ArticlePDF Available

The second-level smartphone divide: A typology of smartphone use based on frequency of use, skills, and types of activities

Authors:

Abstract and Figures

Smartphones have become one of the most frequently used devices for accessing the Internet. Although a growing body of research has examined the second-level digital divide with a focus on general Internet and digital media use, little is known about patterns of smartphone use and smartphone-related skills in the general population. This paper examines inequalities in the use of smartphone technology based on two nationally representative samples of smartphone owners collected in Germany in 2017 and 2020. We identify six distinct types of smartphone users by conducting latent class analyses that classify individuals based on their frequency of smartphone use, self-rated smartphone skills, and activities carried out on their smartphone. Smartphone use differs significantly by sociodemographic characteristics and operating system. The types reflecting more frequent and diverse smartphone use are younger, have higher levels of educational attainment, and are more likely to use an iPhone. Overall, the composition of the latent classes and their characteristics are robust across samples.
Content may be subject to copyright.
The second-level smartphone
divide: A typology of
smartphone use based on
frequency of use, skills, and
types of activities
Alexander Wenz
Mannheim Centre for European Social Research, University of
Mannheim, Germany
Florian Keusch
School of Social Sciences, University of Mannheim, Germany
Abstract
Smartphones have become one of the most frequently used devices for accessing the
Internet. Although a growing body of research has examined the second-level digital div-
ide with a focus on general Internet and digital media use, little is known about patterns
of smartphone use and smartphone-related skills in the general population. This paper
examines inequalities in the use of smartphone technology based on two nationally rep-
resentative samples of smartphone owners collected in Germany in 2017 and 2020. We
identify six distinct types of smartphone users by conducting latent class analyses that
classify individuals based on their frequency of smartphone use, self-rated smartphone
skills, and activities carried out on their smartphone. Smartphone use differs signicantly
by sociodemographic characteristics and operating system. The types reecting more
frequent and diverse smartphone use are younger, have higher levels of educational
attainment, and are more likely to use an iPhone. Overall, the composition of the latent
classes and their characteristics are robust across samples.
Keywords
Digital divide, digital inequalities, smartphone, latent class analysis, user typology, survey
Corresponding author:
Alexander Wenz, Mannheim Centre for European Social Research, University of Mannheim, 68131 Mannheim,
Germany.
Email: a.wenz@uni-mannheim.de
Original Article
Mobile Media & Communication
125
© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/20501579221140761
journals.sagepub.com/home/mmc
Smartphones have become an integral part of peoples daily lives, with 85% of U.S.
adults owning a smartphone in 2021, compared with only 35% a decade ago (Pew
Research Center, 2021). Smartphone ownership has similarly increased across other
countries. For example, Eurostat (2021) shows that the proportion of people in
Germany aged 16 and over who use a smartphone or other mobile phone to access the
Internet increased from 15% in 2011 to 75% in 2019. Although smartphone ownership
rates are still lower in emerging economies, they are rapidly increasing (Silver, 2019).
Despite this general development, inequalities in smartphone access persist (van
Deursen & van Dijk, 2019). Recent data from 2021 show that 96% of U.S. adults
under the age of 30 own a smartphone, whereas only 61% of adults aged 65 and
above own such a device. Smartphone ownership increases with household income
and educational attainment, and U.S. adults living in rural areas are less likely to own
a smartphone than adults living in urban or suburban areas (Pew Research Center,
2021). Similar patterns exist in other countries (Eurostat, 2021; Keusch et al., 2020).
Beyond disparities in digital access, scholars have increasingly paid attention to inequal-
ities in digital skills and the use of digital technologies, referred to as the second-level
digital divide (DiMaggio et al., 2004; Hargittai, 2002). Even among people with
access to the Internet and digital devices, substantial differences can be found with
regard to how skilled and experienced they are in using digital technologies (van Dijk,
2020).
Although a growing body of research has examined the second-level digital divide
with a focus on general Internet and digital media use (e.g., Blank & Groselj, 2014;
Reisdorf & Groselj, 2017; van Deursen & van Dijk, 2014), there is a rather limited under-
standing in such research of inequalities related to smartphone skills and use. Our study
contributes to lling this gap by providing empirical evidence about the second-level
smartphone divide in Germany. We consider this extension important, because over
the last decade smartphones have become one of the most frequently used devices for
accessing the Internet, and for some parts of the population the only means by which
they can do so (Eurostat, 2021; Pew Research Center, 2021; Tsetsi & Rains, 2017). In
addition, smartphones differ in several characteristics from desktop and laptop compu-
ters, most notably in their portability, because smartphone users can easily access the
Internet in their everyday lives whenever they carry around their mobile device. Thus,
the ability to use a smartphone effectively and efciently becomes increasingly important
as the technology gets more embedded into peoples lives. Smartphones can be used for a
large variety of activities including communication and entertainment, but also for activ-
ities that were previously performed mainly on desktop or laptop computers, such as
information seeking or work-related activities (Pearce & Rice, 2013; Perrin, 2017).
Furthermore, there is a growing array of activities that rely exclusively on smartphones
and, thus, cannot be performed on other digital devices. Examples include communica-
tion (e.g., using communication applications such as WhatsApp), commercial activities
(e.g., using mobile banking or mobile payments), mobility (e.g., buying train or plane
tickets or participating in public car-sharing or bicycle-sharing schemes), and
health-related activities (e.g., digital contact tracing or using digital vaccination certi-
cates). Smartphones are also increasingly used as a hub for Internet of Things devices
such as wearable tness trackers and smart home applications.
2Mobile Media & Communication 0(0)
People who lack the required usage experience and skills are less likely to benet
from the opportunities offered by smartphones and are increasingly excluded from the
digital society (van Deursen & Helsper, 2015). Individuals who have difculties
installing new apps on their smartphone, for example, may be unable to use services
that are only accessible via particular smartphone apps. Previous research has
shown that inequalities in general Internet and digital media use are directly linked
to existing social inequalities, including those related to socioeconomic status, educa-
tion, gender, ethnicity, and age (Helsper, 2021; Robinson et al., 2015; van Dijk, 2020).
However, much less is known about the patterns of smartphone use and
smartphone-related skills in the general population and how they relate to social
inequalities.
To shed more light on these inequalities in the use of smartphone technology, we draw
on nationally representative survey data from two independent samples of smartphone
owners collected in Germany in November 2017 and January 2020 and address the fol-
lowing two research questions:
RQ1: Which smartphone usage types can be identied among smartphone owners?
RQ2: How do the smartphone usage types differ in sociodemographic characteristics and in
the operating system used?
With this study, we contribute to the digital divide literature in several ways. First, we
apply the theoretical concept of the digital divide specically to smartphones. We build
on and extend the existing literature on the digital divide with a particular focus on the
skills that smartphone owners report when using their device for different types of activ-
ities in their everyday lives. Second, we provide empirical evidence for the existence of
different smartphone user groups that are clearly differentiated in their activity patterns
and skill sets. Third, measuring the same constructs in two samples allows us to
examine whether the identied smartphone usage types and their correlates are robust
across different time points.
Background
Typologies of digital media and smartphone use
An extensive number of typologies have been constructed over the last two decades to
conceptualize and classify the large variation in Internet and digital media use (Blank
& Groselj, 2014). To create the usage typologies, previous studies often classied
Internet users based on their amount, variety, and types of use (Blank & Groselj,
2014; Holmes, 2011; Horrigan, 2007; Livingstone & Helsper, 2007; Reisdorf &
Groselj, 2017; Selwyn et al., 2005; van Deursen & van Dijk, 2014; Zillien &
Hargittai, 2009). Whereas amount of use is measured as frequency of going online
(e.g., number of hours per day), frequency of engaging in different online activities, or
total years of Internet use, variety of use is measured as the number of activities
Wenz and Keusch 3
carried out online. In turn, types of use are measured by nominal categories that represent
different online activities such as information seeking and communication.
For example, in one of the earlier studies, Selwyn et al. (2005) interviewed 1,001
adults in four regions of England and Wales and identied four categories of Internet
users based on their frequency and type of Internet use: (a) broad frequent users
(13% of the sample); (b) narrow frequent users(18%); (c) occasional users(11%);
and (d) non-users(58%). In a more recent study, Reisdorf and Groselj (2017) use
data from the Oxford Internet Survey, a nationally representative face-to-face survey
of individuals aged 14 and over in Great Britain, to identify ve types of Internet
users: (a) broad users(27% of the population); (b) regular users(30%); (c) low
users(21%); (d) non-users(18%); and (e) ex-users(4%).
Although a large number of such typologies have been proposed for general Internet
and digital media use, smartphone use has received less attention in previous research on
the digital divide. In a few of the existing studies, mobile phone and smartphone use has
been employed as an indicator for general Internet use (Herzing & Blom, 2019; Horrigan,
2007; Yates et al., 2020), but has not been studied on its own. In related areas of research,
typologies of smartphone users have previously been developed, for example, smart-
phone addiction research (Bian & Leung, 2015; Elhai & Contractor, 2018; Kim et al.,
2016) and marketing research (Calvo-Porral & Otero-Prada, 2020; Chen et al., 2019;
De Canio et al., 2016; Hamka et al., 2014; Petrovc
ic
et al., 2018; Sell et al., 2014).
For example, in the area of smartphone addiction research, Elhai and Contractor
(2018) examined the relationship between usage patterns and problematic smartphone
use in a survey of 296 college students in the United States. To classify the students
into types of smartphone users, they conducted a latent class analysis based on the stu-
dentsself-reported frequency of engaging in 11 different activities on their smartphone,
for example, voice/video calls, texting/instant messaging, and using social media. The
authors identied two classes of smartphone users: heavy users,who use their phone
for a large variety of activities; and light users,who particularly make use of social
media and audio entertainment as well as taking photos and recording videos.
In the marketing literature, typologies of smartphone users have been proposed in the
context of customer segmentation research. For example, De Canio et al. (2016) collected
survey data from 264 smartphone users in Italy to conduct a cluster analysis based on their
self-reported use of 10 different smartphone functions. The authors identied ve types of
smartphone users which they called unfriendly users(who do not use any of the 10 smart-
phone functions), utility users(who mainly use their smartphone for voice calling and infor-
mation seeking), gamers(who mainly use their smartphone for games), moderator users
(who use all of the 10 smartphone functions, but information seeking in particular), and
supersmartphoners(who mainly use their smartphone for taking photos, recording videos,
social media, and staying in contact with others, for example, using video calling).
Using a different measurement approach, Hamka et al. (2014) conducted a latent class
analysis based on log les from 129 smartphone users who installed a tracking app on their
device for at least two weeks. The authors identied six types of smartphone users, includ-
ing application ignorant users(who visit a small number of URLs in the mobile browser
and use a small number of apps per day), basic application users(who visit a small
number of URLs but use a medium number of apps), average application users(who
4Mobile Media & Communication 0(0)
visit a medium number of URLs and use a medium number of apps), information seekers
(who visit a large number of URLs but use a small number of apps), app savvy users
(who visit a large number of URLs, use an extensive number of apps, and install a large
number of new apps), and high utility users(who visit an extensive number of URLs,
use an extensive number of apps, but install a small number of new apps).
A limitation of these existing typologies is that they reect the underlying research
questions, for example, smartphone addiction, rather than capturing the multidimension-
ality of smartphone use. In addition, they are often based on small samples or focus on
specic subgroups of the population such as adolescents or college students. The typ-
ology developed in this paper aims to expand upon the existing smartphone usage typolo-
gies by focusing on general smartphone use and drawing on survey data that are more
representative of the general population.
Correlates of smartphone use
Previous research on smartphone use has been relatively sparse compared with the large
body of research on correlates of general Internet and digital media use (Blank & Groselj,
2014; van Deursen & van Dijk, 2014). Frequency and variety of smartphone use have
been found to decrease with age (Andone et al., 2016; Fortunati & Taipale, 2014;
Serrano-Cinca et al., 2018). With regard to the types of activities carried out on smart-
phones, younger individuals are more likely to use their smartphone for entertainment
and games, social interaction, and commercial activities (Andone et al., 2016; Kongaut
& Bohlin, 2016). For other activities such as reading and writing emails on a smartphone,
no signicant age effects have been found (Kongaut & Bohlin, 2016).
Gender also seems to affect patterns of smartphone use. For example, previous
research has shown that women use smartphones for a greater amount of time and
visit a larger number of websites on their mobile browser than men (Andone et al.,
2016; Cotten et al., 2009; Roberts et al., 2014; Wang & Liu, 2018). In contrast,
men use a greater number of apps on their smartphone and generate a larger volume
of uploaded and downloaded data trafc via the cellular network (Wang & Liu,
2018; Zhao et al., 2020). However, with regard to variety of smartphone use, the evi-
dence for gender differences is mixed (Cotten et al., 2009; Fortunati & Taipale, 2014;
Mascheroni & Ólafsson, 2016; Serrano-Cinca et al., 2018; Wang & Liu, 2018; Zhao
et al., 2020). As far as types of use are concerned, women are more likely to use
their smartphone for social interaction and taking photos, whereas men are more
likely to read or watch news on their smartphone (Andone et al., 2016; Cotten
et al., 2009; Kim et al., 2016; Kongaut & Bohlin, 2016; Mascheroni & Ólafsson,
2016; Roberts et al., 2014; Zhao et al., 2020). Mixed ndings were reported for
other smartphone activities, including entertainment and games, commercial activities,
and reading and writing emails (Andone et al., 2016; Cotten et al., 2009; Kongaut &
Bohlin, 2016; Mascheroni & Ólafsson, 2016; Roberts et al., 2014; Wang & Liu, 2018;
Zhao et al., 2020).
Finally, education has been found to affect smartphone use. Individuals with higher
levels of educational attainment use their smartphone for a larger number of activities
(Fortunati & Taipale, 2014; Serrano-Cinca et al., 2018). Furthermore, they are more
Wenz and Keusch 5
likely to use their smartphone for entertainment and games, social interaction, commer-
cial activities, and reading and writing emails than those with lower levels of educational
attainment (Kongaut & Bohlin, 2016).
Whereas the large majority of research on correlates of digital media use has relied on
cross-sectional data that are collected at one point in time, several studies have also exam-
ined how usage behavior has changed over time (Kim et al., 2019; Li et al., 2020; van
Deursen et al., 2015). For example, Kim et al. (2019) collected log data from 139,935
iPhone users between 2012 and 2016 who installed a tracking app on their device, and
found that the use of social media apps and entertainment apps, including photo,
video, and gaming apps, increased during this period, whereas the use of productivity
apps decreased. Similarly, Li et al. (2020) used data collected between 2012 and 2017
from 1,465 Android smartphone users to study longitudinal changes in app use. They
found that diversity of app use increased over the ve-year period, whereas the
number of apps used increased between 2012 and 2014 but decreased between 2014
and 2017.
Data and methods
Samples
To answer our research questions, we use survey data from two samples of smartphone
owners collected in Germany in November 2017 and January 2020.
Sample 1. The rst sample comes from Wave 32 of the German Internet Panel
(GIP) conducted in November 2017 (Blom et al., 2018). The GIP is a probability-
based online panel of the German general population aged 1675 (Blom et al.,
2015). Sample members were recruited face-to-face in 2012 and 2014. Individuals
without computer and/or Internet access were provided with computing equipment
and broadband Internet (Blom et al., 2017). Every two months, panel members are
invited via email to complete web surveys on political and economic attitudes. A
total of 2,648 panel members answered the survey, resulting in a completion rate
of 53.3% and a cumulative response rate of 10.9%. Only smartphone owners
were asked the relevant questions for our analysis, resulting in an analysis sample
of N=2,186.
Sample 2. The second sample comes from a web survey conducted among members
of a German nonprobability online panel in January 2020 (Keusch, 2021). Panel
members were invited to participate through a survey-router system, and 3,350 indivi-
duals started the survey. Only smartphone owners were eligible for the study, and
quotas for gender, age, and frequency of smartphone use were employed. Of the panel
members who started the survey, 669 were screened out because of the quotas,
because they reported not owning a smartphone, or because they did not live in
Germany. Of the 2,681 remaining respondents, 156 broke off the survey (5.8%), resulting
in an analysis sample of N=2,525.
Descriptive statistics for the two samples are shown in Table 1. Both samples are com-
parable in terms of sociodemographics, except for age, with Sample 1 containing a larger
proportion of older adults than Sample 2.
6Mobile Media & Communication 0(0)
Latent class analysis
We conduct a latent class analysis (LCA) to create typologies of smartphone users.
LCA is a clustering method for identifying latent (unobserved) classes in a population
from a set of observed categorical indicators (McCutcheon, 1987). Individuals are
Table 1. Descriptive Statistics.
Sample 1 (2017) Sample 2 (2020)
% Missing % Missing
Gender 1 0
Female 49.2 49.6
Male 50.8 50.4
Age 1 0
1827 / 1829 10.8 27.3
2837 / 3039 18.1 23.4
3847 / 4049 17.6 23.4
4857 / 5059 26.0 19.1
58 +/60+27.5 6.7
Educational attainment 1 0
No high school degree 46.0 44.6
High school degree 22.6 26.2
College degree 31.4 29.1
Operating system 11
iOS 25.0
Android 70.8
Other operating system 4.2
Frequency of smartphone use 1 8
Several times a day 69.2 65.7
Every day 14.4 20.8
Several times a week or less 16.4 13.5
Smartphone skills 1 7
Advanced (5) 23.8 38.2
Intermediate (4) 28.6 34.0
Beginner (13) 47.6 27.7
Smartphone activities
Browse websites 76.7 0 93.6 10
Email 76.2 0 89.5 12
Photo 90.1 0 94.8 10
View content on social media 46.8 0 75.3 10
Post content to social media 26.2 0 61.8 14
Online purchase 33.1 0 67.3 15
Online banking 25.3 0 62.7 25
Install apps 42.3 0 84.5 16
GPS 58.5 0 83.8 19
Bluetooth 33.4 0 67.5 15
Games 29.4 0 64.8 14
Streaming 39.0 0 75.4 17
N2,186 2,525
Wenz and Keusch 7
assigned to the different classes based on their similarity in response patterns on the
indicator variables. LCA has been used in a number of previous studies to construct
typologies of Internet and digital media use (Elhai & Contractor, 2018; Hamka
et al., 2014; Herzing & Blom, 2019; Holmes, 2011; Sell et al., 2014; Yates et al.,
2020).
We use 14 variables in the latent class models that were collected in both samples,
including frequency of smartphone use, self-rated smartphone skills, and 12 smart-
phone activities (Table A1 in the Appendix). Frequency of smartphone use is mea-
suredonave-point rating scale collapsed to three categories (several times a
day, every day, several times a week or less). Self-rated smartphone skills are mea-
sured on a ve-point rating scale collapsed to three categories (advanced [5], inter-
mediate [4], beginner [13]). The categories were collapsed for these two variables
because very few respondents selected the lower end of the scales and, thus, the ori-
ginal response distributions were highly skewed (see Table B1 in the Appendix for
the response distribution of the original categories before recoding). Finally, 12
activities carried out on the smartphone are measured with a series of yes/no ques-
tions, reecting types of activities that were examined in previous research on
Internet and digital media use. These include social interaction (post content to
social media), reading and writing emails (email), entertainment (games, streaming),
commercial activities (online purchase, online banking), and information seeking
(browse websites, view content on social media). The list was complemented by
activities that are specically carried out on smartphones, including taking photos,
using GPS/location-aware apps, installing new apps, and using Bluetooth to
connect the smartphone to other devices. A small proportion of missing values on
these variables (< 2%) were imputed with a chained equations algorithm by using
the R mice package, version 3.13.0 (van Buuren & Groothuis-Oudshoorn, 2011).
Descriptive statistics for the variables included in the LCA are shown in Table 1.
To estimate the latent class models, we use the R poLCA package, version 1.4.1
(Linzer & Lewis, 2011). We vary the number of classes in the LCA from two to 10
and compute model t criteria, including the log likelihood (LL), the Akaike infor-
mation criterion (AIC), and the Bayesian information criterion (BIC), to select the
best-tting model, with lower values indicating a better model t (Nylund et al.,
2007). We also report the size and percentage of the smallest class.
Sociodemographic and smartphone-related correlates
As correlates of smartphone usage types, we collected data on sociodemographic char-
acteristics in both samples (see Table 1 for descriptive statistics). The sociodemo-
graphic characteristics include gender (male vs. female), age (Sample 1: 1827
years, 2837 years, 3847 years, 4857 years, 58 +years; Sample 2: 1829 years,
3039 years, 4049 years, 5059 years, 60 +years),
1
and educational attainment
(no high school degree, high school degree, college degree). In Sample 2, we also col-
lected data on the operating system of the smartphone (iOS, Android, other operating
system). The data preparation and analysis were conducted in R, version 4.0.4 (R Core
Team, 2021).
8Mobile Media & Communication 0(0)
Results
Which smartphone usage types can be identied among smartphone owners?
We rst investigated which types of smartphone users can be identied in the two
samples by conducting a LCA. Varying the number of classes from two to 10 showed
that the BIC reached a minimum at the six-class model in both Sample 1 and Sample
2, with a LL and AIC that did not decrease substantially as more classes were included
in the model (Table 2). The six-class model also resulted in classes with a reasonable size,
with the smallest class containing 166 individuals in Sample 1 (8% of the overall sample)
and 156 individuals in Sample 2 (6% of the overall sample). Therefore, we selected the
six-class solution for our analysis.
Next, we examined the composition of the latent classes. Table 3 shows the predictor
variables (frequency of use, skills, types of activities) by smartphone usage class for
Sample 1, and Table C1 in the Appendix shows the results for Sample 2. We describe
the six usage types as follows.
Advanced users used their smartphone several times a day (Sample 1: 98%; Sample 2:
82%) and mostly rated their smartphone skills as advanced (Sample 1: 72%; Sample 2:
59%). The majority of advanced users engaged in each of the 12 activities on their smart-
phone (Sample 1: each activity used by at least 60%; Sample 2: at least 87%), but there
were a few differences across samples. Advanced users in Sample 1, compared with those
in Sample 2, were less likely to post content on social media (76% vs. 93%), use
Bluetooth (77% vs. 94%), engage with online banking (70% vs. 94%), or play games
on their smartphone (60% vs. 87%). However, they were still much more likely to
engage in these activities than the other types of smartphone users in Sample
1. Advanced users constituted the largest usage group in Sample 2, with almost half of
the sample (44%) categorized into this group, but made up a considerably smaller propor-
tion of Sample 1 (17%).
Broad non-social media users used their smartphone several times a day (Sample 1:
96%; Sample 2: 78%) and mostly rated their smartphone skills as intermediate
(Sample 1: 43%; Sample 2: 40%) or advanced (Sample 1: 39%; Sample 2: 35%).
They used their smartphone for a large variety of activities (Sample 1: each activity
used by at least 40%; Sample 2: at least 58%), except for social media, with only 7%
posting content and 24% viewing content on social media in Sample 2 and none of
them viewing or posting content on social media in Sample 1. These users constituted
8% of Sample 1 and 9% of Sample 2.
Broad non-commercial users mostly used their smartphone several times a day
(Sample 1: 94%; Sample 2: 67%) and rated their smartphone skills as intermediate
(Sample 1: 52%; Sample 2: 40%), advanced (Sample 1: 22%; Sample 2: 30%), or
beginner (Sample 1: 27%; Sample 2: 30%). They used their smartphone for a large
variety of activities except for commercial activities, with only 25% using their smart-
phone for online banking in Sample 1 (37% in Sample 2), and 38% for online pur-
chases in Sample 1 (45% in Sample 2). Similar to advanced users, this group
showed some differences in usage patterns across samples. Broad non-commercial
users in Sample 1, compared with Sample 2, were considerably less likely to use
Wenz and Keusch 9
Table 2. Model Fit and Diagnostic Criteria for Two to Ten Classes of Smartphone Use.
Sample 1 (2017) Sample 2 (2020)
Model t criteria Diagnostic criteria Model t criteria Diagnostic criteria
#
Classes LL AIC BIC
Smallest
class (n)
Smallest
class (%) LL AIC BIC
Smallest
class (n)
Smallest
class (%)
216,942.82 33,951.64 34,139.40 970 44 17,194.68 34,455.36 34,647.88 813 32
316,490.92 33,081.84 33,366.33 565 26 16,716.83 33,533.66 33,825.36 289 11
416,378.65 32,891.31 33,272.52 275 13 16,515.94 33,165.88 33,556.76 306 12
516,244.41 32,656.82 33,134.77 272 12 16,406.31 32,980.62 33,470.67 134 5
616,166.67 32,535.35 33,110.02 166 8 -16,317.85 32,837.70 33,426.93 156 6
716,153.10 32,542.21 33,213.61 133 6 16,277.62 32,791.24 33,479.65 80 3
816,103.31 32,476.62 33,244.75 124 6 16,249.80 32,769.60 33,557.19 82 3
916,065.45 32,434.90 33,299.75 56 3 16,228.25 32,760.50 33,647.26 32 1
10 16,043.34 32,424.68 33,386.27 70 3 16,202.97 32,743.94 33,729.89 26 1
N2,186 2,525
Note.LL=log likelihood; AIC =Akaike information criterion; BIC =Bayesian information criterion. The row in bold indicates the latent class model chosen for the analysis.
10 Mobile Media & Communication 0(0)
Table 3. Predictor Variables by Class of Smartphone Use (Sample 1).
Variables
Advanced
users %
Broad non-social
media users %
Broad non-commercial
users %
Social media and
information users %
Basic general
users %
Camera
users %
Class size 17 8 19 12 25 20
Frequency of
smartphone use
Several times a day 98 96 94 60 69 18
Every day 2 4 6 24 19 25
Several times a week
or less
00 0 15 1357
Smartphone skills
Advanced (5) 72 39 22 11 12 4
Intermediate (4) 23 43 52 19 32 6
Beginner (13) 5 19 27 71 56 90
Smartphone activities
Browse websites 100 99 96 61 84 30
Email 98 97 91 62 76 45
Photo 100 96 99 84 92 72
View content on
social media
100 0 99 93 0 2
Post content to
social media
76 0 46 39 0 0
Online purchase 93 75 38 9 14 1
Online banking 70 54 25 6 15 2
Install apps 97 80 62 4 30 1
GPS 97 92 79 30 61 8
Bluetooth 77 69 43 6 24 3
Games 60 40 33 21 22 11
Streaming 93 77 48 15 26 2
N363 166 419 254 549 435
Wenz and Keusch 11
their smartphone for streaming videos or music (48% vs. 78%), playing games (33%
vs. 61%), installing apps (62% vs. 90%), posting content on social media (46% vs.
71%), or using Bluetooth (43% vs. 59%). Broad non-commercial users constituted
the second largest usage group in Sample 2, with almost one fourth of the sample
(23%) categorized into this group, and they made up a similar proportion of Sample
1(19%).
Social media and information users used their smartphone several times a day
(Sample 1: 60%; Sample 2: 45%) and a majority rated their smartphone skills as begin-
ner (Sample 1: 71%; Sample 2: 62%). They mainly used their smartphone to browse
websites (Sample 1: 61%; Sample 2: 76%), read and/or write emails (Sample 1:
62%; Sample 2: 65%), take photos (Sample 1: 84%; Sample 2: 64%), or use social
media, with 93% viewing content on social media in Sample 1 (74% in Sample 2).
In Sample 2, this usage group also commonly posted content on social media (72%)
although those in Sample 1 were much less likely to do so (39%). These users consti-
tuted 12% of Sample 1 but made up a smaller proportion of Sample 2 (6%).
Basic general users mostly used their smartphone several times a day (69%) in Sample
1, whereas frequency of use was more evenly distributed in Sample 2 (several times a
day: 31%; every day: 35%; several times a week or less: 35%). In both samples, they
mainly rated their smartphone skills as beginner (Sample 1: 56%; Sample 2: 47%) or
intermediate (Sample 1: 32%; Sample 2: 39%). Almost all of them used their smartphone
to take photos (Sample 1: 92%; Sample 2: 91%) and browse websites (Sample 1: 84%;
Sample 2: 90%). Other popular activities included using GPS/location-aware apps
(Sample 1: 61%; Sample 2: 83%) and reading and/or writing emails (Sample 1: 76%;
Sample 2: 79%). Basic general users in Sample 2 also commonly installed new apps
on their smartphone (80%), although those in Sample 1 were considerably less likely
to do so (30%). These users constituted 25% of Sample 1, but made up a smaller propor-
tion of Sample 2 (11%).
Camera users mostly used their smartphone several times a week or less (Sample 1:
57%; Sample 2: 62%), with only 25% using their smartphone every day in Sample 1
(23% in Sample 2) and 18% several times a day in Sample 1 (16% in Sample 2), and
they rated their smartphone skills as beginner (Sample 1: 90%; Sample 2: 82%). They
mainly used their smartphone to take photos (Sample 1: 72%; Sample 2: 77%). Other
popular activities included browsing websites (Sample 1: 30%; Sample 2: 48%) and
reading and/or writing emails (Sample 1: 45%; Sample 2: 45%). Only a small share
of users (13% in both samples) were engaged with the other nine activities. They
constituted 20% of Sample 1 but made up a considerably smaller proportion of
Sample 2 (7%).
In summary, we identied six types of smartphone users with distinct patterns of
smartphone usage frequency, self-rated smartphone skills, and activities carried out on
their smartphone. Three of these types (advanced users,broad non-social media users,
and broad non-commercial users) used their smartphone frequently, rated their skills
as advanced or intermediate, and used their device for a large range of activities. The
other three types (social media and information users,basic general users, and
camera users) used their smartphone less frequently, rated their skills as beginner or
intermediate, and used their device for a rather narrow set of activities.
12 Mobile Media & Communication 0(0)
How do the smartphone usage types differ in sociodemographic characteristics
and in the operating system used?
Next, we examined how the smartphone usage types differ in sociodemographic charac-
teristics and in the operating system. Table 4 shows the characteristics by smartphone
usage class for Sample 1, and Table C2 in the Appendix shows the results for Sample
2. To test whether differences are statistically signicant, we conducted Pearsons
chi-squared tests and reported the p-values. We also carried out post-hoc analyses
based on standardized residuals (Beasley & Schumacker, 1995) to detect which smart-
phone usage types most contributed to overall signicant results, using the R chisq.-
posthoc.test package, version 0.1.2 (Ebbert, 2019).
We found signicant gender differences between the smartphone usage types (Sample
1: p< 0.001; Sample 2: p=0.006). Specically, basic general users were disproportion-
ately male (Sample 1: 57%, p=0.007; Sample 2: 59%, p=0.043). Similarly, broad non-
social media users were disproportionately male (Sample 1: 66%; Sample 2: 58%),
although the difference was only signicant for Sample 1 (p< 0.001). In contrast,
camera users were disproportionately female (Sample 1: 58%; Sample 2: 53%) but,
again, the difference was only signicant for Sample 1 (p=0.001). For the other four
usage types, the gender distribution did not differ signicantly from the entire sample.
The smartphone usage classes also had a signicantly different age composition
(p< 0.001 in both samples). The usage types reecting more diverse and more frequent
smartphone use were disproportionately younger. Advanced users were signicantly
more likely to belong to the youngest age group, with 29% of users aged 1827 in
Sample 1 (p< 0.001) and 38% aged 1829 in Sample 2 (p< 0.001). They were also sig-
nicantly less likely to belong to the oldest age group, with only 4% of users aged 58 and
above in Sample 1 (p< 0.001) and 2% aged 60 and above in Sample 2 (p< 0.001). For
broad non-social media users and broad non-commercial users, the age distribution
was somewhat older, but they were still less likely to be part of the oldest age group,
although the difference was only signicant for Sample 1 (broad non-social media
users: 17%, p=0.042; broad non-commercial users: 14%, p< 0.001). In turn, the
usage types reecting more narrow and less frequent smartphone use were disproportion-
ately older. Camera users were signicantly more likely to be in the oldest age group
(Sample 1: 53%, p< 0.001; Sample 2: 20%, p< 0.001) as were basic general users
(Sample 1: 34%, p< 0.001; Sample 2: 16%, p< 0.001). Finally, the age composition of
social media and information users did not signicantly differ from the entire sample
for most age groups, although they were signicantly less likely to be aged 2837 in
Sample 1 (8%; p=0.001).
The smartphone usage types also differed signicantly by educational attainment
(p< 0.001 in both samples). Advanced users were disproportionately more highly edu-
cated. They were signicantly more likely to have a high school degree (Sample 1:
33%, p< 0.001; Sample 2: 30%, p=0.004) and signicantly less likely to be without a
high school degree (Sample 1: 31%, p< 0.001; Sample 2: 38%, p< 0.001). Similarly,
broad non-social media users in Sample 1 were signicantly more likely to have a
college degree (52%; p< 0.001) and less likely to have no high school degree (26%;
p< 0.001), although the differences in educational attainment were not signicant in
Wenz and Keusch 13
Table 4. Sociodemographic Characteristics by Class of Smartphone Use (Sample 1).
Variables
Advanced
users %
Broad non-social
media users %
Broad
non-commercial
users %
Social media and
information users %
Basic
general
users %
Camera
users %
Chi-squared test
p-Value
Gender <0.001
Female 44 34∗∗∗ 54 57 43∗∗ 58∗∗
Male 56 66∗∗∗ 46 43 57∗∗ 43∗∗
Age <0.001
1827 29∗∗∗ 717
∗∗∗ 84
∗∗∗ 1∗∗∗
2837 32∗∗∗ 36∗∗∗ 23 8∗∗ 15 5∗∗∗
3847 21 19 21 17 18 10∗∗∗
4857 14∗∗∗ 22 26 33 28 31
58 +4∗∗∗ 1714∗∗∗ 33 34∗∗∗ 53∗∗∗
Educational
attainment
<0.001
No high
school degree
31∗∗∗ 26∗∗∗ 41 58∗∗ 44 67∗∗∗
High school
degree
33∗∗∗ 22 27 21 21 13∗∗∗
College
degree
36 52∗∗∗ 32 22∗∗ 34 21∗∗∗
Note. The asterisks indicate p-values from standardized residuals: p<0.05. ∗∗p< 0.01. ∗∗∗p< 0.001.
14 Mobile Media & Communication 0(0)
Sample 2 for this type of smartphone user. In contrast, camera users had a disproportion-
ately lower level of educational attainment. They were signicantly more likely to be
without a high school degree (Sample 1: 67%, p< 0.001; Sample 2: 57%, p=0.013)
and in Sample 1 signicantly less likely to have a high school degree (13%; p< 0.001)
or a college degree (21%; p< 0.001). Similarly, the proportion of people without a
high school degree was disproportionately higher for broad non-commercial users in
Sample 2 (51%; p=0.020) and for social media and information users in Sample 1
(58%; p=0.002). Finally, basic general users did not differ signicantly from the
entire sample in their level of educational attainment.
In Sample 2, we also found signicant differences between the smartphone usage
types by operating system (p< 0.001). Advanced users were disproportionately iPhone
owners (33%; p< 0.001). In turn, most of the other usage types were disproportionately
Android smartphone owners, including camera users (81%; p=0.049), basic general
users (79%; p=0.030), and broad non-commercial users (76%; p=0.014). Finally,
the ratio of iPhone and Android smartphone owners among broad non-social media
users and social media and information users did not differ signicantly from the
entire sample. The latter group, however, were disproportionately more likely to use
an operating system other than iOS or Android on their smartphone (10%; p=0.009).
In summary, we found that the smartphone usage types differed signicantly by socio-
demographic characteristics and operating system. The usage types reecting more fre-
quent and diverse smartphone use were generally younger, had higher levels of
educational attainment, and were more likely to use an iPhone rather than an Android
smartphone compared with the usage types that used their smartphone less frequently
and for a narrower range of activities.
Discussion
Over the last decade, smartphones have become relevant for an increasing number of
activities in peoples daily lives, from communication and information seeking to
health- and work-related activities, some of which rely exclusively on smartphone tech-
nologies. Although a growing number of people in the general population have access to
these devices, it is not yet understood to what extent they have the necessary usage
experience and skills to take full advantage of their smartphone. In this paper, we contrib-
ute to the small body of research on the second-level smartphone divide by constructing a
typology of smartphone users in the general population based on two nationally represen-
tative samples of smartphone owners collected in Germany in 2017 and 2020. We clas-
sify individuals based on their frequency of smartphone use, self-rated smartphone skills,
and activities carried out on their smartphone, and identify six distinct types of smart-
phone users in both samples.
On the one hand, we identify types of smartphone owners who use their device fre-
quently and mostly rate their smartphone skills as advanced or intermediate but differ
in the types of activities they carry out on their smartphone: Advanced users use their
smartphone for the full set of activities that we examined in the study, whereas broad
non-social media users use their device for all activities but viewing and posting
content on social media. Similarly, broad non-commercial users use their smartphone
Wenz and Keusch 15
for a large variety of activities with the exception of online banking and making online
purchases. Some of these groups reect usage types that have also been identied in pre-
vious typologies of general digital media use, including advanced users (e.g., De Canio
et al., 2016: supersmartphoners; Reisdorf & Groselj, 2017: broad users; Yates et al.,
2020: extensive users) and broad non-social media users (e.g., Yates et al., 2020:
general (no social media) users). Smartphone owners who belong to one of these
usage types tend to be younger and have higher levels of educational attainment,
which is in line with previous research in this area (Andone et al., 2016; Fortunati &
Taipale, 2014; Serrano-Cinca et al., 2018). Advanced users also have the highest propor-
tion of iPhone owners, which is consistent with recent research showing that iPhone users
tend to be younger and have higher levels of educational attainment (Keusch et al., 2020).
On the other hand, there are other types of smartphone owners who use their device
less frequently, mostly rate their smartphone skills as beginner or intermediate, and
use their device for a rather specic set of activities: Camera users mainly use their smart-
phone to take photos, whereas social media and information users use their device to
browse websites, read and/or write emails, take photos, and use social media. Finally,
basic general users use their smartphone to take photos, browse websites, use GPS/
location-aware apps, and read and/or write emails. Social media and information users
have also been identied in previous typologies of general digital media use (e.g.,
Yates et al., 2020: social and entertainment media only users) as have basic general
users (e.g., Yates et al., 2020: utility users). Smartphone owners who belong to these
usage types tend to be older and have lower levels of educational attainment, consistent
with previous research (Andone et al., 2016; Fortunati & Taipale, 2014; Serrano-Cinca
et al., 2018).
Replicating the analysis in samples collected in November 2017 and January 2020
shows that the smartphone usage types identied and the correlates of smartphone use
are robust, but there are a few notable differences between the samples. Overall, smart-
phone users in Sample 2 (2020) were more likely to engage in certain types of activities
than users in Sample 1 (2017), for example, using Bluetooth to connect their smartphone
to other electronic devices or engaging with online banking. This pattern points to the
increasing popularity of such smartphone activities over the time frame. In addition,
the size of the usage type groups identied is notably different across samples, probably
due to the different age composition of the two samples. Smartphone usage types that typ-
ically consist of younger users, for example, advanced users, make up a larger proportion
of Sample 2, which has a younger age distribution overall, than of Sample 1. The opposite
is the case for usage types mostly consisting of older users, for example, basic general
users or camera users that make up a larger proportion of Sample 1 than of Sample 2.
A key takeaway message from our study is that although smartphones have become
deeply integrated into most peoples daily lives, smartphone usage patterns continue to
be highly diverse, and a sizable proportion of smartphone owners still use their device
for a rather narrow set of activities. The observed variation in smartphone use highlights
the importance of a more nuanced perspective on smartphone users that future research
on the digital divide should take into account. As smartphones are increasingly required
for digital activities, the differential usage patterns of smartphone owners also have impli-
cations for the potential benets they can obtain from using their device, referred to as the
16 Mobile Media & Communication 0(0)
third-level digital divide (van Deursen & Helsper, 2015). Even among the types reecting
more frequent and diverse use in our study, individuals varied in their experience with
mobile banking, mobile payments, and mobile communication, and are potentially
excluded from these activities in their everyday lives. A second key takeaway message
from our study is that inequalities in smartphone use reect existing social inequalities
with regard to age and education, which replicates previous research on general
Internet and digital media use. Smartphone owners using their device less frequently
and for a narrower set of activities are disproportionately older and have lower levels
of educational attainment than more advanced smartphone users.
Our study is not free of limitations. First, we relied on self-reported measures of smart-
phone use and smartphone-related skills that are susceptible to measurement error.
Respondents might not be able to remember the frequency, duration, or variety of their
smartphone use accurately, and the resulting self-reports are likely to be biased (Boase
& Ling, 2013). Social desirability might also lead respondents to overrate their level of
skills (Palczynska & Rynko, 2021). Future research might consider replicating our ana-
lysis with passively collected or performance-based measures of smartphone use and
skills (Festic et al., 2021; Hamka et al., 2014; Hargittai, 2002). Second, although we
aimed to cover the most popular activities carried out on smartphones, our list of activities
might not have been exhaustive. In addition, our measure of smartphone-related skills
consists of one survey item and might not have fully captured the multidimensionality
of digital skills. We would welcome future studies that implement more detailed versions
of these measures. Third, although we investigated the robustness of our typology across
multiple samples, we were only able to examine differences over a relatively short period
of approximately two years. A potential avenue for future research would be to study
changes in the second-level smartphone divide over a longer period of time.
Data availability statement
The data used in this study are publicly available at the GESIS Data Archive for the Social Sciences
(Sample 1: https://doi.org/10.4232/1.13043; Sample 2: https://doi.org/10.7802/2331). The analysis
code is available at https://doi.org/10.17605/OSF.IO/X748K.
Declaration of conicting interests
The author(s) declared no potential conicts of interest with respect to the research, authorship, and/
or publication of this article.
Funding
The author(s) disclosed receipt of the following nancial support for the research, authorship, and/
or publication of this article: This work was supported by the German Research Foundation
(DFG) through the Collaborative Research Center SFB 884 Political Economy of Reforms
(Project A8) [139943784 to Annelies Blom, Florian Keusch, and Frauke Kreuter].
ORCID iD
Alexander Wenz https://orcid.org/0000-0002-4621-2418
Wenz and Keusch 17
Note
1. We included age as an ordinal variable because the released dataset for Sample 1 only provides
age categories. We replicated the analysis in Sample 2 by including age as a continuous rather
than ordinal variable, but the conclusions remained unchanged.
References
Andone, I., Blaszkiewicz, K., Eibes, M., Trendalov, B., Montag, C., & Markowetz, A. (2016).
How age and gender affect smartphone usage. In P. Lukowicz & A. Krüger (Eds.),
UbiComp 16: Proceedings of the 2016 ACM International Joint Conference on Pervasive
and Ubiquitous Computing: Adjunct,1216 September 2016, Heidelberg, Germany
(pp. 912). ACM.
Kim, D., Gluck, J., Hall, M., & Agarwal, Y. (2019). Real world longitudinal iOS app usage study at
scale. ArXiv. https://arxiv.org/abs/1912.12526
Beasley, T. M., & Schumacker, R. E. (1995). Multiple regression approach to analyzing contin-
gency tables: Post hoc and planned comparison procedures. The Journal of Experimental
Education,64(1), 7993. https://doi.org/10.1080/00220973.1995.9943797
Bian, M., & Leung, L. (2015). Linking loneliness, shyness, smartphone addiction symptoms, and
patterns of smartphone use to social capital. Social Science Computer Review,33(1), 6179.
https://doi.org/10.1177/0894439314528779
Blank, G., & Groselj, D. (2014). Dimensions of Internet use: Amount, variety, and types.
Information, Communication & Society,17(4), 417435. https://doi.org/10.1080/1369118X.
2014.889189
Blom, A. G., Felderer, B., Höhne, J. K., Krieger, U., & Rettig, T., & SFB 884 Political Economy
of ReformsUniversität Mannheim (2018). German Internet Panel, Wave 32 (November
2017) (ZA6906 Data le Version 1.0.0) [Data set]. GESIS Data Archive, Cologne.
Blom, A. G., Gathmann, C., & Krieger, U. (2015). Setting up an online panel representative of the
general population: The German Internet Panel. Field Methods,27(4), 391408. https://doi.
org/10.1177/1525822X15574494
Blom, A. G., Herzing, J. M. E., Cornesse, C., Sakshaug, J. W., Krieger, U., & Bossert, D. (2017).
Does the recruitment of ofine households increase the sample representativeness of
probability-based online panels? Evidence from the German Internet Panel. Social Science
Computer Review,35(4), 498520. https://doi.org/10.1177/0894439316651584
Boase, J., & Ling, R. (2013). Measuring mobile phone use: Self-report versus log data. Journal
of Computer-Mediated Communication,18(4), 508519. https://doi.org/10.1111/jcc4.
12021
Calvo-Porral, C., & Otero-Prada, L.-M. (2020). A prole of mobile service users in a mature
market: From uninvolved pragmaticsto potential switchers.Spanish Journal of Marketing,
25(3), 425445. http://dx.doi.org/10.1108/SJME-03-2020-0046
Chen, H., Zhang, L., Chu, X., & Yan, B. (2019). Smartphone customer segmentation based on the
usage pattern. Advanced Engineering Informatics,42(October 2019), 113. https://doi.org/10.
1016/j.aei.2019.101000
Cotten, S. R., Anderson, W. A., & Tufekci, Z. (2009). Old wine in a new technology, or a different
type of digital divide? New Media and Society,11(7), 11631186. https://doi.org/10.1177/
1461444809342056
De Canio, F., Pellegrini, D., & Aramendia-Muneta, M. E. (2016). The smartphoners: Consumer
segmentation by smartphone usage. Mercati E Competitività,1, 123144. https://doi.org/10.
3280/MC2016-001007
18 Mobile Media & Communication 0(0)
DiMaggio, P., Hargittai, E., Celeste, C., & Shafer, S. (2004). Digital inequality: From unequal
access to differentiated use. In K. M. Neckerman (Ed.), Social inequality (pp. 355400).
Russell Sage Foundation.
Ebbert, D. (2019). chisq.posthoc.test: A post hoc analysis for Pearsons chi-squared test for count
data. R package, version 0.1.2. https://cran.r-project.org/package=chisq.posthoc.test
Elhai, J. D., & Contractor, A. A. (2018). Examining latent classes of smartphone users: Relations
with psychopathology and problematic smartphone use. Computers in Human Behavior,
82(May 2018), 159166. https://doi.org/10.1016/j.chb.2018.01.010
Eurostat (2021). Digital economy and society: ICT usage in households and by individuals. https://
ec.europa.eu/eurostat/web/digital-economy-and-society/data/database
Festic, N., Büchi, M., & Latzer, M. (2021). How long and what for? Tracking a nationally repre-
sentative sample to quantify Internet use. Journal of Quantitative Description: Digital Media,
1(April),123. https://doi.org/10.51685/jqd.2021.018
Fortunati, L., & Taipale, S. (2014). The advanced use of mobile phones in ve European countries.
British Journal of Sociology,65(2), 317337. https://doi.org/10.1111/1468-4446.12075
Hamka, F., Bouwman, H., De Reuver, M., & Kroesen, M. (2014). Mobile customer segmentation
based on smartphone measurement. Telematics and Informatics,31(2), 220227. https://doi.
org/10.1016/j.tele.2013.08.006
Hargittai, E. (2002). Second-level digital divide: Differences in peoples online skills. First
Monday,7(4). https://doi.org/10.5210/fm.v7i4.942
Helsper, E. J. (2021). The digital disconnect: The social causes and consequences of digital
inequalities. SAGE Publications.
Herzing, J. M. E., & Blom, A. G. (2019). The inuence of a persons digital afnity on unit non-
response and attrition in an online panel. Social Science Computer Review,37(3), 404424.
https://doi.org/10.1177/0894439318774758
Holmes, J. (2011). Cyberkids or divided generations? Characterising young peoples internet use in
the UK with generic, continuum or typological models. New Media and Society,13(7), 1104
1122. https://doi.org/10.1177/1461444810397649
Horrigan, J. B. (2007). A typology of information and communication technology users. Pew Research
Center. https://www.pewresearch.org/internet/2007/05/06/a-typology-of-information-and-
communication-technology-users/
Keusch, F. (2021). Smartphone usage behavior (Data le Version 1.0.0) [Data set]. GESIS Data
Archive, Cologne.
Keusch, F., Bähr, S., Haas, G.-C., Kreuter, F., & Trappmann, M. (2020). Coverage error in data
collection combining mobile surveys with passive measurement using apps: Data from a
German National Survey. Sociological Methods & Research,138. https://doi.org/10.1177/
0049124120914924
Kim, D., Nam, J. E. K., Oh, J. S., & Kang, M. C. (2016). A latent prole analysis of the interplay
between PC and smartphone in problematic internet use. Computers in Human Behavior,
56(March 2016), 360368. https://doi.org/10.1016/j.chb.2015.11.009
Kongaut, C., & Bohlin, E. (2016). Investigating mobile broadband adoption and usage: A case of
smartphones in Sweden. Telematics and Informatics,33(3), 742752. https://doi.org/10.1016/
j.tele.2015.12.002
Li, T., Zhang, M., Cao, H., Li, Y., Tarkoma, S., & Hui, P. (2020). What apps did you use?:
Understanding the long-term evolution of mobile app usage. In Y. Huang, I. King, T-Y.
Liu, & M. van Steen (Eds.), Proceedings of the World Wide Web Conference 2020 (WWW
20), April 2024 2020, Taipei, Taiwan (pp. 6676). ACM.
Linzer, D. A., & Lewis, J. B. (2011). poLCA: An R package for polytomous variable latent class
analysis. Journal of Statistical Software,42(10), 129. https://doi.org/10.18637/jss.v042.i10
Wenz and Keusch 19
Livingstone, S., & Helsper, E. (2007). Gradations in digital inclusion: Children, young people and the
digital divide. New Media and Society,9(4), 671696. https://doi.org/10.1177/146144480
7080335
Mascheroni, G., & Ólafsson, K. (2016). The mobile Internet: Access, use, opportunities and divides
among European children. New Media and Society,18(8), 16571679. https://doi.org/10.
1177/1461444814567986
McCutcheon, A. L. (1987). Latent class analysis. SAGE Publications.
Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on the number of classes in
latent class analysis and growth mixture modeling: A Monte Carlo simulation study.
Structural Equation Modeling: A Multidisciplinary Journal,14(4), 535569. https://doi.org/
10.1080/10705510701575396
Palczynska, M., & Rynko, M. (2021). ICT Skills measurement in social surveys: Can we trust self-
reports? Quality and Quantity,55(3), 917943. https://doi.org/10.1007/s11135-020-01031-4
Pearce, K. E., & Rice, R. E. (2013). Digital divides from access to activities: Comparing mobile and
personal computer Internet users. Journal of Communication,63(4), 721744. https://doi.org/
10.1111/jcom.12045
Perrin, A. (2017). 10 facts about smartphones as the iPhone turns 10. Pew Research Center. https://
www.pewresearch.org/fact-tank/2017/06/28/10-facts-about-smartphones/
Petrovc
ic
, A., Slavec, A., & Dolnic
ar, V. (2018). The ten shades of silver: Segmentation of older
adults in the mobile phone market. International Journal of Human-Computer Interaction,
34(9), 845860. https://doi.org/10.1080/10447318.2017.1399328
Pew Research Center (2021). Mobile fact sheet. Pew Research Center. https://www.pewresearch.
org/internet/fact-sheet/mobile/
R Core Team (2021). R: A language and environment for statistical computing. R Foundation for
Statistical Computing. https://www.r-project.org/
Reisdorf, B. C., & Groselj, D. (2017). Internet (non-)use types and motivational access:
Implications for digital inequalities research. New Media and Society,19(8), 11571176.
https://doi.org/10.1177/1461444815621539
Roberts, J. A., Yaya, L. H. P., & Manolis, C. (2014). The invisible addiction: Cell-phone activities
and addiction among male and female college students. Journal of Behavioral Addictions,
3(4), 254265. https://doi.org/10.1556/JBA.3.2014.015
Robinson, L., Cotten, S. R., Ono, H., Quan-Haase, A., Mesch, G., Chen, W., Schulz, J., Hale, T. M., &
Stern, M. J. (2015). Digital inequalities and why they matter. Information, Communication and
Society,18(5), 569582. https://doi.org/10.1080/1369118X.2015.1012532
Sell, A., Mezei, J., & Walden, P. (2014). An attitude-based latent class segmentation analysis of
mobile phone users. Telematics and Informatics,31(2), 209219. https://doi.org/10.1016/j.
tele.2013.08.004
Selwyn, N., Gorard, S., & Furlong, J. (2005). Whose Internet is it anyway? Exploring adults(non)
use of the Internet in everyday life. European Journal of Communication,20(1), 526. https://
doi.org/10.1177/0267323105049631
Serrano-Cinca, C., Muñoz-Soro, J. F., & Brusca, I. (2018). A multivariate study of Internet use and
the digital divide. Social Science Quarterly,99(4), 14091425. https://doi.org/10.1111/ssqu.
12504
Silver, L. (2019). Smartphone ownership is growing rapidly around the world, but not always
equally. Pew Research Center. https://www.pewresearch.org/global/2019/02/05/smartphone-
ownership-is-growing-rapidly-around-the-world-but-not-always-equally/
Tsetsi, E., & Rains, S. A. (2017). Smartphone Internet access and use: Extending the digital divide
and usage gap. Mobile Media and Communication,5(3), 239255. https://doi.org/10.1177/
2050157917708329
20 Mobile Media & Communication 0(0)
van Buuren, S., & Groothuis-Oudshoorn, K. (2011). MICE: Multivariate imputations by chained
equations in R. Journal of Statistical Software,45(3), 167. https://doi.org/10.18637/jss.v045.i03
van Deursen, A. J. A. M., & Helsper, E. J. (2015). The third-level digital divide: Who benets most
from being online? In L. Robinson, S. R. Cotton, J. Schulz, T. M. Hale, & A. Williams (Eds.),
Communication and Information Technologies Annual (10th ed., pp. 2952). Emerald Group
Publishing. https://doi.org/10.1108/S2050-206020150000010002
van Deursen, A. J. A. M., & van Dijk, J. A. G. M. (2014). The digital divide shifts to differences in
usage. New Media and Society,16(3), 507526. https://doi.org/10.1177/1461444813487959
van Deursen, A. J. A. M., & van Dijk, J. A. G. M. (2019). The rst-level digital divide shifts from
inequalities in physical access to inequalities in material access. New Media and Society,
21(2), 354375. https://doi.org/10.1177/1461444818797082
van Deursen, A. J. A. M., van Dijk, J. A. G. M., & ten Klooster, P. M. (2015). Increasing inequal-
ities in what we do online: A longitudinal cross sectional analysis of Internet activities among
the Dutch population (2010 to 2013) over gender, age, education, and income. Telematics and
Informatics,32(2), 259272. https://doi.org/10.1016/j.tele.2014.09.003
van Dijk, J. (2020). The digital divide. Polity Press.
Wang, L., & Liu, C. (2018). The mobile Internet underclass: Reality or hyperbole? Journal of
Information Science,44(5), 569579. https://doi.org/10.1177/0165551517690083
Yates, S. J., Carmi, E., Lockley, E., Pawluczuk, A., French, T., & Vincent, S. (2020). Who are the
limited users of digital systems and media? An examination of U.K. evidence. First Monday,
25(7). https://doi.org/10.5210/fm.v25i7.10847
Zhao, S., Xu, Y., Ma, X., Jiang, Z., Luo, Z., Li, S., Yang, L. T., Dey, A., & Pan, G. (2020). Gender
proling from a single snapshot of apps installed on a smartphone: An empirical study. IEEE
Transactions on Industrial Informatics,16(2), 13301342. https://doi.org/10.1109/TII.2019.
2938248
Zillien, N., & Hargittai, E. (2009). Digital distinction: Status-specic types of internet usage. Social
Science Quarterly,90(2), 274291. https://doi.org/10.1111/j.1540-6237.2009.00617.x
Author biographies
Alexander Wenz is a postdoctoral fellow at the Mannheim Centre for European Social Research at
the University of Mannheim. His research examines the quality of novel methods of data collection,
with a focus on mobile web surveys and passive measurement with smartphone apps and wearable
sensors.
Florian Keusch is Professor of Social Data Science and Methodology at the School of Social
Sciences at the University of Mannheim. In his research he develops, implements, and assesses
the quality of novel methods of collecting data for the behavioral and social sciences.
Wenz and Keusch 21
Appendix A. Questionnaire
Table A1. Variables Used in the Latent Class Analysis.
Variable Question text Response options
Frequency of
smartphone use
How often do you use a smartphone for activities
other than phone calls or text messaging?
Several times a day
Every day
Several times a
week or less
Smartphone skills Generally, how would you rate your skills of using
your smartphone?
Advanced (5)
Intermediate (4)
Beginner (13)
Do you use your smartphone for the following
activities?
Browse websites Browsing websites Yes
No
Email Reading and/or writing email Yes
No
Photo Taking photos Yes
No
View content on
social media
Looking at content on social media websites/apps
(for example looking at text, images, videos on
Facebook, Twitter, Instagram)
Ye s
No
Post content to
social media
Posting content to social media websites/apps (for
example posting text, images, videos on Facebook,
Twitter, Instagram)
Ye s
No
Online purchase Making purchases (for example buying books or
clothes, booking train tickets, ordering food)
Ye s
No
Online banking Online banking (for example checking account
balance, transferring money)
Ye s
No
Install apps Installing new apps (for example from iTunes, Google
Play Store)
Ye s
No
GPS Using GPS/location-aware apps (for example Google
Maps, Foursquare, Yelp)
Ye s
No
Bluetooth Connecting to other electronic devices via Bluetooth
(for example smart watches, tness bracelets, step
counter)
Ye s
No
Games Playing games Yes
No
Streaming Streaming videos or music Yes
No
22 Mobile Media & Communication 0(0)
Appendix B. Descriptive statistics
Table B1. Descriptive Statistics of Measures before Recoding.
Sample 1 (2017) Sample 2 (2020)
% Missing % Missing
Frequency of smartphone use 1 8
Several times a day 69.2 65.7
Every day 14.4 20.8
Several times a week 8.6 8.9
Several times a month 3.7 2.1
Once a month or less 4.1 2.6
Smartphone skills 1 7
Advanced (5) 23.8 38.2
(4) 28.6 34.0
(3) 30.1 21.6
(2) 12.4 4.6
Beginner (1) 5.1 1.5
N2,186 2,525
Wenz and Keusch 23
Appendix C. Sample 2
Table C1. Predictor Variables by Class of Smartphone Use (Sample 2).
Variables
Advanced
users %
Broad
non-social
media users
%
Broad
non-commercial
users %
Social media
and
information
users %
Basic
general
users %
Camera
users %
Class size 44 9 23 6 11 7
Frequency of
smartphone
use
Several times a
day
82 78 67 45 31 16
Every day 13 19 27 29 35 23
Several times a
week or less
53 6 26 3562
Smartphone skills
Advanced (5) 59 35 30 8 14 7
Intermediate (4) 32 40 4 0 3 0 39 11
Beginner (13) 9 25 30 62 47 82
Smartphone
activities
Browse
websites
100 100 99 76 90 48
Email 99 100 92 65 79 45
Photo 100 97 100 64 91 77
View content
on social
media
100 24 100 74 11 11
Post content
to social media
93 7 71 72 0 0
Online
purchase
100 96 45 40 20 2
Online banking 94 85 37 38 28 3
Install apps 100 97 90 31 80 5
GPS 99 91 85 42 83 11
Bluetooth 94 73 59 23 41 7
Games 87 58 61 49 34 13
Streaming 99 76 78 42 38 5
N1,103 227 582 156 289 168
24 Mobile Media & Communication 0(0)
Table C2. Sociodemographic Characteristics and Operating System by Class of Smartphone Use (Sample 2).
Variables
Advanced
users %
Broad non-social
media users %
Broad
non-commercial
users %
Social media and
information users %
Basic
general
users %
Camera
users %
Chi-squared test
p-Value
Gender 0.006
Female 51 42 52 49 4253
Male 49 58 48 51 5947
Age <0.001
1829 38∗∗∗ 16∗∗ 26 31 8∗∗∗ 8∗∗∗
3039 29∗∗∗ 23 23 21 13∗∗∗ 14
4049 20∗∗ 31 27 22 28 20
5059 12∗∗∗ 21 18 19 36∗∗∗ 39∗∗∗
60 +2∗∗∗ 96 8 16
∗∗∗ 20∗∗∗
Educational
attainment
<0.001
No high school
degree
38∗∗∗ 43 5146 51 57
High school
degree
30∗∗ 28 24 26 19 19
College degree 32 30 25 28 29 24
Operating system <0.001
iOS 33∗∗∗ 21 19∗∗ 28 16∗∗ 12∗∗
Android 64∗∗∗ 78 7663 7981
Other operating
system
30 4 10
∗∗ 67
Note. The asterisks indicate p-values from standardized residuals: p<0.05. ∗∗p< 0.01. ∗∗∗p< 0.001.
Wenz and Keusch 25
... The descriptive statistics for these variables are summarised in Table 1. When we examine the five groups of smartphone users in terms of age, education and position in the labour market (see Table 2), we find that the results are consistent with the literature [38,44]. Advanced users are generally younger (only 7.7% of respondents aged 60 and older belong to this group), while social media use is more consistent across the sample. ...
Article
Full-text available
The main question of this paper is what factors influence willingness to participate in a smartphone-application-based data collection where participants both fill out a questionnaire and let the app collect data on their smartphone usage. Passive digital data collection is becoming more common, but it is still a new form of data collection. Due to the novelty factor, it is important to investigate how willingness to participate in such studies is influenced by both socio-economic variables and smartphone usage behaviour. We estimate multilevel models based on a survey experiment with vignettes for different characteristics of data collection (e.g., different incentives, duration of the study). Our results show that of the socio-demographic variables, age has the largest influence, with younger age groups having a higher willingness to participate than older ones. Smartphone use also has an impact on participation. Advanced users are more likely to participate, while users who only use the basic functions of their device are less likely to participate than those who use it mainly for social media. Finally, the explorative analysis with interaction terms between levels has shown that the circumstances of data collection matter differently for different social groups. These findings provide important clues on how to fine-tune circumstances to improve participation rates in this novel passive digital data collection.
... With the rapid development of information and communication technology, the definition of the digital divide is also constantly being enriched. In 2001, the OECD [34] classified the digital divide into stages: a first-level digital divide stage, with "access" to internet technology as the main difference [35]; a second-level digital divide stage, with "whether there is a need to learn how to use" internet technology as the main difference [31,36]; and a third-level digital divide stage, with the difference in the transformation of internet skills into benefits as the main difference [12,14,30,37]; however, this ignores the motivations and attitudes of the public in terms of digital use [38,39]. In the further development of digital technology, a sense of relative deprivation appears due to the fact of a gap in access to material and technology among digitally vulnerable groups, which reduces the positive participation of individuals in a digital society and an attitude gap reappears [8,40]. ...
Article
Full-text available
In the digital era, digital technology enables political, economic and social fields. However, while the digital divide is pervasive due to the influence of population, geography, the economy and other factors, blindly pushing digital change often leads to the further expansion of the risks of social inequality. This paper constructed a three-dimensional analytical framework of digital gap-bridging policies from the perspective of policy objectives, policy tools and policy effectiveness. Based on this framework, this paper used the content analysis method to conduct a quantitative analysis of 81 digital gap-bridging policies published in China and selected the PMC index model for policy evaluation. The result shows that the policy goal of bridging the digital divide in China mainly focuses on bridging the material access gap and the technology gap, while inadequate attention was paid to the attitude gap and the use gap; in terms of policy tools, there exists an unbalanced situation regarding supply tools, environment and demand tools, with the first one being emphasized more than the others; there is still policy “fatigue” regarding weak feedback and less incentive, though the internal consistency level of the policy is good and the policy structure is reasonable. Therefore, we should optimize policy objectives and policy tools to bridge the digital divide. We should not only consider whether there is “or not” on the equipment and whether there is “or not” on the technology, but also consider whether we want to “or not” on the attitude and feel “or not” on the use; we should not only “give” unilaterally through supply-oriented policies but also “cooperate” with each other through demand-oriented and environment-oriented tools to release the maximum potential of digital empowerment.
Article
Full-text available
Testing communication theories requires a valid empirical basis, yet especially for usage time measures, retrospective self-reports have shown to be biased. This study draws on a unique data set of 923 Swiss internet users who had their internet use tracked for at least 30 days on mobile and desktop devices and took part in a survey covering internet usage as well as person-level background variables. The analysis focuses on active usage time overall and on the major services Google Search, YouTube, WhatsApp, Instagram, Facebook, and the online newspaper 20 Minuten. The results showed that overall internet usage time was lower for older and higher-educated users based on both the tracking and survey data, and the reported usage time was consistently higher than the tracked usage time. The tracking data further revealed that internet users in all social groups spent the majority of their time online on a mobile device. The number of users of the major services varied mainly between age groups. These differences were less pronounced when it came to the time users spent engaging with these services. Over the course of a day, the major services varied in their frequency of use: for example, messaging peaked before noon and in the late afternoon, whereas online news use was comparably constant at a lower level.
Article
Full-text available
Self-reports are the most common way of measuring information and communications technology (ICT) skills in social surveys. Studies comparing self-reported computer skills with objective assessments have found evidence of significant overreporting of skills, but were conducted only among non-representative groups of individuals. This paper fills an important gap by analysing the degree to which ICT skills are overreported in the working-age population of Poland, and the potential causes of this behaviour. We compare answers to Eurostat questions on ICT usage with direct assessments of the corresponding tasks. The results suggest that those individuals who are most likely to possess ICT skills are also most likely to overreport having these skills. The propensity to overreport decreases with age and increases with years of education and numeracy level; women are less likely than men to overestimate their skills. The positive relationship between the probability of a group to overreport their own skills and their expected levels of skills suggests that social desirability bias may explain this phenomenon.
Article
Full-text available
Mobile services are expanding rapidly and in this tremendously dynamic environment companies should provide value-added services to meet users’ demand. In this context, the present research aims to determine whether different user groups exist in this market and profile them. Based on information of 568 mobile service users, a research was developed in the context of a mature mobile services market –Spain-. A behavior-based cluster analysis is developed by means of confirmatory factor analysis (CFA), followed by two-step clustering. Then, an Anova and post hoc Tuckey tests are conducted to confirm differences among the obtained clusters. Research findings show that mobile service users cannot be perceived as a homogenous group, since different users with different behavior coexist in this market. More specifically, four behavior-based segments emerge in the mobile service sector: “service connoisseurs”, “uninvolved pragmatics”, “potential switchers” and “delighted loyal”; being the “potential switchers” the most challenging segment for mobile service companies. This study reports mobile service users’ heterogeneity; and in turn, mobile service managers should consider customers as four different types, instead of considering them as one single customer.
Conference Paper
Full-text available
The prevalence of smartphones has promoted the popularity of mobile apps in recent years. Although significant effort has been made to understand mobile app usage, existing studies are based primarily on short-term datasets with limited time span, e.g., a few months. Therefore, many basic facts about the long-term evolution of mobile app usage are unknown. In this paper, we study how mobile app usage evolves over a long-term period. We first introduce an app usage collection platform named carat, from which we have gathered app usage records of 1,465 users from 2012 to 2017. We then conduct the first study on the long-term evolution processes on a macro-level, i.e., app-category, and micro-level, i.e., individual app. We discover that, on both levels, there is a growth stage enabled by the introduction of new technologies. Then there is a plateau stage caused by high correlations between app categories and a pareto effect in individual app usage, respectively. Additionally, the evolution of individual app usage undergoes an elimination stage due to fierce intra-category competition. Nevertheless, the diverseness of app-category and individual app usage exhibit opposing trends: app-category usage assimilates while individual app usage diversifies. Our study provides useful implications for app developers, market intermediaries, and service providers.
Article
Full-text available
Researchers are combining self-reports from mobile surveys with passive data collection using sensors and apps on smartphones increasingly more often. While smartphones are commonly used in some groups of individuals, smartphone penetration is significantly lower in other groups. In addition, different operating systems (OSs) limit how mobile data can be collected passively. These limitations cause concern about coverage error in studies targeting the general population. Based on data from the Panel Study Labour Market and Social Security (PASS), an annual probability-based mixed-mode survey on the labor market and poverty in Germany, we find that smart-phone ownership and ownership of smartphones with specific OSs are correlated with a number of sociodemographic and substantive variables.
Article
Full-text available
The integration of 5G networks and AI benefits to create a more holistic and better connected ecosystem for industries. User profiling has become an important issue for industries to improve company profit. In the 5G era, smartphone applications have become an indispensable part in our everyday lives. Users determine what apps to install based on their personal needs, interests, and tastes, which is likely shaped by their genders - the behavioral, cultural, or psychological traits typically associated with one sex. It is possible to profile users' gender based simply on a single snapshot of apps installed on their smartphones. With this inference based on easy to access data, we can make smartphone systems more user-friendly, and provide better personalized products and services. In this paper, we explored such possibility through an empirical study on a large-scale dataset of installed app lists from 15,000 Android users. More specifically, we investigated the following research questions: 1) What differences between females and males can be explored from installed app lists? 2) Can user gender be reliably inferred from a snapshot of apps installed? Which snapshot feature(s) are the most predictive? What is the best combination of features for building the gender prediction model? 3) What are the limitations of a gender prediction model based solely on a snapshot of apps installed on a smartphone? We found significant gender differences in app type, function, and icon design. We then extracted the corresponding features from a snapshot of apps installed to infer the gender of each user. We assessed the gender predictive ability of individual features and combinations of different features. We achieved an accuracy of 76.62% and AUC of 84.23% with the best set of features, outperforming the existing work by around 5% and 10%, respectively. Finally, we performed an error analysis on misclassified users and discussed the implications and limitations of this work.
Article
This paper presents findings on the correspondence of levels of digital systems and media use with a range of socio-economic and demographic measures in the U.K. Most research on inequalities in regard to digital systems and media has focused on access and skills. Building on prior work (Yates and Lockley, 2018; Yates, et al., 2015) we argue that inequalities in regard to digital systems and media are better understood around types of user and their correspondence to other key social variables — rather than solely individual skills and access. The analysis presented here covers a range of key demographic variables, especially those that are markers of distinct social disadvantage. We find that those not using the Internet have distinct characteristics — predominantly around age, education and deprivation levels. We also find that those undertaking limited uses (overall limited use or a very narrow range of uses) are all predominantly from lower socio-economic status backgrounds with variations due to age and education. The data used for the analysis is the recent U.K. Ofcom 2018–19 (n = 1,882) media literacy survey. The paper uses latent class analysis methods to inductively define user types. Multinomial and binary logistic regression are used to explore the correspondence of latent class group membership to key demographic variables. These insights have direct U.K. and international policy relevance as they are key to the development of strategies to tackle ongoing digital inequalities in U.K. society.
Article
The dimension used to measure user heterogeneity plays a key role in the customer segmentation. For most traditional products, customer requirements (CRs) for products are often related to their basic characteristics and psychological characteristics. Therefore, in the traditional market segmentation theory, the dimensions used to distinguish customer differences include the demographic attributes such as gender, age, income etc., or the customer psychology. Customer behaviour, especially the purchasing behaviour, is also used as an important dimension for market segmentation. However, for the smart product like smartphone with rich functionality and multi-interactions, the customer’s interaction preference can be fully released. That makes the heterogeneity of customer stems from the usage characteristics rather than the traditional demographic attributes. Hence, the usage pattern is defined and proposed as the description of the usage characteristics and be used to measure the heterogeneity of customers in this research. The Equivalence CLAss Transformation (ECLAT) algorithm is employed to identify the customer’s APP frequent sets from the operating data and to construct the usage pattern. Thereafter, the customer can be segmented based on the distance among customers’ usage pattern. Compared with the demographic attributes, the usage pattern can provide more reliable and truthful measures for the smartphone customer segmentation.