Investigating The Effects of Audio Exposure toward Podcast Listener
Loyalty via Theory of Planned Behavior
(A Study on Urban Millennials in Jakarta)
JERRY S. JUSTIANTO1, MTS ARIEF2, INDAH SUSILOWATI2, MUHAMAD ARAS2
1Doctor of Research in Management Binus Business School, Binus University Jakarta Indonesia
Jl. K. H. Syahdan No. 9, Kemanggisan, Palmerah,Jakarta, 11480,
INDONESIA
2Management Department, BINUS Business School Doctor of Research in Management,
Bina Nusantara University, Jl. K. H. Syahdan No. 9, Kemanggisan, Palmerah,Jakarta, 11480,
INDONESIA
Received: June 8, 2021. Revised: January 22, 2022. Accepted: February 10, 2022. Published: February 25, 2022.
Abstract:- The purpose of this study was to determine the effect of audio exposure on podcast listener loyalty
through the theory of planned behavior. This research method uses quantitative research methods using SEMPls.
The results showed that there was an indirect effect on the frequency of AOD exposure that was positively related
to individual attitudes towards AOD loyalty. While the frequency of AOD exposure is positively related to
accepted subjective norms, there is no indirect effect of AOD exposure frequency related to subjective norms on
AOD loyalty. While the frequency of AOD exposure has a positive but not significant effect on individual AOD
loyalty, there is no indirect effect on the frequency of AOD exposure that is positively related to the individual's
perceived behavioral control on AOD loyalty. Meanwhile, AOD has a positive and significant effect on AOD-
Loyalty. Perceived subjective norm has a positive and significant effect on AOD loyalty attitudes. Perceived
subjective norms have a positive but not significant effect on individual AOD loyalty. Perceived subjective norm
has a positive and significant effect on AOD loyalty perceived control. From these results obtained a hypothesis
which states that Perceived behavior control related to AOD is positively related to AOD loyalty attitudes and
Perceived Behavior Control over AOD is positively related to AOD Loyalty. Perceived Behavior Control has a
positive and significant effect on AOD Loyalty.
Key-Words:- Podcast Perceived Behavior Control, loyalty perceived control, Perceived subjective norm, subjective
norm
1 Introduction
The phenomenon of podcasting, dating its origins to
2004, can be understood as a partial fulfillment of that
prophecy. While the Internet has not yet killed the
radio station, both Internet radio and podcasting
specifically have continued to grow in popularity,
each with monthly audiences of over 35 million
people [1]. Podcasts have several advantages over
traditional radio, in that they generally have little to no
advertising, allow audiences to time-shift their
listening, and allow for a more personalized listening
experience. Podcasting can therefore be seen as both a
boon and a challenge to traditional broadcasting[1].
With the rise of the trend of Audio on Demand,
the radio industry has to move to this new platform.
As the industry is known for identic to audio media,
now the content has been shifting to another format:
radio streaming and podcast. Some experts believe
that Podcasting as Radio’s Revival [2]. Podcasting is a
technology that allows users to download, store, and
play back both audio and audio-visual content files
[3]. In facts, not only radio industry is the one who
instinctively shall be adapt to these two platforms, but
other media is also looking to audio content, including
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
726
Volume 19, 2022
TV stations, Magazines, Newspaper, all are having
great podcast channel lineup.
One of the reasons of the increasing demand for
audio-media is due to the multi-tasking generation
which busy with their smartphone running several
tasks at the same time was shown by Federica Furlan
that “when media are experienced simultaneously, the
foreground medium dominates over the background in
the audience’s attention.” The results of the study
conclude that persons engaging in multi-tasking in
device have lower attention levels than non-media
multi-taskers [4].
If Podcasting is the main format that many
industries cherish and recommend for content creator
from conventional media and new media to look for,
we have to see several reasons, why they choose
Podcasting as the main source for their audio-content
distribution[5,6]. Unlike radio broadcasting that need
license, and internet radio that need 24 hours
programming and streaming cost, Podcast is a
convenient way of storing and distributing the audio
content. Dario Llineres stating in his book Podcast the
parameters of the new aural culture as “The processes
of production and the creation of content affords new
freedoms with regard to the communication of
knowledge[7]. The medium’s hybridity of thought,
sound and text perhaps even fosters a reinvigoration
of the dialectic, an exchange of ideas beyond what is
possible in purely written form—be it in a magazine
or academic journal[8]. Podcasting, for us, taps into
something fundamental about oral communication,
argument and even the tension between subjective and
objective knowledge that has been amplified in the
digital age [9].
The fact that the media industry reacts to the
opportunistic premise of Podcasts as the dominant
format of audio on demand, while the quantity or time
spent listening among all audio exposures tracked by
Edison Research, Podcasts comprise only 4% for the
general audience, and 28% for users heavy podcasts
[26]. Given the demand for Audio on Demand
increased significantly by 58.7%, we will see the
inability of Podcasts to capture the time spent
listening to the audience [27]. These issues will raise
several research questions to answer what the industry
is facing, one of which, for example is why the media
industry puts so much effort into creating content via
podcasts as the preferred audio on demand channel
while acknowledging only a small amount of effort.
“Share of Ear” which is only 4%. Another question
might be through what channels or tools in particular
the industry should focus on, so that more user-
friendly content for millennials' multitasking behavior
can be found more easily and hook them up to
increase time spent listening.
This research which hopefully finding the role of
active and passive type of content which will be
researched using the introduction of Content Density
using Quantitative Content Analysis incorporating
MMI switch ability to see the effect on multitasking
environment that can lead to the increasing number of
users and time spend of listening.
2 Literature Review
2.1 Theory of Planned Behavior
According to the theory of planned behavior (TPB)
[10], behavioral intention, or the precondition of
behavior, is predicted by three factors: behavioral
attitude (the degree to which one is in favor of or
opposed to performing a behavior), perceived
behavioral control (an individual’s perceived ability to
perform a behavior), and subjective norms (an
individual’s perception of the degree to which
important others think whether he or she should
perform a behavior). Additional research reported that
the three independent variables: applied attitude,
subjective norms, and perceived behavioral control to
predict ICT adoption intention found that all three
variables were significant predictors of college
students’ intention for utilizing an online learning
system[11]. Several studies also consistently using
TPB combine with or without implementing
subjective norm perceived norms such as:
Factors of Online Learning Adoption: A
Comparative Juxtaposition of the Theory of Planned
Behavior and the Technology Acceptance Model; The
results show that both TAM and TPB predict e-
learning adoption intention well [12].
2.2 Hypotheses Development
Tai-Kuei Yu in “An empirical study combining the
task technology fit model with the theory of planned
behavior” offers a new perspective on the task
technology fit model (TTF) into TPB model construct
via the attitude variable along with the other TPB
variables[3,13]. Mou in predicting the podcast
intention, also refer to non-technological factors that
give stronger predictor via the perceived behavior
control from Xigen Li study on technology attributes
toward perceived value of information[3,14].
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
727
Volume 19, 2022
Following are Hypotheses for structural equation
models for the research model. The empirical study of
Lin and Mou, the journal that studied the podcast
adoption based from the theory of planned behavior,
the subjective norm in the model, Lin split it into
descriptive norm and injunctive norm for the purpose
finding which one have more effects [3]. However,
since the independent variable is exposure frequency,
instead of communication frequency, the use of
original TPB is preferred to the separation of
subjective norms like in Mou 2015. The constructs of
perceptions, subjective norms, as well as perceived
behavior control are believed to be educated by
underlying beliefs. The TPB has been used by many
scientists to anticipate a number of behaviors [15].
And these lead to following hypotheses:
Audio Exposure or Exposure Frequency is the
method of looking and what is the most audio media
that reach the audiences ear, and how the effect of
toward listener loyalty, attitude, subjective norms, and
perceived control. This correspond to the study of
Havlena [16] in synergistic effects of cross-platform
media channels exposure toward awareness of
message[16].
H1 : AOD exposure frequency is positively related to
individual attitude toward AOD loyalty
H2 : AOD exposure frequency is positively related to
subjective norm toward AOD loyalty
H3 : AOD exposure frequency will positively related
to individual's AOD Loyalty
H4 : AOD exposure frequency is positively related to
individual perceived behavior control toward AOD
loyalty
For Attitude connecting to the exposure frequency
then to loyalty is a modification from Mou from
communication frequency to intention [3].
H5 : Attitude toward AOD is positively related to
AOD-Loyalty
Subjective norm has been discussed a lot as
determinant factor for intention and behavior, the
approach for the model is using social norm for the
dimension indicator to be measured. Unlike Mou
original research model which dividing the subjective
norm into descriptive and injunctive norm, this will
return it back to its original format [3,10]
H6 : Perceived subjective norm associated toward
AOD is positively related to AOD loyalty attitude
H7 : Perceived subjective norm associated with AOD
use is positively related to individual’s AOD loyalty
H8 : Perceived subjective norm associated toward
AOD is positively related to AOD loyalty perceived
control
Perceived Behavior in this case is about the
comfortableness of users toward tools adoption,
which mostly in this case is Apps. With current apps,
nowadays, not only giving accessibility or user
interface that is easy to use, moreover with the
advancement of deep learning algorithm, but the data
mining of user’s preference had also been recorded
and monitor closely. Recommendation algorithms are
several of the most powerful machine learning
systems today due to their capability to shape the info
we consume. For example, YouTube's algorithm,
especially, has an outsize impact. The platform is
believed to be second only to Google in net visitors,
and 70% of what customers watch is given to them
through suggestions. In recent years, this impact has
come under serious scrutiny. Because the algorithm is
enhanced for getting folks to participate with movies,
it is likely to provide decisions that reinforce what
someone probably wants or perhaps believes, which
could produce an addictive experience which shuts
out some other non-relevant contents [16].
H9 : Perceived behavior control associated toward
AOD is positively related to AOD loyalty attitude
H10 : Perceived Behavior Control over AOD is
positively related to AOD Loyalty
3 Methods
The main model is SEM on 5 variables: Exposure
Frequency, Subjective Norms, Attitude, Perceived
Control, and Loyalty. Modified version of Mou and
Lin [3] research on Podcast, especially on the variable
Attitude toward podcast use. The approach that I will
use is looking at Edison Research way in constructing
the antecedent variable toward the “share of ear”
rating, which are: source, content type, location, and
device. In this case, the source is the first thing that
people use to get the AOD, which are the channel of
the contents pass through the audience. They are:
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
728
Volume 19, 2022
Fig. 1: Share of Ear Sources [18]
As a result, the dimension for Attitude will be the
audio source comes first, and in Indonesia the sources
are Radio, Owned Music, Streaming Service, TV, and
Podcast. These are the sources, which can be accessed
via any devices.
3.1 Profile Respondent
In reversal of Mou [3] research in using purposive
sampling of people who have not used podcast, this
survey is selecting millennials for young adults such
as university students; that have grown up with
smartphones and have experienced in consuming
audio on demand and that include audience that solely
using video on demand, such as YouTube, for the
reason that most people consume the audio section
more than the video itself when busy doing multi-
tasking [25].
Among the 339 responders via a web and mobile
based survey using SurveyHero.com, the completed
answered is by 236 responders which of those are
audio on demand users, 95.8% are using some kind of
apps or tools to access the content for example:
Spotify are 88.6%, Joox 23.7%, and Apple Music is
only 16.1%, and 90.3% of the completed surveys are
also consuming video streaming service. The final
number of responders that are used for statistic
calculation are 206 responders for both the switch-
ability testing of content density and SEM. With
median age of 25.2, Male and Female consist of
41.7% and 58.3%. Students are consisting of 40.3%
and employees are 46.6%. And the responders also
stating that they aware of the audio exposures that
they encounter, with average of 4.8 hours. And the
average minutes of consuming Podcast or Video Cast
in average of 45 minutes per day or approximately
15.6% using the "share of ear" method [18],
comparing to the data for purposive podcast users in
United States, the share of ear is 28% (vs 4% for all
audience), and millennials AOD users based on
Westwood One study spending 6.4 hours per week
consuming audio podcast or translate to 55 minutes a
day [18]. This prove the wave of listening audio on
demand is correspond to Jakarta urban millennials.
Table 1. Table Respondents
Age
25.2
Male
86
41.7%
Female
120
58.3%
Students
83
40.3%
Workers
96
46.6%
Average of Audio Exposure per
day
4.8
Hours
288
Minutes
Average of Podcast
Listening
45
Minutes
***(own data: 1 Final CD Original DATA CD TEST
JS_ContentDensity_rev3 copy.xlsx)
Another interesting fact is the tools adoption (include
application or website) used to explore the audio on
demand among millennials, based on Westwood One
comparing to this research are as follow [19]:
Table 2. Table tools adoption
Name of Tools
This Study
US Millennials
Spotify
89%
43%
YouTube
44%
40%
Joox
24%
-
Apple Podcast
16%
37%
Source: processed data (2021)
4 Analysis and Discussion
4.1 Goodness of Fit Model
In testing the suitability of the model (Goodness of Fit
– Model), there are 3 types of tests including the
suitability test of the measurement model (outer
model), structural model fit test (inner model) and
Evaluation of Goodness of Fit (GoF) – Inner Model.
4.2 Outer Model
For testing the outer model by looking at the results of
convergent validity and discriminant validity.
Convergent validity is measured by loading factor and
Average Variance Extracted (AVE) parameters;
Construct Reliability is measured by Cronbach’s
Alpha and Composite Reliability parameters. An
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
729
Volume 19, 2022
indicator is declared valid if it has a loading factor
above 0.70 for the intended construct while the
Average Variance Extracted (AVE) value generated
by all constructs must be above > 0.5. And to see the
results of construct reliability, the Cronbach's Alpha
and Composite Reliability values must be above 0.7
[20].
Table 3. Reliability and Validity Test Results
Source: processed data (2021)
Furthermore, discriminant validity is used to test
the validity of a model, measured by the cross-loading
value parameter which contains the magnitude of the
correlation between constructs and their indicators and
indicators from other constructs. The cut off value
used for cross loading must be greater than 0.70. The
discriminate validity test using the AVE value is
carried out by comparing the root value of the AVE of
each construct with the correlation between the
constructs and other constructs. It is recommended
that the AVE value should be greater than 0.50 [21].
Table 3 shows that the root value of the AVE of each
construct is greater with the correlation between
constructs and other constructs. So it can be
concluded that it has good discriminant validity.
4.3 Inner Model
Assessing the inner model is to see the relationship
between latent constructs by looking at the estimation
results of the path parameter coefficients and their
level of significance. In the assessment of the inner
model by looking at the R-square for each dependent
latent variable[22]. The following is the R-square
value in the construct:
Table 4. R-Square
Dependent variable
R Square
Attitude
0.364
Listener Loyalty
0.384
Perceived Control
0.068
Subjective Norm
0.048
Source: processed data (2021)
In the table ***4.13 the R-Square value for the
Attitude construct gives a result of 0.364. This means
that the Attitude construct can be explained by the
Exposure Frequency, Subjective Norm and Perceived
Control constructs of 36.4%.
The R-Square value of Listener Loyalty construct
is 0.384. This means that the construct of Listener
Loyalty is explained by the constructs of Attitude,
Subjective Norm, Exposure Frequency and Perceived
Control of 38.4%.
The R-Square value of the Perceived Control
construct is 0.068. This means that the construct of
Perceived Control is explained by Subjective Norm
and Exposure Frequency of 6.8%. And finally, the R-
Square value of the Subjective Norm construct is
0.048. This means that the Subjective Norm construct
is explained by the Exposure Frequency construct of
4.8%.
4.4 GOF Evaluation of Inner Model
Research
constructs and
research items
Loading
factor
> 0.5
Composite
reliability >
0.7
Ave. var.
extracted
>0.5
Exposure
Frequency
0.875
0.585
EF101
0.715
EF102
0.828
EF103
0.770
EF104
0.794
EF105
0.710
Usefulness
0.807
0.926
AT101
0.918
AT102
0.917
AT103
0.859
Genre
0.571
0.842
AT111
0.731
AT112
0.797
AT113
0.756
AT114
0.738
Ease of Use
0.634
0.896
PC101
0.766
PC102
0.866
PC103
0.847
PC1041
0.768
PC1042
0.726
Auto Suggestions
0.668
0.857
PC201
0.875
PC202
0.863
PC203
0.703
Social Norm
0.650
0.902
SN01
0.713
SN02
0.823
SN03
0.762
SN04
0.836
SN05
0.885
Continuance
0.801
0.941
LL101
0.920
LL102
0.946
LL103
0.906
LL104
0.802
Advocating
0.597
0.816
LL201
0.764
LL202
0.794
LL203
0.759
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
730
Volume 19, 2022
To evaluate the GoF – Inner Model, it is done by
looking at the values of R-Square (R2), Q-Square (Q2)
and GoF. Following is the test of Inner Model with Q2
(predictive relevance) :
Q2 = 1 – (1 – R12) (1 – R22) (1 – R32) (1 – R42)
Q2 = 1 – (1 – 0.364) (1 – 0.384) (1 – 0.068) (1 –
0.048)
Q2 = 1 – 0.348
Q2 = 0,652
From these results in Q2 (0.652) > 0, it shows that
the model has predictive relevance [22]. The Q2
predictive relevance value of 0.652 indicates that the
model is strong.
Next, look for the Goodness of Fit (GoF), which
is calculated by the square root of the average
communality index value with the R-Square (R2)
average.
GoF =
The Com value is obtained from the average
AVE value of 0.72 so that the GoF value can be
calculated as follows:
GoF =
GoF =
GoF = 0,566
The GoF value of 0.566 indicates that the model
in this study is included in the Strong criteria.
4.5 Hypotheses Testing and Discussions
To test the hypothesis using the output path
coefficients (Mean, STDEV, T-Values) provided that if
the t statistic value obtained from the table is greater
than 1.96 and the p-value <0.05, the hypothesis
between the existing variables is accepted. On the
other hand, if the value of t statistics is less than 1.96,
the p-value > 0.05, then the hypothesis is rejected.
There are two submodels in a structural equation
model; the inner model specifies the relationships
between the independent and dependent latent
variables, whereas the outer model specifies the
relationships between the latent variables and their
observed indicators. In SEM, a variable is either
exogenous or endogenous. An exogenous variable has
path arrows pointing outwards and none leading to it.
Meanwhile, an endogenous variable has at least one
path leading to it and represents the effects of other
variable(s).
Table 5. Path coefficients (Mean, STDEV, T-Values)
– 1
Original
Sample
(O)
T
Statistic
s
(|O/STE
RR|)
P-Value
Result
Direct Effect
Exposure
Frequency ->
Attitude
0.124
2.037
0.042
Significant
H4
Exposure
Frequency ->
Subjective
Norm
0.220
3.537
0.000
Significant
H5
Exposure
Frequency ->
Listener
Loyalty
0.019
0.293
0.769
Not
significant
H6
Exposure
Frequency ->
Perceived
Control
0.109
1.576
0.115
Not
significant
H7
Attitude ->
Listener
Loyalty
0.450
6.701
0.000
Significant
H8
Subjective
Norm ->
Attitude
0.123
2.024
0.043
Significant
H9
Subjective
Norm ->
Listener
Loyalty
0.124
1.782
0.075
Not
significant
H10
Subjective
Norm ->
Perceived
Control
0.214
2.519
0.012
Significant
H11
Perceived
Control ->
Attitude
0.526
8.964
0.000
Significant
H12
Perceived
Control ->
Listener
Loyalty
0.154
2.213
0.027
Significant
H13
Indirect Effect
Exposure
Frequency ->
Attitude ->
Listener
Loyalty
0.056
2.011
0.044
Significant
Exposure
0.027
1.549
0.122
Not
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
731
Volume 19, 2022
Frequency ->
Subjective
Norm ->
Listener
Loyalty
significant
Exposure
Frequency ->
Perceived
Control ->
Listener
Loyalty
0.017
1.171
0.242
Not
significant
Total Indirect Effect
Exposure
Frequency ->
Listener
Loyalty
0.156
3.717
0.000
Significant
Subjective
Norm ->
Listener
Loyalty
0.139
2.977
0.003
Significant
Perceived
Control ->
Listener
Loyalty
0.236
5.346
0.000
Significant
Source: processed data (2021)
Thus, the SEM Path Diagram will be:
Hypothesis 1 : AOD exposure frequency is positively
related to individual attitude toward AOD loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (2.037) > 1.96 and P-Value (0.042) < 0.05, and
the original sample value is 0.124 (positive sign)
From these results, the hypothesis which states that
the AOD exposure frequency is positively related to
individual attitude is received, the higher the AOD
exposure frequency, the higher the individual attitude.
Meanwhile, to determine the indirect effect (AOD
exposure frequency is positively related to individual
attitude toward AOD loyalty) the T-statistical value
(2.011) > 1.96 and P-Value (0.044) < 0.05, and the
original sample value of 0.056 (positive sign). From
these results, it can be stated that there is an indirect
effect of AOD exposure frequency is positively
related to individual attitude toward AOD loyalty.
This is an important finding, in contrary of Mou
which the communication frequency do not have
direct effect toward attitude except via subjective
norm, in this case the audio exposure has determinant
effect for the listener's attitude [3]. But this one is
making more sense, it is like a listening habit once,
someone accustomed to specific background audio.
Respondents also stated that they were aware of the
audio exposure they encountered, with an average of
4.8 hours. And the average minute consuming Podcast
or Video Cast is 45 minutes per day on average or
about 15.6% using the “share of ear” method.
Hypothesis 2: AOD exposure frequency is positively
related to subjective norm toward AOD loyalty.
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (3.537) > 1.96 and P-Value (0.001) < 0.05, and
the original sample value is 0.22 (signed from these
results, the hypothesis states that the AOD exposure
frequency is positively related to subjective norm is
received. The higher the AOD exposure frequency,
the higher the subjective norm.
Meanwhile, to determine the indirect effect (AOD
exposure frequency is positively related to subjective
norm toward AOD loyalty) the T-statistical value
(1.549) < 1.96 and P-Value (0.122) > 0.05, and the
original sample value of 0.027 (positive sign). From
these results, it can be stated that there is no indirect
effect of AOD exposure frequency related to
subjective norm toward AOD loyalty.
This is corresponding to Mou finding that
subjective norms is being dependent on the frequency
of conversation, in this case is audio exposure [3].
From then on, the barrier between the usage of radio,
as is using the radio frequency and on demand audio-
based media, such as an example of podcasting, is
narrowing. Added with the trend that most
conventional media usage is in decreasing trend, the
audio-based media is increasingly up[28].
Hypothesis 3: AOD exposure frequency will
positively relaed to individual's AOD Loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (0.293) < 1.96 and P-Value (0.769) > 0.05, and
the original sample value is 0.019 (positive sign)
From these results, the hypothesis which states that
AOD exposure frequency will be positively related to
individual's AOD Loyalty is rejected. AOD exposure
frequency has a positive but not significant effect on
individual's AOD Loyalty.
This hypothesis is derived from Mou finding that
communication frequency regarding podcast have
positive affect toward the intention to listen and with
addition of the way Edison Research data came from
audio exposure frequency, the result is not proven in
this test [18,3]. Hanyang Luo introduced different
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
732
Volume 19, 2022
website type as moderating variable to explore
website quality whether has different influence on
user loyalty related to different website styles, this
object can make up the gap that former research only
chose one kind of website. In this paper also using the
Content Density as the “Website Information Quality”
variable to see how ‘hedonism’ content affecting user
loyalty[29].
Hypothesis 4: AOD exposure frequency is positively
related to individual perceived behavior control
toward AOD loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (1.576) < 1.96 and P-Value (0.115) > 0.05, and
the original sample value is 0.109 (positive sign)
From these results, the hypothesis which states that
AOD exposure frequency is positively related to
individual perceived behavior control is rejected.
Meanwhile, to determine the indirect effect (AOD
exposure frequency is positively related to individual
perceived behavior control toward AOD loyalty), the
T-statistical value (1.171) < 1.96 and P-Value (0.242)
> 0.05, and the original sample value of 0.017
(positive sign). From these results, it can be stated that
there is no indirect effect of AOD exposure frequency
is positively related to individual perceived behavior
control toward AOD loyalty.
A number of experiments in this place received
extensive media attention and also the usually
promulgated findings are that individuals are terrible
multitaskers, that folks are powerless to multitask, or
even that multitasking actively damages concentration
and cognition [12]. The assumption of assuming
exposure frequency will affect the tools adoption is
not proven, with the advancement of technology of
recommendation service and great user interface and
experience from tools toward millenials stating
actually the other way around. Tools that bring the
audio exposure [17,23].
Hipotesis 5: Attitude toward AOD is positively
related to AOD-Loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (6.701) > 1.96 and P-Value (0.000) < 0.05, and
the original sample value is 0.45 (signed From these
results, the hypothesis which states that Attitude
toward AOD is positively related to AOD-Loyalty is
received. Attitude toward AOD has a positive and
significant effect on AOD-Loyalty. The higher the
Attitude toward AOD, the higher the AOD-Loyalty.
Attitude is always a determining factor toward
behavior, this study same as Mou and Lyn, the direct
and indirect affects toward behavior is through
Attitude as on the driving variables [3]. Hanyang Luo
whose research was based on the literature relevant to
the connection between site quality and users' attitude
or behavior, we adopt the technology acceptance
model (TAM) and stimulus – organism - response (S-
O-R) principle to create a conceptual design, learning
how the site quality impact user devotion through
users' perceived commitment and value through
hedonism value [24].
Hypothesis 6: Perceived subjective norm associated
toward AOD is positively related to AOD loyalty
attitude
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (2.024) > 1.96 and P-Value (0.043) < 0.05, and
the original sample value is 0.123 (positive sign)
From these results, the hypothesis states that
Perceived subjective norm associated toward AOD is
positively related to AOD loyalty attitude is received.
Perceived subjective norm has a positive and
significant influence on AOD loyalty attitude. The
higher Perceived subjective norm, the higher AOD
loyalty attitude will be. Another conformation on the
previous study of Mou, which social norm have direct
confirmatory effect toward the attitude, such as peer
pressure [3,24]. Mou stated Lee’s study as “By the
same token, a study focusing on the intention to play
online games suggested that player attitude, perceived
subjective norms, and perceived behavioral control, in
addition to flow experience and perceived enjoyment,
all had an influence on an individual’s intention to
continue playing these online games”[24].
Hypothesis 7: Perceived subjective norm associated
with AOD use is positively related to individual’s
AOD loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (1.782) < 1.96 and P-Value (0.075) > 0.05, and
the original sample value is 0.124 (positive sign)
From these results, the hypothesis which states that
Perceived subjective norm associated with AOD use
is positively related to individual's AOD loyalty is
rejected. Perceived subjective norm has a positive but
insignificant effect on individual's AOD loyalty. The
descriptive norm instead of injunctive norm as part of
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
733
Volume 19, 2022
the subjective norm is shown the positive effect
toward behavior, so this result is more relates to
injunctive norms in Mou's Theory of Planned
Behavior model [3]. The past decade indicates a sharp
increase in investigation examining people's capacity
to multitask. A fair level of that analysis is centered
on a certain kind of multitasking named media
multitasking[24].
Hypothesis 8: Perceived subjective norm associated
toward AOD is positively related to AOD loyalty
perceived control.
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (2.519) > 1.96 and P-Value (0.012) < 0.05, and
the original sample value is 0.214 (positive sign)
From these results, the hypothesis which states that
Perceived subjective norm associated toward AOD is
positively related to AOD loyalty perceived control is
received. Perceived subjective norm has a positive and
significant influence on AOD loyalty perceived
control. The higher the Perceived subjective norm, the
higher AOD loyalty perceived control. Social norm is
driving force for tools recommendation and this one is
correspond toward Mou finding in injunctive norm as
well as the report from Westwood One on how
listeners get their tips for podcast exploration [3,19].
Given these two components, the authors can certainly
then define success by attaining some level of overall
performance in the multitasking condition with
respect to the control condition. To have the ability to
evaluate results across studies and theoretical
approaches, the studies selected for this analysis and
reported here use any activity problem while the
control condition.
Hypothesis 9: Perceived behavior control associated
toward AOD is positively related to AOD loyalty
attitude.
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (8.964) > 1.96 and P-Value (0.001) < 0.05, and
the original sample value is 0.626 (positive sign)
From these results, the hypothesis which states that
the Perceived behavior control associated toward
AOD is positively related to AOD loyalty attitude is
received. The higher the Perceived behavior control,
the higher the AOD loyalty attitude.This hypothesis
confirming also previous study from Mou [3]. Nic
Newman stating this changing behavior are for the
advantages for audio media due to all this trend in
hearables equipment will improve the interruption of
radio schedules and activate more on demand audio
usage[19].
Hypothesis 10 : Perceived Behavior Control over
AOD is positively related to AOD Loyalty
From the table ***4.20 (Path coefficients (Mean,
STDEV, T-Values) it is found that the T-statistical
value (2.213) > 1.96 and P-Value (0.027) < 0.05, and
the original sample value is 0.154 (positive sign)
From these results, the hypothesis which states that
Perceived Behavior Control over AOD is positively
related to AOD Loyalty is received. Perceived
Behavior Control has a positive and significant
influence on AOD Loyalty. The higher the Perceived
Behavior Control, the higher the AOD Loyalty. In
contrary from Mou's finding, perceived behavior
control in this variable is looking the recommendation
service for helping the users find what they want.
With the advancement of recommendation
technology, this might change toward the behavior of
users by getting better recommendation at the time of
the research [3,17]. Many study is either ignoring the
variety of streaming audio in general and treated as
digital channel or becoming too niche in studying one
sub channel of AOD and ignoring the others [30].
Geographical location also affecting the availability of
variation of AOD, for example, satellite audio stream
play a big roles for coast to coast driving commuter in
North America.
5 Conclusion
Based on the analysis and discussion above, it can be
concluded that the frequency of AOD exposure is
positively related to the individual's attitude of being
accepted, that there is an indirect effect that the
frequency of AOD exposure is positively related to
the individual's attitude towards AOD loyalty. While
the frequency of exposure to AOD is positively
related to accepted subjective norms. From these
results it can be stated that there is no indirect effect
of the frequency of AOD exposure related to
subjective norms on AOD loyalty. While the
frequency of AOD exposure is positively related to
individual AOD Loyalty is rejected, the frequency of
AOD exposure has a positive but not significant effect
on individual AOD Loyalty. From these results, the
hypothesis which states that the frequency of AOD
exposure is positively related to individual perceived
behavioral control is rejected. From these results it
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
734
Volume 19, 2022
can be stated that there is no indirect effect of the
frequency of AOD exposure that is positively related
to individual perceived behavioral control on AOD
loyalty. While the attitude towards AOD is positively
related to AOD-Loyalty is accepted. Attitude towards
AOD has a positive and significant effect on AOD-
Loyalty. From these results, the hypothesis states that
Perceived subjective norm is positively related to the
attitude of loyalty AOD is accepted. Perceived
subjective norm has a positive and significant effect
on AOD loyalty attitudes. From these results the
hypothesis which states that the perception of
subjective norms related to the use of AOD is
positively related to individual AOD loyalty is
rejected. Perceived subjective norms have a positive
but not significant effect on individual AOD loyalty.
Perceived subjective norm is positively related to
AOD loyalty perceived control is accepted. Perceived
subjective norm has a positive and significant effect
on AOD loyalty perceived control. From these results
obtained a hypothesis which states that Perceived
behavior control related to AOD is positively related
to AOD loyalty attitudes. From these results obtained
a hypothesis which states that Perceived Behavior
Control over AOD is positively related to AOD
Loyalty. Perceived Behavior Control has a positive
and significant effect on AOD Loyalty.
5.1 Theoretical Implication
The perceived behavior control is a perception of user
experience toward the control factors within
experience, in this case, this study might give some
exploratory study toward the tools adoption,
especially the research recommendation service
satisfaction, user interface, and user experience.
The recommendation service is one of the strong
points on tools adoption. It has been known in
predicting what people want to see, the same
algorithm that start the functionality from ads
targeting, but now more toward the content
recommendation also to increase the time-spent of
stickiness. The best system for multimedia content
recommendation is YouTube, Netflix, Pandora, and
Spotify, and it expanding to another platform as well.
Behind the deep learning capabilities powered
with artificial intelligent, are an abundant of big data
in term of variety and velocity of it for example
YouTube current published algorithm include three
parts:
Scale: Many current recommendation
algorithms shown to work nicely on problems
that are small fail to run on the scale of
YouTube. Extremely specialized distributed
learning algorithms and cost-efficient serving
methods are important for managing
YouTube's significant pc user.
Freshness: YouTube has an extremely
powerful corpus with a lot of hours of video
are uploaded per next. The recommendation
product must be responsive adequate to model
freshly uploaded content along with the
newest actions taken by the computer user.
Balancing brand new content with well-
established videos can be known from an
exploration/exploitation viewpoint.
Noise: Historical pc user conduct on YouTube
is inherently hard to predict because of
sparsity and an assortment of unobservable
outside factors. We rarely get the soil fact of
consumer pleasure plus as an alternative unit
loud implicit feedback signal.
According to Karen Hao[3] among the proposed
update for YouTube algorithm, the scientists
specifically focus on an issue they recognize as
"implicit bias." It describes how recommendations
themselves can impact operator behavior, making it
difficult to decipher whether you clicked on a video
and audio recording since you liked it or perhaps since
it had been highly recommended. The impact is the
fact that over time, the device is able to push users
even further and further separate from the movies they
want to watch or listen, and this is also explain even
though YouTube is intentionally serving for video, but
it is a popular knowledge, that people are mostly
listening to the video while there eyes are watching
the second screen and the algorithm to find that
"implicit bias" will cover the audio channels also.
Westwood One even reporting that many podcast
users are listening and discovery through YouTube
along with other podcast platform that they use.
This research also find that YouTube is the
number two choice among Millennials in Jakarta
Indonesia to explore and finding the audio-based
content. With this content density metric approach, it
will be easier for the recommendation service to
suggest the right content based on the noise and multi-
media tasking that the users are encounter. And this
also correspond to this finding that tools adoption
(perceived behavior control) has significant effect
toward loyalty.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
735
Volume 19, 2022
5.2 Managerial Implication
This study confirms that tools adoption via tools
adoption is the important aspect that affecting
listeners perception toward the content or its loyalty to
the content itself. We had learned how YouTube
create a rabbit-hole for its users, and YouTube also
becoming the major player in audio-based media
indirectly due to its highly contagious
recommendation system. Secondly, the peer pressure
through social norm, is also affecting attitude and
tools adoption, in this case big players are having both
capabilities: managing big users based to create social
norm effects, while they also sharpening the big data
usage to improve the personalization to the users.
Now, as content owners who do not have those
capabilities, the best way is to adapt and distribute the
content to a larger hub. We can learn from the podcast
industry that when IHeartMedia use its platform to be
exclusive platform for its exclusive contents, but
finally it changes the strategy to expanding its content
through other application tools also, naming Spotify,
Apple Podcast, Pocket Cast (own research). For this
reason, the strategy of content distribution to popular
hub or content aggregation platform is one of the keys
to get more audience.
Podcast has been popular in United States because
of the popularity of Podcast, audio and mobility is
corelating tightly, and with audio content that can
have the time-shift ability, when Apple put the
podcast as part of the iTunes software that can be
synchronized to the iPod (which inspiring the name of
Podcast), the popularity of podcast becoming more
mainstream. In Indonesia, podcast became popular
just the same time Spotify became popular in
Indonesia, within short time Podcast becoming the
household talks in many radio lovers in Indonesia.
With the same analogy of chicken or egg that
comes first, we can see that between exposure
frequency or tools adoption, the research shows that
tools adoption has direct effect through loyalty of
users, while exposure frequency has indirect effect to
loyalty via social norm then tools adoption. In this
case tools adoption is very important factors that
hooked listeners to content loyalty.
This research also shows that there is peer
pressured toward the using of certain applications, and
the tools adoption also will influence the attitude
which then will lead to loyalty to listen more on
involved in advocating the channels that they listen.
And since tools adoption is also influencing the
attitude of users, through user experience, this will
have double advantage when the media take
advantage of available tools effectively to expand
their media reach.
References:
[1] Markman, Kris M., and Caroline E. Sawyer.
"Why pod? Further explorations of the
motivations for independent podcasting." Journal
of Radio & Audio Media 21.1, 2014,20-35.
[2] Markman, Kris M. "Considerations—Reflections
and future research. Everything old is new again:
podcasting as radio's revival." Journal of Radio
& Audio Media 22.2, 2015, 240-243.
[3] Mou, Yi, and Carolyn A. Lin. "Exploring podcast
adoption intention via perceived social norms,
interpersonal communication, and theory of
planned behavior." Journal of Broadcasting &
Electronic Media 59.3,2015, 475-493.
[4] Bressan, Federica. "A framework for the
description of age-related symptoms in audio
media: Definition and implementation." Studies
in Conservation 65.4, 2020,189-199.
[5] Cebeci, Zeynel, and Mehmet Tekdal. "Using
podcasts as audio learning
objects." Interdisciplinary Journal of E-Learning
and Learning Objects 2.1,2006,47-57.
[6] Ractham, Peter, and Xuesong Zhang.
"Podcasting in academia: A new knowledge
management paradigm within academic
settings." Proceedings of the 2006 ACM SIGMIS
CPR conference on computer personnel
research: Forty four years of computer personnel
research: achievements, challenges & the future.
2006.
[7] Cook, Ian M. "Critique of podcasting as an
anthropological method." Ethnography,
2020,1466138120967039.
[8] Llinares, Dario, Neil Fox, and Richard Berry.
"Introduction: Podcasting and podcasts—
Parameters of a new aural culture." Podcasting.
Palgrave Macmillan, Cham, 2018,1-13.
[9] Llinares, Dario, Neil Fox, and Richard Berry,
eds. Podcasting: New aural cultures and digital
media. Springer, 2018.
[10] Ajzen, Icek. "Attitude personality and behaviour
2nd." New York: Open University Pres (2005).
[11] Knabe, Ann Peru. Applying Ajzen's theory of
planned behavior to a study of online course
adoption in public relations education.
Marquette University, 2012.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
736
Volume 19, 2022
[12] Ndubisi, Nelson. "Factors of online learning
adoption: A comparative juxtaposition of the
theory of planned behaviour and the technology
acceptance model." International Journal on E-
learning 5.4, 2006,571-591.
[13] Yu, Tai‐Kuei, and Tai‐Yi Yu. "Modelling the
factors that affect individuals’ utilisation of
online learning systems: An empirical study
combining the task technology fit model with the
theory of planned behaviour." British Journal of
Educational Technology 41.6 (2010): 1003-1017.
[14] Xigen, L. I., and Li Zeng. "Technology
Attributes, Perceived Value and Quality of
Information, and Social Utility as Predictors of
Podcast Adoption and Use." 59th Annual
Conference of the International Communication
Association (ICA). 2009.
[15] Moss, Nathan D., Erin L. O’Connor, and
Katherine M. White. "Psychosocial predictors of
the use of enhanced podcasting in student
learning." Computers in Human Behavior 26.3,
2010,302-309.
[16] Havlena, William, Robert Cardarelli, and
Michelle De Montigny. "Quantifying the isolated
and synergistic effects of exposure frequency for
TV, print, and internet advertising." Journal of
Advertising Research 47.3, 2007, 215-221.
[17] Hao, Karen. "YouTube is experimenting with
ways to make its algorithm even more
addictive." MIT Technology Review,
September 27, 2019,2019.
[18] Edison Research. THE PODCAST CONSUMER
2017 #PodCon17. Retrieved from
http://www.edisonresearch.com/wp-
content/uploads/2017/04/Podcast-Consumer-
2017.pdf
[19] Westwood One. Westwood One’s Podcast
Download Fall 2020 Report. In WestWood One.
Retrieved from
https://www.westwoodone.com/wp-
content/uploads/2020/09/Westwood-Ones-
Podcast-Download-Fall-2020-Report.pdf
[20] Raykov, Tenko. "Estimation of composite
reliability for congeneric measures." Applied
Psychological Measurement 21.2, 1997,173-184.
[21] Hair Jr, Joe F., et al. "PLS-SEM or CB-SEM:
updated guidelines on which method to
use." International Journal of Multivariate Data
Analysis 1.2, 2017,107-123.
[22] Ardi, Noper. "Structural Equation Modelling-
Partial Least Square to Determine the Correlation
of Factors Affecting Poverty in Indonesian
Provinces." IOP Conference Series: Materials
Science and Engineering. Vol. 846. No. 1. IOP
Publishing, 2020.
[23] Punnett, Ian. "Digital is the future—and the now:
EmPosium on digital platform listening trends."
(2016): 4-19.
[24] Christian, J., A. Rivis, and P. Sheeran.
"Descriptive norms as an additional predictor in
the theory of planned behavior: A meta-
analysis." Planned Behavior 22.3, 2018,43-62.
[25] Carey, John. "Audience measurement of digital
TV." International journal of digital
television 7.1, 2016,119-132.
[26] Boling, Kelli S., and Kevin Hull. "Undisclosed
information—Serial is My Favorite Murder:
Examining motivations in the true crime podcast
audience." Journal of Radio & Audio Media 25, no. 1
(2018): 92-108.
[27] Bhatt, Swati. The attention deficit: unintended
consequences of digital connectivity. Springer Nature,
2019.
[28] Lee, Mark JW, Charlynn Miller, and Leon Newnham.
"Podcasting syndication services and university
students: Why don't they subscribe?." The Internet
and Higher Education 12, no. 1 (2009): 53-59.
[29] Luo, Hanyang, Xinwei Han, and Yanan Yu. "The
impact of website quality on user loyalty through
perceived value and commitment." In 2016 13th
International Conference on Service Systems and
Service Management (ICSSSM), pp. 1-5. IEEE, 2016.
[30] Bonini, Tiziano. "Doing radio in the age of
Facebook." Radio Journal: International Studies in
Broadcast & Audio Media 12, no. 1-2 (2014): 73-87.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
Jerry S Justianto is the main author of this article
and also plays a role in analyzing the statistical data
generated in this study.
M Arief is an expert in the field of research
management, he is very helpful in providing input and
input in this research so that it can produce quality
research.
Indah Susilowati is an expert in the field of
economics and research management, she is also an
expert in the field of economics so she is very
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
737
Volume 19, 2022
instrumental in providing input on the use of theories
in this research in order to produce quality research.
Muhamad Aras is an expert in marketing
communications, online behavior and social
communication, he plays a role in providing input and
input in this research so that it can produce quality
research.
Sources of Funding for Research Presented in
a Scientific Article or Scientific Article Itself
All funding in this study came from private
funding.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en_
US
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2022.19.64
Jerry S. Justianto, Mts Arief,
Indah Susilowati, Muhamad Aras
E-ISSN: 2224-2899
738
Volume 19, 2022
