched, what programs they switched to
and from, and other
 
            questions assessing their attitudes toward RCDs and television
viewing habits.
        One study using a questionnaire found that of cable and non-cable viewers who
use their
 
            RCDs, about half (49%) change channels during programs. "Heavy
grazers (17.9%) switched
 
            channels more than once every two minutes; moderate grazers (35.8%)
changed one to three
 
            times every six and a half minutes and light grazers (46.3%) changed
one time every 20
 
           minutes" (Selnow, 1989, p. 35).
        Data collected from over 1500 adults and 400 children from five different
surveys
 
       revealed 3.4 channel changes occurred per hour during regularly-watched
shows.  Shows
 
          watched on a non-regular basis had a slightly higher number of changes
per hour (4.6) (
 
           Heeter & Greenberg, 1985b).
        A telephone survey asked 583 randomly-selected adults: "during a typical hour
of TV
 
         viewing yesterday, how often did you change the channel?" (Ferguson,
1991a, p. 9). The
 
           respondents reported a mean of 4.9 channel changes per hour.  When
students at two di
 
         fferent universities were asked a similar question the results were 3.4
(N=455) (Walker &
 
            Bellamy, 1989), and 3.2 changes per hour (N=219) (Wenner & Dennehy,
1993).
        It is difficult for a respondent who may have changed channels a multitude of
times over
 
            the course of an evening to remember which channels were viewed,
when they were viewed,
 
            and which channels were merely scanned.  Channel changing can occur
with such frequency
 
            and with such automatic response that viewers may not remember how
often they changed
 
          channels, thus decreasing the reliability of the survey results
(Ferguson, 1994).
        Ferguson (1994) suggests that asking viewers how many times they use an RCD
when watching
 television is akin to asking them how many times during a typical hour they
look at their
 watches or perform  other "mundane" behaviors.  The measurement of mundane
behaviors is
 
            problematic and questionable when done by self-report.
        It is not that self-report is useless in examining motives or recalling
behaviors, but
 
            that, ideally, reliably-monitored behaviors coupled with
self-reports increase our
 
       knowledge of the association between motives and actions.
        In-home observation and electronic techniques.  The disadvantages of
questionnaires and
 
            telephone surveys spurred the need for more accurate methods of
measuring television
 
         viewing and RCD use.  To overcome the weaknesses of self-reports,
researchers studying RCD
 use are struggling with many of the same issues of gauging television viewing
and are
 
           experimenting with various types of measurement. Some studies have
coupled self-report
 
           with in-home observations, while other studies have used in-home or
laboratory observation
 alone or in combination with videotapes of programs and of the viewing area.
 
   Additionally, electronic monitoring of programs has taken place in either
in-home or
 
         laboratory settings.
          Using a primarily in-home observation procedure, Eastman & Newton (1993),
sent a team
 
            of student observers to 115 solo viewing situations and 44 group
watching conditions
 
         during primetime viewing hours.  The participants' RCD activities were
recorded by
 
       observers with whom they were familiar.  The results indicate that when
counting the
 
         number of channel changes during 30-minute and one-hour long programs
only (excluding
 
          sporting events and movies), an average of only 1.7 changes occurred
during each half hour
 of viewing (Eastman & Newton, 1993).  The results of this study indicate that
viewers use
 the RCD mainly to change channels between programs and, overall, use their RCDs
less than
 claimed in self-reports.
        Moriarty (1991) conducted in-home observations of 10 small groups and two
families over
 
            12 evenings of 45 minutes each.  The observers asked friends to
participate.  It was found
 that RCD use occurred about once every three minutes and twice as often in
commercials as
 in programs.
        The combination of videotaping television programs and camera observation of
the viewing
 
            area was the procedure used by Cornwell, et al. (1993).  The
equipment was placed in 10
 
            homes for six days.  The cameras were used to record the television
viewing area and the
 
            VCRs were programmed to tape programs from 5 p.m. to 11 p.m. each
day.   The videotapes
 
            show 13.6 "uses" per hour.  An RCD "use" was defined as any
television viewing function
 
            such as channel changing, muting, on/off power made with an RCD.
        In a laboratory-experimental study, Bryant & Rockwell (1993) observed the
television
 
          viewing behaviors of 80 adults in a living room-like setting.  Four
different conditions
 
            were created: 1) no RCD, 2) an RCD with only a power on/off button,
3) an RCD with on/off
 
            and channel changing functions, and 4) an RCD with on/off, channel
changing functions and
 
            volume control.  One result of the study supports the supposition
that viewers armed with
 
            an RCD with channel-changing functions switch channels more
frequently (a total of about
 
            67.5 times) than those without an RCD (41.0 times) and those with an
RCD with only a power
 on/off button (35.0 times).
        Going beyond videotaping, observation, and self report, one study monitored
cable
 
       television viewing via a computer located in a cable system's headend. A
random sample of
 
            cable subscribers was monitored (Heeter, D'Alessio, Greenberg, and
McVoy, 1988). The
 
         computer recorded the number of channel changes and the time each
channel was changed.  In
 their electronic assessment of cable viewing, Heeter, et al. (1988) discovered
that
 
         viewers made an average of 4.4 channel changes per hour.
        In a laboratory study, the number of channel changes subjects made while
watching
 
       television were recorded by a specially-designed electronic counter. The
number of
 
       electronically-monitored channel changes ranged from three to 396 times
per hour, with a m
 
            ean of 107 (Ferguson, 1994).
        The techniques used in the above-mentioned studies share some common
limitations:  1)
 
           reliance on self-report (Ferguson, 1991a; Heeter & Greenberg, 1985b;
Selnow, 1989; Walker
 & Bellamy, 1989; Wenner & Dennehy, 1993); 2) small sample size (Ferguson 1994;
Moriarty,
 
            1991; and Cornwell, et al., 1993);  3) observational obtrusiveness
(Eastman & Newton,
 
          1993; Moriarty, 1991);  4) RCD activity monitored only during limited
viewing times
 
        (Krendl et al., 1993; Eastman & Newton, 1993; Ferguson, 1994; Moriarty,
1991; Traudt,
 
          1993; and Cornwell, et al., 1993);  5) laboratory setting (Bryant &
Rockwell, 1993;
 
        Ferguson, 1994); and 6) non-random sample (Eastman & Newton, 1993;
Moriarty, 1991).
        The methodologies employed in these studies have ranged from self-report
(telephone
 
         surveys and questionnaires) to observations to electronic measures. As
a consequence of
 
            differing procedures the outcomes (number of channel changes) vary
widely from study to
 
            study.   The number of channel changes in any time period varies
widely.  This is the case
 regardless of whether channel changing activity is recalled, observed,
videotaped or
 
          electronically monitored (refer to Table 1).
        With the advent of 60+ cable television, VCRs, and remote control devices, the
 
    development of more accurate measurement methodologies is of great concern.
Though it is
 
            clear that RCDs are changing the way people watch television,
precise, valid and reliable
 
            methods of monitoring RCD use need to be developed.  One approach is
an electronic
 
       monitoring technique that will accurately record the RCD activity of a
random sample of
 
            participants in a naturalistic environment.
        For the present study a custom-designed microcontroller connected to a cable
converter
 
            box was developed that allowed the researcher to precisely monitor
the following
 
     channel-changing activities in participants' homes: how often changing
takes place, the
 
            length of time channels are watched before a change is made, and in
what time periods
 
          changing occurs. The data also revealed which channels are watched
most frequently and how
 many channels are regularly scanned and/or watched.  Additionally, the results
from
 
         monitoring actual RCD behavior are compared with self-reports. The
comparison highlights
 
            any inconsistencies between people's recall of their RCD use and
actual use.
 
RCD Activity
        Number of channel changes.  This study electronically monitored the number of
channel
 
           changes as well as the length of time each channel was watched before
the next switch
 
          occurred.  A channel switch is the instantaneous push of a button (or
buttons) on an RCD
 
            which results in a change from one channel to another.  Therefore, a
channel change is
 
           operationalized as the switch to the channel and the duration of time
the channel is on
 
            before the switch to the next channel is made.  This leads to the
first two research
 
         questions regarding the frequency of channel changes.
        RQ1. What is the overall average number of channel changes made using an RCD
per hour        of
 viewing ?
 
RQ2.  Does channel changing activity vary from daypart to daypart in the length
of time
 
                  that channels are watched before a change is made and in the
number of channel switches?
 
        Recalled vs actual RCD use.  Another objective of this study was to discover
 
  inconsistencies, if any, that exist between viewers' recall of RCD use and
actual use.
 
            The common methodology employed in past RCD studies has been the
survey which is dependent
 on self-report. In the present study, a post-monitoring questionnaire queried
recall of
 
            RCD use.  This, coupled with the precise record of channel changing
activity, allowed an
 
            analysis of inconsistencies between actual and reported RCD use.
        Comparing actual to self-reported RCD use, the following research question was
examined:
RQ3.  Are estimates of RCD use based upon recall higher or lower than the actual
number
 
                  derived from electronic monitoring?
 
        Viewing Styles and RCDs.  Counting the number of channel changes is just one
measure of
 
            RCD activity.  Examining the ways in which viewers watch programs is
another (Heeter, et
 
            al., 1988).  An individual who averages 20 channel changes per hour
might be classified as
 a "grazer" who does not watch programs in their entirety.  However, the
presumption of
 
            grazing could be erroneous.  One channel may be watched for a long
period of time; the 20
 
            channel changes could occur between programs.
        Heeter, et al. (1988) operationalized three modes of viewing: 1)
scanning--watching one
 
            or more channels each for four minutes or less, 2) extended
sampling--watching a channel
 
            from four to 15 minutes, and 3) stretch viewing--watching a channel
for 15 minutes or
 
          longer.  Data from electronically-monitored channel-changing activity
were utilized to
 
           answer the following questions:
RQ4. What percentage of time do viewers spend in: a) scanning; b) extended
sampling; and
 
                  c) stretch viewing?
 
RQ5. Does an analysis of channel-switching behavior within shorter time periods
(e.g. less
 than five seconds) reveal other viewing styles in addition to scanning,
extended sampling
 and stretch viewing?
 
        Cable television provides viewers with a large selection of channels from which
to
 
        choose.  Larger channel repertoires together with the RCD have
contributed to the erosion
 
            of the major network audiences.  This study examines the audience
share of available
 
         channels in terms of the percentage of overall viewing time each of the
channels received.
   It is presumed that viewers are "grazing" and not really "watching" a channel
if it is
 
            on for less than four minutes at a time.  Thus, a channel was
considered part of a
 
       viewer's repertoire only if it was on for a least four minutes at a time
without a switch.
  This leads to the final research question:
        RQ6.  What percentage of their overall viewing time do participants spend
watching each
 
            of the various types of channels (i.e. networks, cable news, etc.)?
 
METHODOLOGY
 
Sample
        For this study, television remote control usage was monitored in a medium-sized
southern
 
            city from February 28th through April 11th, 1994.  Participants were
randomly selected
 
           from those who responded to a prior telephone survey assessing
television viewing.  As
 
           part of the survey, respondents were asked if they had an RCD, cable
service, and a cable
 
            converter (N=386).  For the present study, in addition to the
aforementioned criteria,
 
           basic-only service or basic and HBO were required.  These are the
only channels that could
 be received by the RCD monitoring unit (specially-adapted cable converter box
attached to
 a microprocessor).  Only 44 respondents reported meeting all of the criteria
necessary to
 participate in the study.  Of the 23 respondents who consented to participate,
15 resided
 in households with other family members who also agreed to take part. Thus, a
total of 44
 participants (40 adults and 4 children under 12 years of age) were included in
this
 
         study.
        The methodology consists of three parts: 1) electronic monitoring of RCD use in
a
 
       naturalistic setting, 2) a viewing diary, and 3) a post-monitoring
questionnaire.
 
Procedure
        Monitoring RCD use.   A specially-designed microcontroller (RCD monitoring
unit)
 
      connected to a cable converter box recorded the number of channel changes,
the channels
 
            that were selected, the time the channels were switched, and the
length of time each
 
         channel was selected.   Additional viewing information obtained from
the diary was later
 
            added to the data file.     The monitoring unit did not require any
attention from the
 
         participants. Additionally, it did not give any indication (lights,
noise) that it was
 
           recording RCD use.   The unit was left in each participant's home for
a period of four
 
           days.  These four-day periods varied from household to household.
For example, data were
 
            recorded at one household Monday through Thursday and at another
household on Wednesday
 
            through Saturday, and so on. Thus, across all 23 households, RCD
activity was monitored
 
            and analyzed from all seven days of the week as well as across six
weeks1.  This allowed
 
            for potentially thousands of minutes of television viewing during
which substantial RCD
 
            activity was possible.
        Viewing Diary.  In addition to recording RCD use, the participants filled out a
diary
 
           indicating who was watching television and who was operating the RCD.
A record of which
 
            viewers were watching television and using the RCD permits analyses
of the number of
 
         channel changes, etc. to be based on the individual RCD user rather
than the household.
 
            For each 15-minute time block throughout the day in which the
television was on,
 
     participants were asked to indicate who was watching and who was making the
channel
 
        changes using the RCD.
        Participants were also asked to indicate whether a videotape was being viewed
or whether
 
            Nintendo or some other game was being played. This was required
because either channel 3
 
            or channel 4 is displayed on a cable converter box when a videotape
or game is being
 
         played. The RCD monitoring unit simply records the channel number and
cannot differentiate
 between these activities (games) and off-air television viewing.
        Post-monitoring questionnaire.  After the RCD monitoring period, a
questionnaire was
 
          filled out by each participants 12 years of age and older to assess
RCD and television
 
           viewing behaviors.  To measure RCD activity, participants were asked
to recall "as closely
 as you can estimate, during a typical hour of TV viewing how many times do you
use the
 
            remote control device to change channels."
 
RESULTS
 
Profile of Participants
 
        This study monitored the number of channel changes made by 44 participants (40
adults, 4
 
            children) in 23 households.  The gender ratio is 47.5% male to 52.5%
female.  The adults'
 
            ages ranged from 17 to 68 years (children's ages ranged from 3 to 11
years), and 70.5%
 
            are white.  Slightly more than half (55.0%) had three or more years
of college and the
 
           modal household income ranges were $10,000-$19,999 and $20,000-29,000
(n=9 for each).
 
RCD Activity
        Number of channel changes.  The first research question examines the total
number of
 
          channel changes.  Further, comparisons are made between the number of
channel switches
 
           made in this study and those found in previous RCD studies.
        RCD activity was monitored during slightly more than 374 hours (374 hours, 8
minutes and
 
            4 seconds) of television viewing across six weeks. The total number
of channel changes
 
           made by the study's participants was 13,680. The average number of
channel changes made
 
            per hour was 36.6 or one channel change every one minute and 38
seconds.  The mean number
 
            of channel changes per hour across individuals ranged from 1.23 to
178.2 (SD = 41.3).
 
            The average number of channel switches found in this study was 7.5
to 15 times more than
 
            revealed in previous studies that used self-report to measure RCD
activity.  When compared
 to previous studies that electronically assessed RCD use, the participants in
this study
 
            changed channels almost 8.5 times more often as those subjects in
Heeter, et al. (1988),
 
            but were well below the rate found by Ferguson (1994) (see Table 1).
        Channel viewing by daypart.  Examination of channel changes ascertained the
overall
 
         average length of time channels are watched for various three-hour
dayparts.  Daypart
 
          comparisons allow an examination of channel-changing behavior in prime
time (8 - 11 p.m.)
 
            as well as in other popular viewing times.   A  total  of  79.3%  of
all channel switches
 were made during the 15 hours comprising five dayparts:
        1) 6 - 9 a.m.  Morning viewing (news, etc.).
        2) 12 - 3 p.m. Afternoon viewing (soap operas, talk shows).
        3) 5 - 8 p.m. Late afternoon (local news, syndication).
        4) 8 - 11 p.m. Prime time (dramas, sitcoms, etc.).
        5) 11 p.m. - 2 a.m. Late night (talk shows, news, etc.).
 
        A one-way analysis of variance of the total number of seconds between channel
changes
 
           revealed a significant difference in the length of time that channels
were watched before
 
            a change was made among the five dayparts (F [4, 10,848] = 8.38 (p <
.05). Follow-up
 
         pairwise comparisons between the dayparts were conducted. The
participants in this study
 
            watched channels for a significantly longer duration (three minutes
and five seconds)
 
          without a change during the early afternoon daypart (noon to 3 p.m.)
(Duncan's Multiple
 
            Range test [43.7 - 48.7, a=.05]).  The number of seconds viewers
remained on channels was
 
            lowest during the late night viewing period (11 p.m. to 2 a.m.) with
one change every 51
 
            seconds (see Table 2).
        Dayparts were also examined by the average number of channel changes per hour.
During
 
            the late night viewing period channels were switched at the rate of
70.2 changes per hour.
  This compares with 19.6 changes per hour made while watching television in the
early
 
           afternoon. Table 2 depicts the average number of channel switches per
hour by each
 
       daypart.
        Recalled Vs Actual RCD Use.  The third research question assesses any
inconsistencies
 
           that exist between viewers' recall of RCD use and electronic records
of actual use.  The
 
            mean number of self-reported channel switches per hour was 4.8
compared with the
 
     electronically-monitored mean (44.2)2.  The self-reported number of channel
changes per
 
           hour ranged from 0 to 20.0.
        Participants made 9.2 times more channel changes in an average hour of
television viewing
 than they had estimated.  A paired two-sample t-test found a significant
difference
 
         between the number of recalled channel changes per hour and the actual
number of switches
 
            per hour (t [35] = 5.82 (p < .025, two-tail).
        Viewing styles.  The three viewing styles identified by Heeter, et al. (1988)
were the
 
            focus of the fourth research question.  As seen in Table 3, nearly
all of the channel
 
          changes were made for scanning purposes.  However, scanning accounted
for only 8.3% of the
 total viewing time.  In contrast to scanning, stretch viewing comprised more
than three
 
            quarters (79.0%) of the total viewing time, but only 2.5% of all
channel switches.
 
       Additionally, viewers spent 12.7% of their viewing time engaged in
extended sampling.
        Other viewing styles.   To more closely examine channel switching behavior the
three
 
          viewing styles were reconstituted.  The scanning period was subdivided
into four periods,
 
            the remaining two viewing styles were also subdivided into shorter
time frames to check
 
            for patterns of channel changing within each style.
        When viewing styles are examined using the 10 time periods, it is found that
eight in 10
 
            (80.1%) channel switches were made after less than five seconds on a
channel.  Thus, in
 
            addition to the three viewing styles identified by Heeter, et al.
(1988), a fourth style
 
            of "rapid-fire" channel switching (changes made in less than five
seconds of viewing)
 
          emerges.  However, this large percentage of changes only accounts for
slightly less than
 
            one percent (0.9) of the overall viewing time. This compares with
stretch viewing of 60
 
            minutes or more which accounts for just 0.6% of all channel changes
but 42.6% of the
 
         participants' viewing time.  Table 4 profiles the percent of channel
changes and share of
 
            overall viewing time spent on channels for each viewing period.
        Rapid-fire channel switching may be quick program evaluation and/or simply a
means of
 
           getting from one channel to another without attending to in-between
channels.  This can be
 distinguished from staying on channels for at least five seconds to four
minutes where
 
            changes may be more for the purpose of program evaluation rather
than as a way of getting
 
            from one end of the dial to the other.  In other words, rapid-fire
switching may occur for
 the purposes of both evaluation and for traversing through channels, whereas
scanning may
 take place primarily for program evaluation.
        Viewing time.   The last research question examined the length of overall
viewing time
 
            participants spent watching various types of channels. The types of
channels were
 
      collapsed into groups.  Almost half (49%) of the viewers' time was spent
watching the
 
          three broadcast networks (ABC, CBS, NBC) and Fox, though these
channels only account for
 
            14.3% of all channels available (35) in the study's market3.  In
comparison, 14 general
 
           cable channels accounted for one quarter of the participants' viewing
time though they
 
           make up 40% of the available channels (refer to Table 5).
 
DISCUSSION
 
RCD Activity
        Selective exposure is associated with choosing preferred mass media content and
the
 
         behavior that leads to it is known as selectivity.  When wielding RCDs,
viewers can be as
 
            selective as they wish with the push of a button.  RCDs also allow
viewers to create their
 own style of selectivity.  RCDs aid selectivity by increasing viewers'
programming
 
        choices and by allowing easy access to those choices.  Some RCD users
may constantly graze
 through channels rarely watching any program from the beginning to its
conclusion. Some
 
            viewers may select a program by grazing through all available
channels or a subset of
 
          channels before deciding on one program to watch.  Still other viewers
will turn to a
 
          channel using their RCDs and watch it from beginning to end without
evaluating other
 
         programs first.  This study electronically-monitored and examined
viewers' levels of
 
         selectivity: how often and in what ways they use an RCD.
         The results indicate a selective audience which scans the dial at the rate of
one
 
        channel every one minute and 38 seconds.  When compared with two
previous studies that
 
           also electronically monitored channel changes, this study found a
more selective audience
 
            than Heeter, et al. (1988) - 36.6 changes per hour vs 4.4 changes
per hour, respectively.
  However, the participants in Ferguson's (1994) laboratory study made almost
three times
 
            as many changes per hour (107) as those in the present study.
         The rate of channel changes varies widely among the present study and those
conducted by
 Heeter, et al. (1988) and Ferguson (1994).  The high rate of switches counted
by Ferguson
 (1994) may be attributed to a laboratory setting wherein each student's RCD
activity was
 
            monitored for only one hour.  To switch channels in the households
studied by Heeter, et
 
            al. (1988) the predetermined channel number had to be entered on the
numeric keypad and
 
            thus did not allow viewers to rapidly scan the dial simply by
pushing an up/down arrow
 
           key.  Therefore, the reported rate of channel switching excludes
rapid-fire grazing and
 
            includes only changes made with the purpose of going from one
channel to another channel
 
            already in mind.
        RCD activity was examined within the context of three viewing styles identified
by
 
        Heeter, et al., (1988): scanning, extended sampling and stretch viewing.
Further, the
 
           three viewing styles were subdivided into shorter time frames
revealing a fourth method of
 viewing;  rapid-fire switching within five seconds on a particular channel.  As
the
 
         viewing audience quickly flips through channels it may be making
decisions of whether or
 
            not to watch a program within five seconds of tuning in. It may also
be possible that
 
          these rapid-fire changes are simply a means of quickly moving from one
dial position to
 
            another with little evaluation taking place.
        It was found that eight in ten channels switches were made in quick succession;
yet these
 rapid-fire changes only account for slightly less than one percent of the total
viewing
 
            time.  By comparison, participants spent more than three-quarters of
their time stretch
 
            viewing during which only a tiny portion of all channel changes were
made.  It is apparent
 that the overall pattern of channel changing behavior indicates that viewers
make
 
       numerous channel switches in short bursts but settle down to watch
programs for longer
 
           periods of time.
        The monitoring device used in the present study precisely recorded the length
of time
 
           (even when less than one second) that channels were selected before
the next change
 
        occurred.   This revealed the pattern of rapid-fire changes which has
been masked in
 
         previous studies which simply averaged the number of switches per hour.
The mean number
 
            of switches alone does not accurately reflect channel changing
behavior.  For example, two
 viewers may each make an average of 15 changes per hour; the first viewer may
make 14
 
           switches in rapid succession then stay tuned to one channel for 59
minutes, whereas, the
 
            second viewer may change channels throughout the hour.  Though the
mean number of channel
 
            changes each viewer makes is the same, the RCD behaviors vastly
differ. This indicates
 
           that a more comprehensive measure of RCD use should include both the
number of channel
 
           changes and the length of time a channel is watched before a switch
is made.
        An important aspect of this study is the determination of the accuracy of
self-reported
 
            channel changes relative to actual flipping behavior. Participants
in this study switched
 
            channels almost 10 times more often in a typical hour of viewing
than they recalled (44.2
 
            to 4.8, respectively).  Ferguson (1994) found that participants
underestimated the number
 
            of channel changes they made by a factor of more than three times.
Both of these studies
 
            lend support to the contention that RCD use cannot be accurately
assessed by self-report
 
            alone.
        Channel changes may have been underestimated because viewers only counted the
number of
 
            meaningful switches as opposed to "mundane" changes.  A meaningful
change is one where
 
           viewers consciously decide to make a change to a specific
predetermined channel.  In
 
         comparison, mundane switches could be those made without an end channel
in mind.
 
      Alternatively, mundane switches could be those that are made only as a
means of traversing
  channels on the way to finding a desired program.
        Daypart analyses were conducted to ascertain whether channel-changing behavior
varied
 
           during different times of the day.   When the early afternoon time
period is considered,
 
            viewers stayed on channels slightly more than three and a half times
longer (three minutes
 and five seconds) than when watching late night television (51 seconds).       The
difference
 
            in viewing between the two dayparts may be explained by the types of
programs offered.
 
           Early afternoon television typically includes soap operas, talk shows
and game shows as
 
            compared with late night fare of entertainment/talk shows, news
programs and movies.  It
 
            could be that early afternoon shows, especially soap operas, have
gained program loyalty
 
            that late night programs have not.  Program loyalty may foster more
instrumental viewing
 
            as the audience tunes in to a regularly-watched show and thus is
less inclined to change
 
            the channel.
        Television may also be watched more ritualistically during the late night hours
than
 
          during the early afternoon simply due to the time of day.  Viewers may
watch television at
 night as a way to pass the time or to help them get to sleep. It seems probable
that
 
          under these conditions viewers would be more likely to frequently
switch channels as they
 
            may not be interested in attending to one program, especially one of
long duration.
        This study also examined the overall percent of viewing time the participants'
spent
 
          watching various types of channels.  According to the results of this
study, viewers spent
 almost half of their time watching the major broadcast networks (ABC, CBS, NBC
(2), and
 
            Fox), although these channels account for only 14.3% of the channels
available on the
 
          35-channel cable system available in the test city.
        In the television industry there is concern that RCDs contribute to the
reduction of
 
          network audience share by facilitating the viewing of cable channels.
To offset the
 
        effects of network erosion, the television industry is discussing new
programming stra
 
          tegies to keep viewers tuned to one channel (Eastman & Neal-Lunsford,
1993).   One such
 
            plan, channel tiering or the clustering of like programs together on
the channel dial, is
 
            being considered, especially since grazers tend to search up and
down through several
 
          adjacent channels.  (Eastman & Neal-Lunsford, 1993).
        Other techniques include "top-loading" in which attention-getting action is
featured at
 
            the beginning of a program to capture an audience.  In addition to
top-loading, program
 
            producers are creating faster-paced shows as an attempt to overcome
viewers' ennui and to
 
            circumvent RCD use (Eastman & Neal-Lunsford, 1993).   Starting with
the 1994-95 season,
 
            ABC, CBS and Fox introduced seamless programming by postponing
commercials well into a
 
           program's plot development to decrease the amount of flipping
(Eastman & Neal-Lunsford,
 
            1993; Goldman, 1994).
 
Limitations
        There are four major limitations to this study: 1) the microcontroller's
dependence on a
 
            cable converter box to enable it to capture channel changes, 2) the
small number of
 
        participants, 3)  the small number of monitoring devices, and 4)
monitored RCD activity
 
            limited to only four days in each household.
        The design of the monitoring unit required a cable converter box to aid the
 
 microcontroller in capturing the RCD button pushes.  Thus, this system excluded
non-cable
 
            households, households that received premium channels other than
HBO, and households that
 
            have cable-ready televisions and thus do not require a cable
converter box.
        Due to monetary constraints only six microcontrollers were manufactured.  If a
larger
 
           supply of microcontrollers had been available the entire sample could
have been monitored
 
            within a shorter period of time thus increasing control over any
differences in weekly
 
           programming schedules and for any special programs.
        Lastly, RCD activity was monitored for only four days in each household rather
than for a
 full week.  This may have impacted RCD activity as individuals' viewing habits
may vary
 
            from day to day.  For example, a viewer may change channels less
frequently during weekend
 viewing  than during the weekdays when his/her RCD activity was recorded.
 
Future Research
        Television sets and RCDs are undergoing changes that will make them
technologically more
 
            sophisticated with capabilities beyond the models currently in use.
These changes will
 
            offer RCD users more ways to create their own viewing styles.
Studies might look at the
 
            implications of newer more elaborately designed RCDs on selective
exposure to television
 
            programs and interactivity between viewers and their television
sets.  Future research
 
           might examine how the increasing functionality of RCDs is impacting
the television and
 
           advertising industries as well as viewing behavior.  The
effectiveness of new programming
 
            and advertising strategies in curbing the amount of channel changing
might also be
 
       researched.  Additionally,  a more technologically advanced electronic
monitoring device
 
            is needed to accurately record RCD and television viewing behaviors.
 
 
Table 1
Comparison of the Number of Channel Changes Across Studies
 
                Number of               Range of
        Type of         Channel Changes Channel Changes
        Studies         Measurement       Per Hour              Per Hour
 
Present study (1994)    electronic      36.6    1.2 - 178.2
(n = 44)
 
-------------------------------------------------------------
 
        Ferguson (1994) electronic (lab)        107.0   3.0 - 396.0
(n = 49)
 
Heeter, et al., (1988)  electronic      4.4     N/A1
(n =    197 HH)2
 
Cornwell, et al., (1993)        videotape       13.6    .40 -  31.1
(n = 17)
 
Eastman & Newton (1993) in-home observation          1.73               1.3 - 2.23
(n = 253)
 
Wenner & Dennehy (1993) self-report     3.2     N/A
(n = 219)
 
Ferguson (1991a)        self-report     4.9     N/A
(n = 583)
 
Walker & Bellamy (1989) self-report     3.4     N/A
(n = 455)
 
Ainslie (1989)  self-report     2.5     N/A
(n = 494)
 
Heeter & Greenberg (1985b)self-report   4.6     N/A
(n = 1900)
 
 
 
1 Not available.
2 Household. Other studies reported behavior of individuals.
3 Channel changes per half-hour.
 
 
 
Table 2
Average Length of Time Spent Viewing a Channel
Without Switching by Daypart
 
 
Daypart
        6 am -  12 pm-  5  pm-  8 pm-   11 pm-  Overall
        9 am      3 pm  8 pm    11 pm           2 am    Average
 
 
All Viewing Styles      1 min.  3 min.  1 min.  1 min.          1 min.
Combined        15 sec.ab       5 sec.c 33 sec.ab       50 sec.b        51 sec.a        44 sec.
 
 
 
 
Average Number of Channel Changes per Hour
of Viewing by Daypart
 
 
Daypart
 
        6 am -  12 pm-  5  pm-  8 pm-   11 pm-  Overall
        9 am      3 pm  8 pm    11 pm           2 am    Average
 
 
All Viewing Styles
Combined        48.2    19.6    38.6    32.8    70.2    41.9
 
 
 
Note: Means with different letters differ significantly at p <.05 alpha level
using Duncan's
 
        Multiple Range Test.  For example,  when all viewing styles are
combined, participants watched
 
          channels for a significantly longer length of time before switching
during the 12 p.m. - 3 p.m.
 
           period than during 11 p.m. - 2 a.m.
 
 
 
 
 
 
Table 3
 
Viewing Time Spent Scanning,
Extended Sampling and Stretch Viewing
 
                                                Present Study   Heeter, et al. (1988)
 
        Channel Changes*        Viewing Time**  Viewing Time
        (Freq.)          (%)    (%)                     (%)
 
Scanning        12,962          94.8    8.3     7.9
 
Extended Sampling       371     2.7     12.7    9.9
 
Stretch Viewing 346     2.5     79.0    82.0
 
Total   13,680
 
 
Note:  Present study compared with Heeter, et al., 1988
 
* A channel change includes the switch to the channel and the length of time the
channel is on until
 the switch to the next channel occurs.
 
**  Of the overall time channels were viewed, 8.3% was spent scanning  (changing
channels after
 
           viewing them for less than four minutes at a time).
 
 
 
Table 4
Viewing Styles
Subdivided Into Shorter Lengths of Time
 
     Length of Time          Percent of                 Percent of
     Spent Viewing      Channel Changes Overall Viewing Time
 
        Scanning:
                0 - 5 seconds   80.1    0.9
                6 - 60 seconds  10.6    2.2
                1 - 2 minutes   2.1     1.8
                2 -4 minutes    2.0     3.4
        Extended Sampling:
                4 - 8 minutes   1.6     5.5
                8 - 12 minutes  0.8     4.6
                12 - 15 minutes 0.3     2.6
        Stretch Viewing:
                15 - 30 minutes 1.1     14.2
                30 - 60 minutes 0.8     22.2
                More than 60 minutes    0.6     42.6
 
 
Table 5
 
Percent of Channel Repertoire
and Percent of Time Spent Viewing Channels
 
 
        % of    % of
Type of Channel Available Channels      Viewing Time
 
Networks (5 Channels
        (NBC (2), CBS, ABC, Fox)        14.3    49.0
 
General Cable (14 Channels)     40.0    25.0
        (Family, USA, TNT, etc.)
 
Specialized Cable (7 Channels)  20.0    11.7
        (ESPN, ESPN2, VH1,
        MTV, Weather, QVC,
        Program Guide)
 
Cable News (3 Channels)
        (CNN, Headline, CSPAN)  8.6     5.0
 
PBS (2 Channels)        5.7     3.0
Premium (4 Channels*)   11.4    6.3
 
 
*15 Households (30 Participants) received HBO. During the monitoring period "The
Disney Channel",
 
             "Showtime" and "Cinemax" were provided free of charge for two days
each as a special promotion by
 
             Comcast Cable.
 
 Notes
 
 
 
1 An RCD electronic monitoring device was left in each of 23 households for four
days
 
          yielding approximately 2,200 potential viewing hours across the study.
RCD activity was
 
            recorded during slightly more than 374 hours of television viewing
indicating that in each
 household the television was turned on for about six hours per day on average.
 
2 This mean differs from the overall mean of 36.6 (n = 44) due to the difference
in the
 
            number of participants who responded to this questionnaire item and
were thus included in
 
            this analysis (n = 36).
 
3 The study market received NBC on two channels.
 
 
 
 
 
 REFERENCES
 
 
Ainslie, P. (1989). The new TV viewer. In C. Magazine (Ed.), How Americans Watch
TV: A
 
           Nation of Viewers. C.C. Publishing.
 
Bellamy, R. V. (1993). Remote control devices and the political economy of a
changing
 
          television industry. In J. R. Walker and R. V. Bellamy (Eds.), The
Remote Control in the
 
                  New Age of Television. (pp. 211-220). Westport: Praeger
Publishers.
 
Bellamy, R. V. & Walker, J. R. (1990).  The use of television remote control
devices: A
 
            transactional study. News Computing Journal. 6(1), p. 1-18.
 
Bryant J. & Rockwell S. C. (1993). Remote control devices in television program
selection:
 Experimental evidence. In J. R. Walker and R. V. Bellamy (Eds.), The Remote
Control in
 
            the New Age of Television. (pp. 73-86). Westport: Praeger
Publishers.
 
Copeland, G. A. (1989). The impact of remote control tuners of family viewing.
Paper
 
         presented at the Speech Communication Association,  San Francisco, CA.
 
Copeland, G. A., & Schweitzer K. (1993). Domination of the remote control during
family
 
            viewing. In J. R. Walker and R. V. Bellamy (Eds.), The Remote
Control in the New Age of
 
                  Television. (pp. 155-168). Westport: Praeger Publishers.
 
Cornwell, N. C., Everett, S., Everett, S. E., Moriarty, S., Russomanno, J. A.,
Tracey, M.,
 & Trager, R. (1993). Measuring RCD use: Method matters. In J. R. Walker and R.
V. Bellamy
 (Eds.), The Remote Control in the New Age of Television. (pp. 43-56). Westport:
Praeger
 
            Publishers.
 
Eastman, S. T. & Neal-Lunsford, J. (1993). The RCD's impact on programming and
promotion.
 
            In J. R. Walker and R. V. Bellamy (Eds.), The Remote Control in the
New Age of Television.
 (pp. 189-210). Westport: Praeger Publishers.
 
Eastman, S. T. & Newton, G. D. (1993). The myth of grazing: Observations of
remote control
 use.  A paper presented at the Broadcast Education Association conference, Las
Vegas,
 
           Nevada: April.
 
Ferguson, D. A. (1991a). Gender differences in the use of remote control
devices.  A paper
 presented at the Speech Communication Association conference, Atlanta, Georgia:
November.
 
Ferguson, D. A. (1991b). Selective exposure to television: Predicting
inheritance effects
 
            from VCR and cable penetration.  Dowden Center Journal. 1 (1).
 
Ferguson, D. A. (1992a). Channel repertoire in the presence of remote control
devices,
 
           VCRs and cable television. Journal of Broadcasting & Electronic
Media, Spring, (36) 83-91.
 
 
Ferguson, D. A. (1994). Measurement of mundane TV behaviors: Remote control
device
 
       flipping frequency. Journal of Broadcasting & Electronic Media. 38 (1),
35-47.
 
Goldman, K. (1994, July 28). Stayed tuned for no commercials between shows.  The
Wall
 
          Street Journal. pp. B1, B6.
 
Griffin, E. (1991). A First Look at Communication Theory. New York: McGraw-Hill,
Inc.
 
Heeter, C. (1985). Program selection with abundance of choice. Human
Communication
 
       Research, 12(1), 126-152.
 
Heeter, C., D'Allessio, D., Greenberg, B., & McVoy, S. D. (1988). Cableviewing
behaviors:
 
            An electronic assessment. In C. Heeter, & B. Greenberg (Ed.),
Cableviewing. (pp. 51-63).
 
                  Norwood, NJ: Ablex Publishing Company.
 
Heeter, C. & Greenberg, B. S. (1985a). Cable and program choice. In D. Zillmann,
& J.
 
          Bryant (Ed.), Selective Exposure to Communication. (pp. 203-225).
Hillsdale, NJ: Lawrence
 
                  Erlbaum.
 
Heeter, C. & Greenberg, B. S. (1985b). Profiling the zappers. Journal of
Advertising
 
         Research, 25(2), 15-18.
 
Klopfenstein, B. C. (1993). From gadget to necessity; The diffusion of remote
control
 
          technology. In J. R. Walker and R. V. Bellamy (Eds.), The Remote
Control in the New Age of
 Television. (pp. 23-40). Westport: Praeger Publishers.
 
Krendl, K. A., Troiano, C., Dawson R., & Clark, G. (1993). "O.K. where's the
remote?"
 
          Children, families, and remote control devices. In J. R. Walker and R.
V. Bellamy (Eds.),
 
                  The Remote Control in the New Age of Television. (pp.
137-154). Westport: Praeger
 
            Publishers.
 
Moriarty, S. E. (1991).  Explorations into the commercial encounter.  A paper
presented at
 the American Academy of Advertising conference.
 
Perse, E. M. (1990). Audience selectivity and involvement in the newer media
environment.
 
            Communication Research, 17(5), 675-693.
 
Perse, E. M. & Ferguson, D. A. (1993). Gender differences in remote control use.
In J. R.
 Walker and R. V. Bellamy (Eds.), The Remote Control in the New Age of
Television. (pp.
 
            169-186). Westport: Praeger Publishers.
 
Selnow, G. (1989). Why grazing can't be ignored and what to do about it. In C.
Magazine
 
            (Ed.), How Americans Watch TV: A Nation of Viewers. C.C. Publishing.
 
Traudt, P. J. (1993).  Surveillance and cluster viewing: Foraging through the
RCD
 
      experience.  In J. R. Walker and R. V. Bellamy (Eds.), The Remote Control
in the New Age
 
                  of Television. (pp. 57- 72). Westport: Praeger Publishers.
 
Umphrey D., & Albarran, A. B. (1993). Ethnic differences in remote control
device
 
      motivations. A paper presented at the Broadcast Education Association
conference, Las
 
                Vegas: Nevada, April.
 
Walker, J. R. (1988). Inheritance effects in the new media environment. Journal
of
 
       Broadcasting & Electronic Media, 32(4), 391-401.
 
Walker, J. R., & Bellamy, R. V. (1989). The Gratifications of grazing: Why
flippers flip.
 
            A paper presented at the Speech Communication Association
conference, San Francisco, Ca.
 
Walker, J. R. & Bellamy, R. V. (1992). Gratifications derived from remote
control devices:
 A survey of adult RCD use. A paper presented at the International Communication
 
     Association conference, Miami, Florida:
 
Walker, J. R., Bellamy, R. V.. & Traudt, P. J. (1993). Gratifications derived
from remote
 
            control devices: A survey of adult RCD use. In J. R. Walker and R.
V. Bellamy (Eds.), The
 
                  Remote Control in the New Age of Television.(pp. 103-112).
Westport: Praeger Publishers.
 
Webster, J. G. (1986). Audience behavior in the new media environment. Journal
of
 
      Communication. 36 (3) 77-91.
 
Webster, J. G., & Wakshlag, J., J. (1983). A theory of television program
choice.
 
      Communication Research, 10(4), 430-446.
 
Wenner, L. A., & Dennehy, M. O. (1993). Is the remote control device a toy or
tool?
 
        Exploring the need for activation, desire for control, and technological
affinity in the
 
                  dynamic of RCD use. In J. R. Walker and R. V. Bellamy (Eds.),
The Remote Control in the
 
                  New Age of Television.(pp. 103-112).  Westport: Praeger
Publishers.
 
Yorke, D. A., & Kitchen, P. J. (1985). Channel flickers and video speeders.
Journal of
 
           Advertising Research, 25(2), 21-25.