Int. J.

Human–Computer Studies 170 (2023) 102965

Contents lists available at ScienceDirect

International Journal of Human - Computer Studies

journal homepage: www.elsevier.com/locate/ijhcs

Why people choose Apps: An evaluation of the ecology and user experience
of mobile applications
Ons Al-Shamaileh a, Alistair Sutcliffe b, *
a
College of Interdisciplinary Studies, Zayed University, UAE
b
University of Manchester, UK

A R T I C L E I N F O A B S T R A C T

Keywords: Purpose: To investigate the reasons for users’ choice of mobile applications and how their choice relates to their
Mobile applications experience of use.
User experience (UX) Method: A mixed methods study of the factors influencing users’ choice to adopt or abandon mobile applications.
Technology acceptance
Seventy-nine respondents completed a questionnaire recording their top four favourite applications, the fre­
quency of use and user experience measures: aesthetics, content, usability, pleasurable interaction, and overall
experience. They also reported up to four abandoned Apps, with any alternatives considered and the reasons for
use or abandoning. Follow-up interviews probed the reasons for users’ choice of specific applications.
Results/Conclusions: Social media was the most favoured category of App, followed by leisure, e-commerce, and
communication. Quantitative data shows that content, usability and pleasure predict overall user experience and
App acceptance. Interview data indicate that user’s choice of downloading and abandoning applications is also
influenced by usefulness, usability, content, reliability and contextual factors such as networking and recom­
mendations. Most user App choices appear to be fast-path decisions made without systematic comparison of
products.

1. Introduction appear to be pragmatics (an amalgam of utility and usability) and aes­
thetics/hedonics (Diefenbach and Hassenzahl, 2009; Lavie and Trac­
Applications on mobile phones have developed into a fiercely tinsky, 2004). However, interactive features may also influence user
competitive marketplace on both Android and iOS models (Auxier and judgement of effectiveness, efficiency and overall attitude (Teo et al.,
Anderson, 2021). While positive reviews, recommendations and word of 2003; Cyr et al., 2009; De Angeli et al., 2006). In a study of product
mouth may drive market success for many Apps, there is little under­ acceptance amongst medical students, Hart and Sutcliffe (2019) found
standing about the relative influences of utility, usability, and other that functionality of the device (iPAD) and Apps was the most important
factors on why people download and use particular Apps or reject them, influence on acceptance, and useful functions overcame poor perceived
outright or after a period of use. Several researchers have investigated usability.
the importance of mobile applications and studied factors influencing In information systems, IT product choice has been investigated in
user choice, including personalisation (Tunney, 2018), practicability technology acceptance models (TAM) (Venkatesh et al., 2003, 2012;
(Gefen et al., 2003), influence of society (Venkatesh et al., 2003) and Williams et al., 2015). TAM models have evolved from a core set of
performance (Malik et al., 2017). In this paper we investigate whether or concepts: behavioural intention, PEoU (perceived ease of use), PU
not user experience is an important influence on users’ choice and de­ (perceived utility), to include other influences such as trust, subjective
cisions to keep using or abandoning Apps. norms and facilitating conditions (Venkatesh et al., 2012). Model elab­
User Experience (UX) research has identified usability, service oration produced UTAUT, containing several variables describing user
quality (similar to utility), classic and expressive aesthetics (Lavie and attitudes (performance and effort expectancy, hedonic motivation), user
Tractinsky, 2004) as important influences on users’ perceptions of characteristics (e.g., age, gender, experience, self-efficacy, and habit),
products. While no overall consensus of variables influencing overall and economic considerations such as price and value (Venkatesh et al.,
judgement has emerged in UX research, the more important components 2003; Williams et al., 2015). Other antecedent variables that may

* Corresponding author.
E-mail addresses: [email protected] (O. Al-Shamaileh), [email protected] (A. Sutcliffe).

https://doi.org/10.1016/j.ijhcs.2022.102965
Received 29 March 2022; Received in revised form 10 November 2022; Accepted 13 November 2022
Available online 17 November 2022
1071-5819/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

influence behavioural intention have included user self-image, culture, (Hartmann et al., 2007; Sutcliffe, 2009; Hart and Sutcliffe, 2013). Kujala
predisposition towards technology (Magni et al., 2010), and compati­ et al. (2011) reported that both pragmatic and hedonic qualities
bility with the users’ tasks and organisational setting (Yousafzai et al., contributed to attractiveness over a 6–12 month period; this is supported
2007; Williams et al., 2015). by Mendoza et al. (2005) finding of decreased frustration over time, as
In spite of several studies, the rationale for users’ choice of IT users overcome initial usability problems. However, overall preferences
products generally and mobile Apps in particular remains unclear. for websites with similar content but different designs may depend on
Domain-specific factors may be important; however, a more general framing effects of tasks and users’ characteristics (Porat and Tractinsky,
model of technology acceptance for mobile Apps is needed. Previous 2012). Interactive system features may also influence user judgement of
studies have not illuminated how users make decisions to download and effectiveness, efficiency and overall attitude (Teo et al., 2003). In a study
use Apps; for instance, do they just follow peer pressure, as appears to be of product acceptance amongst medical students, Hart and Sutcliffe
the case in social media (Aloudat et al., 2019; Lee et al., 2012), or do (2019) found that functionality of the device (iPAD) and Apps was the
they make more rational decisions? Decision Making Theory (Cacioppo most important influence on acceptance, with useful functions over­
et al., 1986; Kahneman, 2011) asserts most people, most of the time coming poor perceived usability. However, contextual factors such as no
make ‘fast-path’ choices based on a few dominant attributes and little perceived need, lack of training and poor fit with working practices also
conscious effort. In contrast ‘slow-path’ choices are evaluated more influenced rejection of the iPAD. Although several conceptual frame­
carefully with trade-offs and mental models of the choice space (John­ works or models of UX have been proposed with a variety of constructs
son-Laird, 1983). Choice of social media Apps may be fast path; how­ such as aesthetics, immersion, and presence (Hartmann et al., 2009)
ever, other Apps may trigger more systematic, slow-path reasoning. This there appears to be little consensus between generic and contextual
perspective forms one of the motivations for this enquiry, not only which interpretations of UX (Bargas-Avila and Hornbaek, 2011). UX studies
variables influence choice but also how users make choices across have not interpreted empirical data on influencing variables in light of
different App types. This paper investigates criteria in determining Decision Making Theory, although a proto-theory of variables and
users’ decisions to adopt/abandon Apps on mobile phones, with the temporal influences on users’ perceptions of experience was proposed
following research questions: by Hartmann et al. (2009).
IT product choice has been investigated extensively in technology
RQ1. What types of App are present in users’ portfolios on their acceptance models (TAM) which have evolved over several decades
mobile phones? (Williams et al., 2015), to include a variety of influences such as trust
RQ2. What are the more important determinants of users’ choice ? and risk, subjective norms and facilitating conditions (Venkatesh et al.,
RQ3. How is users’ choice influenced by the App type/purpose? 2012; Maruping et al., 2017), user characteristics (e.g. age, gender,
RQ4. Which cognitive processes (slow/fast path) do users follow experience, self-efficacy and habit), and economics, such as price and
when making App choices? value (Venkatesh et al., 2003, 2012; Williams et al., 2015). Other
antecedent variables that may influence behavioural intention have
The paper is organised as follows: first we review related research, included user self-image, culture, predisposition towards technology
followed by a description of the study methods. The quantitative and (Magni et al., 2010), and compatibility with the users’ tasks and
qualitative results are presented, and the paper concludes with discus­ organisational setting (Yousafzai et al., 2007; Williams et al., 2015). Lee
sion on technology acceptance, user experience and reasons for choice of et al. (2012) proposed a Unified Theory of Acceptance based on a survey
Apps. of college students and office workers, finding that personalisation has a
positive effect on performance expectancy, intention to use, and rec­
2. Related research ommendations to other users. Pan and Zhao (2018) reported reasons
that led to abandoning mobile healthcare Apps, ranging from difficulty
With the growth of the Internet and mobile communication tech­ faced in understanding the user interface to constraints on data entry
nologies, a wide range of mobile applications have been produced requirements, and failure to update diagnosis. Some of the major factors
(Islam et al., 2010). Malik et al. (2017) analysed factors influencing for improving the continuance of the mHealth App were found to be
customers’ perspectives in adopting a particular mobile App, including integration with wearable healthcare devices and enhancing
ease of use, social influence, trust, performance and incentives. Harris patient-to-physician interaction (Pan and Zhao, 2018). Hornbaek and
et al. (2016) identified reputation of the App in the market and famil­ Hertzum (2017) reviewed TAM and UX research, concluding that there
iarity as influences leading users towards installing or rejecting an App. was little common ground between the models in the two research areas,
Noh and Lee (2016) investigated the factors influencing consumers to and that TAM models did not account for the experiential perspective of
use banking Apps and showed that the intention to use is based on use. They argued that construct definitions needed to be improved, as
features such as safety and service quality; however, this research was well as the links to cognitive models, tasks and the context of use. While
conducted on the South Korean banking sector and more diverse studies the psychological Theory of Reasoned Action (Ajzen 1991) underpins
in different cultures are necessary to validate these results. Wei et al. TAM models as antecedents of behavioural intent (users’ choice), TAM
(2015) investigated how mobile gaming Apps are downloaded, and why studies have not investigated the users’ reasoning processes while
95% of them are abandoned within a short period. They considered that making product or App choices.
pricing, initial attractiveness, quality, followed by enhanced benefits Overall, six main factors appear to be the more important influences
and monetary rewards, foster user loyalty and play a significant role in on user preferences and acceptance of mobile phone Apps: person­
the continuous usage of any gaming App. alisation of the level at which the desired service or information provi­
In User Experience (UX), components that contribute to users’ sion and features are customisable by the user (Tunney, 2018; Morosan
overall judgement of IT products have been identified as usability, ser­ and DeFranco, 2016; Sang-Hyeon and Sang-Hyun, 2002; Dong-Wook
vice quality (similar to utility), classic and expressive aesthetics (Lavie et al., 2008; Wei et al., 2015); integrity, including users’ trust in the App
and Tractinsky 2004); while pragmatics (an amalgam of utility and us­ and/or its supplier/ provenance (Wang et al., 2006; Davis et al., 1989;
ability) and hedonics were proposed by Hassenzahl (2004) as anteced­ Dai-yon and Hyun-Jung, 2008; Baldwin et al., 2017; Adhikari et al.,
ents to judgement of general product qualities of ‘goodness’ and 2014; Pan and Zhao, 2018; Addonizio, 2017); performance of the App for
‘beauty’. However, overall preferences for websites with similar content improving users’ work and effectiveness, and also performance in sup­
but different designs can be swayed by framing effects of tasks and users’ porting other user tasks (Gefen et al., 2003; Venkatesh et al., 2003;
characteristics (O’Brien 2010). Interactive system features may also Malik et al., 2017); utility, perceived usefulness of the App, its func­
influence user judgement of effectiveness, efficiency and overall attitude tionality and services (Gefen et al., 2003; Venkatesh et al., 2003; Ding

2
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

and Chai, 2015; Al-Shamaileh et al., 2012a, b); external influences on the How long they had used their chosen Apps (as reported in the
decision to choose an application including overall popularity, recom­ survey).
mendation from friends, family, reviews, social media networking effect The reasons for participants’ choice of Apps which they continued to
(Aloudat et al., 2019; Gye-Soo, 2002; Al-Shamaileh and Sutcliffe, 2012a, use.
Al-Shamaileh and Sutcliffe, 2012b; Al-Shamaileh et al., 2011;Al-Sha­ Whether any alternatives were considered in their choice.
maileh, O., 2018; Lee et al., 2012); and facilitating conditions, which may Comments on user experience.
be social and organisational reasons for choice (e.g. company stan­ List any abandoned Apps or any they had downloaded but used
dards), training and technological support for the App (Noh and Lee, infrequently.
2016; Wei et al., 2015; Hart and Sutcliffe, 2019). Reasons for abandoning or not using these Apps.
Comments on user experience.
3. Study design & methods
Interviews were audio recorded and subsequently transcribed. Both
A mixed methods approach (Creswell et al., 2003) was chosen with a the interview transcripts and text responses in the survey were analysed
survey to collect quantitative data on App choice, user experience and to investigate reasons for using or abandoning Apps. Analysis followed a
preferences, with free-format replies to collect qualitative data on cho­ hybrid approach (Cresswell et al., 2003) with preset categories and open
sen or abandoned Apps and users’ reasons for their choice. Participants coding (Holton, 2007) for choice reasons and other utterances which
were recruited through advertising the study on Facebook and the Zayed were not associated with preset categories. User experience responses
University portal. The survey was completed in person and com­ were coded using the UX concepts derived from the survey: usability,
plemented by interviews to gain further insight into users’ decision content, aesthetics, satisfaction and pleasure. Sentences and utterances
processes for App choice. which did not fall into these categories were analysed using open coding
to classify emergent themes and reasons for choice. The main themes
3.1. Survey design which emerged from open coding were illustrated with excerpts.
The interview transcripts were analysed in more depth to investigate
Respondents were asked to complete a questionnaire consisting of the decision processes for App choice, to categorise the evidence users
demographic information (age, gender, nationality and occupation) and reported to justify their choice, based on the UX concepts listed above
the following sections: plus any emergent categories such as cost/free App, convenience,
communication (social media) and how many alternatives were
(i) Identify the top four mobile applications that they used. considered in their choice (if any). Analysis of the abandoned or infre­
(ii) Rate the frequency of use, satisfaction and mood based on their quently used Apps followed the same process. The relative complexity or
experience with their first application on a 1 to 7 scale where 1 = simplicity of the reported decisions (Sloman 1996; Caciappo et al. 1986)
frequent or very positive. were considered to classify the user’s decision process as fast/slow path
(iii) Evaluate their experience of their first application on five scales: using the criteria:
expressive aesthetics, usability, pleasurable interaction (Lavie
and Tractinsky, 2004), content adapted from Bernier Instruc­ Fast path: one or two reasons for choice cited, no alternative
tional Design (De Angeli et al., 2006) and their overall experience considered for adopted Apps. For abandoned Apps 1–2 reasons and
with three questions (O’Brien, 2010) on a 1 to 7-point scale. possibly an alternative selected.
(iv) List any alternative they considered as competitors for their first Slow path: >2 reasons for choice cited, evidence for considering al­
chosen application (free-format reply). ternatives for adopted Apps. For abandoned apps more detailed
(v) State their reasons for selecting their first App (free-format reply). reasons, evidence for systematic search and selection of an
(vi) Repeat steps (ii) to (v) for their second App. alternative.
(vii) Repeat steps (ii), (iv-v) for their third and then fourth Apps.
(viii) Identify four applications that they had downloaded and subse­ The authors coded a sample of ten interviews independently then
quently stopped using. compared utterance codings, resulting in an inter-coder agreement of
(ix) Rate the frequency of use and then repeat steps (iii) to (v) for their 87%. Differences were reconciled and a list of emergent categories
first abandoned App. produced. The first author then completed the interview coding with a
(x) Repeat step (ix) for their second abandoned App. further sample of ten interviews dual coded to check inter-coder
(xi) Rate frequency of use and repeat steps (iv) and (v) for third and agreement. The second sample produced agreement of 98%.
fourth choice abandoned Apps (if any cited). The duration of each survey and interview was approximately one
hour. The survey was carried out online between January and April
The user experience scales (step iii) were chosen because they have 2019 using Qualtrics. Participants were sent a link to the survey and
an established track record in the HCI literature and covered the aspects specific date/time completion sessions were scheduled with a Zoom
of App choice more appropriately that TAM/UTAUT models (Venkatesh video conference so the first authors could introduce the survey and
et al., 2012) which use more limited usability and utility scales. Users answer any questions. Participants were then asked to complete the
had access to their mobile phones while completing the survey; how­ online survey at their own pace and a sub-set (30/79) were interviewed
ever, they primarily relied on their memory for listing popular and directly after they completed the survey. Interviews were self-selected
abandoned Apps. Limited data sets (frequency of use, alternative and by volunteering in advance for the post-survey interview. Most in­
reasons) were collected for Apps 3 and 4 because several respondents terviews were conducted in the same Zoom session as the survey;
did not cite any Apps for their third and fourth choices; also in pilot however six respondents preferred to be interviewed at a later date in
interviews they reported that their memory of experience for these Apps 2019 and early 2020. These interviews followed the same structure as
was limited. The questionnaire is illustrated in Appendix A. the previously completed interviews.
Qualitative data analysis from the interviews and responses to the
3.2. Interviews free format survey questions on reasons for accepting/rejecting app
produced the following thematic categories (see Table 1) which we
After they had completed the survey, 30 participants were inter­ describe as factors that influence users’ decisions.
viewed to elicit their opinions on the reasons behind using/abandoning Excerpts from interviews are given in the results Section 4.4. for each
each application. The interview was structured with seven questions: of these themes. In the discussion of results factors refers to both the

3
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

Table 1 4. Results
Theme category definitions followed in qualitative data analysis.
Theme Inclusion criteria 4.1. Descriptive statistics
Usefullness; General evaluations of the utility and quality of service of Apps,
with some specific comments on the usefulness of features and Frequencies of the applications selected and used by respondents are
functions. illustrated in Table 2. Social networking applications were reported
Usability: Positive and negative comments on user interface features and most frequently, followed by communications and other applications
operations. Reports of problems experienced while operating with an even spread (4–7%) amongst other categories. Within other
Apps
Content: Comments on quality of information delivered by Apps or
social media Twitter and LinkedIn were more frequent. Communication
content in their web site. Apps reflected the prevalence of e-mail and audio and video commu­
Alternatives: Identification of specific apps or category of Apps either nication, although some of the expected market leaders, (Skype, Zoom,
investigated and/or adopted in response to the ‘Alternative Teams) were not reported frequently. Google as a search engine was
considered” survey question and references to alternative
probably under-reported as it may have been an assumed App by re­
choices in interviews.
Reliability: Comments relating to App failures, crashes, or other reasons for spondents who cited Google Maps, Translate and News. There was a
not operating as expected. considerable variation in choice of individual Apps within the other
Security Concerns about potential loss of personal data, poor security categories. After the four most popular social media, only four more
protection, potential exposure to fraud and malevolent actor in frequent Apps were chosen by 2–3 users, while all the others were re­
the Internet environment.
Addiction: Specific comments about the addictive nature of social media
ported by a single user.
Influence of Reasons for choice influenced by family, friends, word of mouth Overall, social media dominated our respondents’ choice, account­
others: or internet/press reviews ing for 75.9% of first choices, with the tail of other (13%) including
Popularity: Reasons for choice where the App was perceived to be the information (Google Maps 5%) and finance Apps (2.5%). Second choices
market leader/ most popular.
were also dominated by social media (65.5%), with a long tail of single
Design: General comments on the design quality of Apps.
Overall Statements summarizing respondents’ general opinion of Apps. user Apps. Third and fourth choices demonstrated more diverse choice
judgement: with leisure Apps (games, video streaming services), finance (online
banking) and communications with video conference Apps (BOTIM,
Viber). Some choices might reflect the demographics of respondents
above influences on users’ decisions as well as quantitative data mea­ residing in the UAE, e.g. Indian communications and social media Apps.
sures analysed in the survey: usability, aesthetics, content, pleasure, Two-thirds of respondents (64.5%) did not consider any alternatives
mood, satisfaction and overall experience. for their first choice, and of those that did nearly all were other popular
social media (WhatsApp, Instagram, Facebook). Similarly, 70.8% did
3.3. Participant demographics not consider an alternative for their second choice, and other social
media Apps formed the majority of alternatives. A similar pattern was
Seventy-nine respondents participated in the experiment; 34 were found for third and fourth choice Apps, with most respondents (75.9%
male (43%) and 45 female (57%). Forty worked in the private sector third and 72.2% fourth choices) not considering any alternatives,
(51%), 11 were unemployed (14%), ten worked in the public sector although the alternatives that were considered were split between social
(13%), nine were self-employed (11%) and nine were students (11%). media and other Apps in the same category as their chosen App. The
Eight respondents were aged 18–25 (10%), 33 were aged 26–35 (42%), totals for abandoned Apps are illustrated in Table 3.
32 were 36–45 (41%), five 46–55 (6%) and one was older than 55 (1%). The more popular social media, WhatsApp, Facebook, Instagram and
The respondents were from Jordan (30, 37.9%), India (26, 32.9%), UAE Snapchat, which are shown separately in Table 2, are aggregated in
(6, 7.6%), non-Middle East (Canada, USA and France each 5, 6.4%), and Table 3 because overall frequencies were lower. Few social media Apps
1 to 3 respondents each from other countries (Egypt, Lebanon, Pakistan, were abandoned, although this may reflect the limited choice of avail­
Palestine, Sri Lanka and Syria, 13, 16.5%). The only large demographic able alternatives, including regional variations such as Orkut (an Indian
sub-group was private sector employees who were Jordanian or Indian Facebook rival). Leisure, e-commerce, communication and other cate­
nationals (39% of respondents); otherwise, there were no intersecting gories accounted for most of the abandoned Apps. Abandoning of
groups of occupation, nationality and age. Google Apps was limited to communications, Hangout (meetings) and

3.4. Classification of Apps

The Apps identified in the survey were classified into the following Table 2
categories: Totals of Apps used by category, in choice 1–4 order.
First Second Third Fourth Total
Social media (most popular), with individual totals for the more (%)

frequently reported: Facebook, Instagram, WhatsApp, SnapChat. WhatsApp 36 11 7 6 60

Other social media: Twitter, LinkedIn, etc. (21.1)
Instagram 6 13 14 5 38
Google: all applications: Maps, Hangouts, Meet, etc.
(13.3)
Finance: e-banking and personal finance management Apps. Facebook 10 21 8 8 47
Leisure: film and video streaming services, YouTube, games, other (16.5)
entertainment Apps. Snapchat 2 6 5 7 20 (7.0)
Communications: Internet audio and video utilities (Skype and ri­ Other social media 5 4 3 3 15 (5.3)
Google 4 2 4 2 12 (4.2)
vals), e-mail, VPNs, Zoom, Teams, video conferencing. Finance 2 2 6 8 18 (6.3)
E-commerce: e-shops, purchase and delivery services, e-booking Leisure 2 0 9 4 15 (5.3)
services. Communications 4 7 4 12 27 (9.5)
Information: maps, weather, any advice and information delivery. E-commerce 3 4 6 5 18 (6.3)
Information 2 4 3 2 11 (3.9)
Other: including utilities (file share), functional Apps, e-government.
Other 3 5 7 7 22 (7.4)
Total 79 79 76 69 303
(26.1%) (26.1) (25.1) (22.7)
None cited 0 0 3 10

4
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

Table 3 were higher for abandoned Apps indicating more diversity in users’
Totals of Apps abandoned, by category, in choice 1–4 order. judgement.
First Second Third Fourth Total (%) The analysis of UX scales’ differences with paired-sampled t-test
showed that differences in favour of retained Apps were significant for
Social media 1–4 5 4 4 2 13 (5.1)
Social media >4 4 3 2 3 12 (4.7) all measures: aesthetics t 78 = 5.7, p<.000, usability t78 = 9.7, p < .000,
Google 1 2 1 1 5 (2.0) pleasurable interaction t78= 5.97, p<.000, content t78=4.4, p<.000,
Finance 4 6 6 3 19 (7.5) overall judgement t78 =9.7, p=<0.000.
Leisure 16 7 3 10 36 (14.2) Similar differences were found when the first choice Apps used/
Communications 13 20 7 7 47 (18.4)
E-commerce 12 15 15 14 56 (21.9)
abandoned were compared (t tests all p<.000, apart from content p<.01)
Information 3 1 4 3 11 (4.3) and second choice Apps used/abandoned (t tests all p<.000). There were
Other 15 15 13 13 56 (21.9) no significant differences for all UX measures between first and second
Total 73 73 55 56 257 choices for either used or abandoned Apps, so it appears the choice order
None reported 6 6 24 23 59
made little difference. There were no significant differences in overall
experience, satisfaction, and other UX measures on used/abandoned
Duo. The low frequency of abandoned finance Apps may reflect limited App by age, nationality or occupation, although small numbers in some
choice, constrained by users’ choice of bank, payment service, and demographic categories limited the scope of testing (ANOVAs).
personal finance manager Apps. The higher frequency of abandoned
communications Apps may be a facet of our international participants 4.2.1. Influence of UX on satisfaction and overall experience
living in the Middle East, where a number of video and audio call Apps Regression tests were carried out with UX factors as predictors for
are available from India and other countries. The first and second choice four dependant variables: frequency of use, mood, satisfaction, and
scores indicate possible experimentation in evaluating video and audio overall experience; with gender as the control variable.
call services, which may have been abandoned in favour of established Frequency of use was not predicted by UX factors for App1 or App2
market leaders, e.g. Skype, Zoom, Teams, although Skype did score four used, and the same result was found for Apps1 and 2 not used/aban­
abandons. Abandoning e-commerce Apps may indicate users sampling a doned (r2 models all n/s).
variety of local (UAE) and international purchase and delivery services, Satisfaction r2.325 p<.000 was predicted by pleasure (β.487,
e-shops, e-malls, and booking services. Abandoning other Apps may p<.001) and weakly by usability (β 0.328, p<.05) for App1 used, mood
reflect dissatisfaction with downloaded utilities (e.g. file sharing), (r2 0.369 p<.000) only by pleasure (β 0.518, p<.000), while overall
functional Apps (e.g. photo-editing), e-government, and educational experience (r2 0.648 p<.000) was predicted by usability (β.311, p<.01),
services. Few information services were abandoned, and the main pleasure (β.364, p<.000) and content (β.360, p<.000).
reason for rejection was no need for, or poor local traffic information. The predictions for App2 used followed a similar pattern, with
Overall, the lower percentage reporting of abandoned Apps for choices 3 satisfaction (r2 0.426 p<.000) being weakly predicted by pleasure (β
and 4 may reflect poor recall by our participants. 0.307, p<.05), usability (β 0.318, p<.05) and content (β.278, p<.05);
Apps downloaded and later abandoned were diverse, with few in­ mood (r2 0.523, p<.000) by pleasure (β.558, p<.000), and weakly by
dividual Apps being cited by more than two users. Social media and e- usability (β.255. p<.05); while overall experience (r2 0.721 p<.000) was
commerce (Careem, Noon, Talabat) were more frequent categories, each predicted by usability (β.482, p<.000), pleasure (β.312, p<.001) and
with 33% of the total first cited as abandoned, with information (Road content (β.248, p<.01).
Traffic Authority, RTA), entertainment (Shahid) and communications For abandoned Apps only the overall experience was recorded. This
(VPN) each accounting for 11.3%. The second-place abandoned Apps (r2 0.480 p<.000) was predicted only by content (β.401, p<.001) for the
were dominated by a long tail of single user Apps (75.9%), with com­ first abandoned App and the same for App2 (r2 0.639, p<.000), content
munications 15% (Google, IMO video calls), and social media 9.1%. (β.388, p<.000).
Respondents considered more alternatives for abandoned Apps, with The results of the regressions for predicting overall experience,
just over half reporting that they had alternatives (first 51.9%, second satisfaction and mood by UX factors for used Apps are summarised in
58.3%, third 52.7% and fourth 54.6%) and about half of those re­ Fig. 1.
spondents citing specific alternatives that they considered (first 61.8%, Regression for the third and fourth Apps used tested the frequency of
second 42.5%, third 52%, fourth 40%), of which most were Apps in the use with mood and satisfaction as predictors. No significant results were
same class as the abandoned one. found for App3, and only a weak relationship for App4 with mood
predicting frequency (r2 0.131, p<.05, β.- 0.335, p<.05).

4.2. User experience 4.3. Qualitative data results

Cronbach alphas were calculated to explore the internal consistency 4.3.1. Apps used: survey data
of the user experience questionnaire scales. Values ranged from 0.85 to Usefulness and usability were most frequently mentioned by re­
0.95 for all scales, so aggregate averages for all scales were used in spondents, accounting for 76% of all comments. Comments on the used
subsequent statistical tests. applications were nearly all positive since they reflected users’ reasons
Users rated Apps they retained higher than abandoned Apps on all for their choice; see Table 5 for the frequency of comments for used Apps
user experience measures (see Table 4), although standard deviations by category in citation order. (A small number of responses fell into two
categories, hence some column totals exceed N = 79 respondents.)
Table 4 Usefulness comments were mainly general evaluations of the App,
User experience ratings for used and abandoned Apps. although a minority (10%) gave specific comments on features and
Used Apps 1&2 Abandoned Apps 1&2 functions. Connectivity, networking, keeping in touch and communi­
cation were the most frequent comments for social media, e.g., “It’s fun
Scale/measure Mean SD Mean SD
Aesthetics 5.02 .92 4.11 1.19 and keeps me updated with people and celebrities I care about” (P-9).
Usability 5.97 .89 4.48 1.29 Usability comments tended to be general evaluations of ease of use:
Pleasurable interaction 5.14 1.04 4.14 1.21 “Easy to use; easy to post pictures and videos” (P-15). Other comments
Content 5.22 1.04 4.48 1.27 included social reasons for choice, (recommended by friends/family):
Overall judgement 5.77 .96 3.86 1.43
“My parents still use it” (P-10), general amusement or passing time and

5
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

Fig. 1. Relationships of user experience factors and overall measures of experience, satisfaction and mood for Apps used 1 (black lines) & 2 (red lines).

Table 5 Table 7
Frequency of comments for Apps used: survey answers. Frequencies of comments for Apps abandoned, by citation order.
Application Used Apps citation order

1 2 3 4 Total (%) Comment category First Second Third Fourth Total (%)

Usefulness 50 52 46 46 194 (61.2) Usefulness 22 21 22 17 82 (31.6)

Usability 10 12 12 8 42 (13.2) Usability 14 14 11 8 47 (18.1)
Content 1 2 4 3 10 (3.1) Alternatives 10 6 9 10 35 (13.5)
Alternative 2 1 1 1 5 (1.6) Content 4 3 2 2 11 (4.2)
Other 15 12 11 14 52 (16.4) Value/Price 2 2 5 1 10 (3.8)
No Reasons 1 2 4 8 14 (4.4) Other 14 15 5 7 41 (15.8)
80 81 78 80 317 No Reasons 4 11 8 10 33 (12.7)
70 72 62 55 259

occasionally social media addiction. Nearly all comments were positive,

although comments in some other categories were neutral. The fre­ anymore” (P-11), with others giving some detail about poor function­
quencies for usefulness and usability of the more popular (social media) ality or usability problems, e.g., “Not very easy to use despite its great
Apps are shown in Table 6. content and nice design” (P-22). Content and value for money were
Of the three more frequent comment categories, usefulness was infrequent reasons. Other comments ranged from social contextual
associated with general observations about connectivity, contacting justification for choice, e.g., moving country so the App was no longer
friends and family, while other comments on social media reported relevant: “I only used it in Egypt and once in Jordan because I didn’t
networking effects, the addictive nature of social media and tacit coer­ have my car” (P-4), to criticisms of quality, “The quality of the voice was
cion, e.g., “everyone’s on it”, “difficult to avoid”. WhatsApp attracted not good, I couldn’t hear my friend” (P-23).
more positive usability comments than other social media, with general
ease of use opinions. 4.3.3. Apps used: qualitative data (interviews & survey1)
The focus of the interview questions was on the reasons for first and
4.3.2. Apps abandoned: survey data second App choices. Most responses were allocated to the following UX
Poor usefulness and a few specific comments on functionality (see and open derived themes: usefulness, usability, content, reliability, se­
Table 7) were the most common reasons cited for abandoning Apps, curity, addiction to social media, unavailability of alternatives, influ­
followed by usability and then consideration of alternatives and finding ence of others, popularity, design, and overall judgement, although
one. These comments were nearly all negative criticisms, with approx­ there were several other low frequency comment themes: see Table 8.
imately 40% being general observations, e.g., “Don’t find it very useful Although there were more comments for the first choice App (64.5%)
than the second choice (35.5%), there were no marked differences in the
distribution of comment themes, apart from usability, where the first
Table 6 choice App received a much higher frequency. Overall, the interview
Frequency of usability and usefulness comments for more popular social media data was consistent with the survey open-format replies, with most
and other App categories. comments focused on usefulness, usability and content. First and second
Usefulness Usability Other choice Apps were nearly all social media that had been used for several
WhatsApp 48 17 13
years, (range 2 - 10+, median 5 years). Non-social media first and sec­
Facebook 39 4 10 ond choice Apps had been used for shorter durations, median 2 years.
Instagram 31 4 9 Respondents considered usefulness and usability as the most
Snapchat 12 1 2 important reasons to choose an application: category frequencies 55%
Other social media 9 1 2
and 21% respectively; see Table 7. The main social media usefulness
Google 11 1 0
Finance 13 1 2 themes were social connections, keeping up with friends, communica­
Leisure 0 1 5 tion and chatting, as well as photo sharing. Facebook themes tended
Communications 11 1 6
E-commerce 12 1 2
Information 7 1 2
Other 1 2 8 1
Excerpts P-1 to 30 are interviewed participants, P-31 to 79 free-format
194 36 52
comments from the survey

6
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

Table 8 most popular texting App” (P-37). Comments on aesthetics and user
Interview data: frequency of comments for Apps used. interface design were rare.
App used 1 App used 2
4.3.4. Apps abandoned: interview data
Category Comment % Comment % Totals %
frequency frequency For abandoned applications, the majority of the comments were
negative on usefulness and usability (57%); other statements included
Usefulness 71 54% 42 58% 113 55%
Usability 34 26% 8 11% 42 21%
finding better alternatives (23%), criticism of content (12%), and lack of
Content 7 5% 2 3% 9 4% time to try out the App and overall judgement; see Table 9. In contrast to
Reliability 5 4% 1 1% 6 3% used Apps, no tail of low frequency comment themes emerged from the
Security 3 2% 1 1% 4 2% analysis. There were no marked differences in the distribution of theme
Addiction 3 2% 2 3% 5 2%
frequencies between the first and second abandoned Apps, although
No alternatives 2 2% 4 5% 6 3%
Influence of 2 2% 2 3% 4 2% overall frequencies were slightly higher for the first abandoned App
others (56.5/43.5%). As observed with the used Apps, the interview and survey
Popularity 2 2% 4 5% 6 3% data were consistent with poor usefulness and usability, availability of
Design 1 1% 2 3% 3 1% better alternatives and poor content accounting for most of the reasons
Overall 1 1% 5 7% 6 3%
judgement
cited for choice.
Totals 131 64.5% 72 35.5% 203 The most frequent reason for both abandoned applications was
inadequate usefulness (35% of the total comments). Participants either
found no useful features and so deleted the application, or the applica­
towards social communication, networking and keeping in touch, tion was only useful for a particular context of use: e.g. “This App helped
whereas WhatsApp was valued for synchronous communication and me when I was in a sharing accommodation, to identify and split the cost
group chat. Usefulness comments focused on social media Apps features between our roommates. For that purpose, only I was using it. Then I
and how helpful these features were, e.g. “WhatsApp helps me to con­ moved from that room. Now I’m staying in a separate room, so I deleted
nect with most of the friends and family, also the feature of groups in that App” (P-4); or there was limited incentive for use: “Zomato, I just
WhatsApp, helps to connect with and share with multiple people at used to last one year only I can say, because there was a special discount
once” (P-18); “Networking with family and friends, marketing own given” (P4).
businesses social media” (P-2); “Sometimes it shares memory of old Usability comments (22% of total comments) were general negative
posts that we have shared. Those specific features make it interesting” statements that the App was not user friendly, had poor navigation, or
(P-17); “The reason for using it as LinkedIn is mainly professional. Good was too complex, and difficult to operate. Several respondents
connections particularly. You can be able to see the vacancies of com­ mentioned that poor usability influenced them to stop using an appli­
panies, whatever they’re posting on” (P-5); and “Facebook usually. Also cation: “It’s very confusing, not user friendly itself. I downloaded one
I just like to just see like friends, and you know, what are my friends song, and it’s not showing, it is really confusing” (P-24); “I stopped using
doing…… and then LinkedIn I am using it for professional purposes” (P- it because it is inconvenient and not easy to do backward and forward
15). while watching a movie” (P-52).
Usefulness for non-social media Apps was frequently linked to the Alternatives were either specific preferences, “Because I prefer to use
context of use, for example connection to the user’s bank; “Actually IMO, it’s more useful” (P-70), or more general choices, “I use other
Payzapp is mainly linked with HDFC Bank. So it’s fully secured; the user communication applications” (P-7). Several respondents found better
interface, its completely secured” (P-1); or e-government, “Yeah, I mean, Apps in terms of features and/or price: “There are other better games
nowadays whatever you are getting fines, mainly car registration and applications available free” (P-32); “I shifted to booking.com because it
the car fines, you know that you need to do this Dubai police website” (P- provides me with better and pleasant service” (P-21). Change in location
9); and family communication, “And Botim we are obliged to use also affected user preferences, “Yes, now that I’ve moved to India, I no
because I cannot talk to my son studying in India, video call. Only it is longer need D4D, as there are other options here. So, I feel there is no use
the official App. So I subscribe and I mostly use this one for making video for me now” (P-23).
calls” (P-8). Other infrequent reasons for not using an application include con­
Usability was also considered as an important theme and participants tent: “Too much information about people and I’m not interested on it”
frequently commented positively in general terms; “It’s user friendly, (P-45), not having enough time: “I don’t have time to use it” (P-75), and
it’s easily downloadable. It’s freely downloadable App. That’s why we general comments: “I don’t like it” (P-55).
continue to use it” (P-71); “Easy to communicate with the folks and
available in all the smartphones” (P-66). Although only 4% of the 4.3.5. Decision process: interview data
comments mentioned the content, these were positive and specific: The majority of user decisions in selecting and rejecting Apps were
“Facebook has interesting information, news, interactions, all in one fast path (82% in the interview sample of 30 users). Of the 12% slow-
platform” (P-20); “The content of LinkedIn is useful and can be used to
educate myself” (P-51). Few respondents mentioned reliability,
although comments were positive: “There are several map applications, Table 9
but this Google Map is very clear. Like wherever I’m going for the Frequency of comments for Apps abandoned, by category.
driving, it’s getting correct locations and it’s getting updated automat­ Abandoned App 1 Abandoned App 2
ically” (P-41); and similarly, for security, “I chose Zoom because of se­ Category Comment % Comment % Totals %
curity reasons” (P-6). Addiction comments were mainly relating to frequency frequency
games and social media: “It is an addictive game” (P-57). The lack of Usefulness 29 37% 19 32% 48 35%
alternatives and limitations imposed by the context of use appeared as Usability 20 26% 11 18% 31 22%
other infrequent reasons for choosing Apps: “I use Botim as WhatsApp Better 14 18% 18 30% 32 23%
calls were blocked in Dubai so there was no option to make video calls” alternatives
Content 8 10% 8 13% 16 12%
(P-42). Some respondents’ choice was influenced by friends and family:
Lack of time 2 3% 3 5% 5 4%
“Most of my friends have Facebook” (P-34); or work contexts, “Yeah. It’s Overall 5 6% 1 2% 6 4%
very useful. Most of us in company are connected through WhatsApp judgement
group only” (P-2); while others used the App for its popularity “It’s the Total 78 56.5% 60 43.5% 138

7
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

path decisions ten were for abandoned Apps and eight for accepted the wide range of available alternatives and reject those with poor
Apps. Interviewed users reported a mean of 3.43 accepted Apps of which functionality. Few social media Apps appear to have been abandoned,
53.9% were social media as first and second choices, consistent with the possibly reflecting limited choice and the network lock-in cost of
survey data. Fast-path decisions for social media Apps reflected norms changing networks of friends and contacts (Phua et al., 2017; Liu and
for communication: “WhatsApp is now currently getting used by Yang, 2016).
everyone I can say for office and personal use” (P-3); “And most of my Communication Apps, specialised for phone call and video group
colleagues and my friends, they use WhatsApp for communication, meetings, formed an important part of our users App ecosystems. Syn­
exchanging photographs, exchanging documents” (P-6); “Facebook is chronous media-rich communication is an important social need (Ali­
interesting information, news, interactions, all in one platform”(P-14). nejad, 2019; Rozzell et al., 2014; Sutcliffe et al., 2022) as reflected in
For non-social media Apps the more common reasons were the App social media (e.g., WhatsApp); however, other video communication
satisfied the user’s need so no alternatives were sought: “Audible is Apps appear to complement social media. Other choices reflect our re­
convenient, because I used to listen to audio books while driving. And spondents’ online life, with e-shopping, home delivery and payment
fitness App helps me to do training at home and it gives me instructions services, and a mix with leisure-related Apps, for games, health, and
to workout at home.” (P-19); followed by convention or habit: “Payzapp hobbies. Competition in the App ecosystem appears to be more intense
almost 3–4 years I was using it, when I was in India” (P-1); and external in third and fourth choices where the non-social media categories are
constraints: “It is because of the smart Dubai of you know the smart more frequent. Many Apps in communication, leisure and e-commerce
Dubai initiative…. you have to book the taxi through the Careem App. were abandoned for better alternatives, indicating more intense
(P-7). The more common fast-path rejection decisions were change in competition in these categories.
users’ requirements so the App was no longer needed: “It’s my banking Our findings pertaining to the second research question on the influ­
App, which I used to access occasionally. But now I closed the account. ence on user choice demonstrated that usefulness followed by usability
So I stopped using the App” (P-14); cost: “CC camcord is actually a good are the most important determinants of choice, reflected in the high
App, only thing I discontinued is because, after some time it asks to be a frequency of comments, with usability and content UX quantitative
paid App actually. It looks expensive” (P-6); and usability: [Hungama] measures predicting overall experience and satisfaction. Since utility
… it’s very confusing, not user friendly itself. I downloaded one song, was not measured in the survey, content may have been treated as a
and it’s not showing, it’s really confusing.” (P-11). The dominance of surrogate for usefulness by our respondents. However, most usefulness
fast-path decisions is consistent with the survey results that the majority comments reflected overall utility, i.e. the aggregate functionality of the
of respondents did not consider any alternatives for the four chosen Apps. This is consistent with Hart and Sutcliffe’s (2019) report that
Apps. functionality of the device (iPAD) and Apps were the most important
The less frequent slow-path decisions for social media focused on influence on acceptance. Functionality as perceived utility has been a
choice between LinkedIn for professional use and other products and the dominant influence on user choice in many TAM studies (Venkatesh
relative merits of Facebook and WhatsApp. For other products more et al., 2012; Williams et al., 2015) and in UX studies through content,
complex slow-path decisions reflected trade-offs: “I have all Apps that I service quality and pragmatic quality measures (Lavie and Tractinsky,
need in my phone. Only if need to use any new App I go to Playstore. 2004; De Angeli et al., 2006; Hassenzahl, 2004). Usability was the sec­
This App needs some improvement, in terms of Android and Apple there ond most important factor found to influence App choice, as evidenced
is a difference, some need password encryption, card payments, etc.” (P- by both survey and interview results. For chosen Apps usefulness and
18); the context of use: “Udemy. … it’s an online learning App. But I just usability comments were positive, while for abandoned Apps both were
used only when I was in Masters, so after that, I’m not using much on negative; reinforcing the interpretation that usefulness and usability are
those things because it’s my phone has some limited capacity, also the main reasons for choice. This is consistent with Diefenbach and
alternative Apps keep on coming” (P-13); and comparing products: “now Hassenzahl’s (2009) reported importance of pragmatics (PQ: a mix of
I am not using Amazon; also, while purchasing online I used, nowadays, utility and usability) influencing users’ overall judgement of product
I use some other way of shopping, so I deleted and stopped using the ‘goodness’. Poor usability may be discounted when utility is perceived to
App. … For Twitter already I have Facebook, almost same features and be good and matching users’ requirements (Lavie and Tractinsky, 2004;
details are available in Facebook itself. For Amazon, I’m using E-bay… Porat and Tractinsky, 2012); however, we found no evidence of any such
because it’s mainly for buying materials and all those things” (P-26). discounting. Our respondents’ usability comments were favourable for
Apps used, while negative usability was an important reason for aban­
4.4. Discussion doning applications.
In contrast to many UX studies which have reported aesthetics as an
Reviewing our first research question, it appears that social media important influence on user choice (De Angeli et al., 2006; Hartmann
Apps dominate users’ portfolios; however, a more diverse ‘ecosystem’ of et al., 2007; Lindgaard et al., 2011), aesthetics did not predict choice or
different App categories appears after users’ top two choices. While overall experience in our study; furthermore, few aesthetics-related
social media dominate users’ portfolios of mobile phone Apps, as ex­ comments were present in the interviews. Survey data demonstrated
pected from large-scale surveys (Auxier and Anderson 2021; Wikipedia, that content and pleasure, but not aesthetics, with usability were the
2020), social media coexist in a mixed ecology of e-commerce, finance, most important predictors of overall experience. The overall experience
leisure, and communications with a long tail of ‘others’. The market scale included questions on recommending the App and visiting it again,
leaders, Facebook, Instagram, WhatsApp with Snapchat, were the more which are strong indicators of choice and similar to behavioural inten­
popular Apps in our survey, a slightly different mix than found in a USA tion in TAM models (Yousafzai et al., 2007). The minor role of aesthetics
survey (Auxier and Andson 2021) where YouTube and Pinterest are in choice contrasts with several UX studies which have reported aes­
more popular. Other categories reflect a mix of utilitarian needs such as thetics and hedonics as important influences on judgement and product
shopping, information seeking and banking, with leisure interest in choice (Lindgaard et al., 2011; Diefenbach and Hassenzahl, 2004). He­
games, hobbies, and recreation. The dominance of social media is donics also appears as an antecedent variable to behavioural intention in
consistent with other surveys (Wikipedia, 2020; Auxier and Anderson, some TAM studies (Magni et al., 2010) and has been reported with
2021); however, there are few reports of the mobile ‘ecosystem’ mix of pragmatics as a longer-term influence on product acceptance (Kujala
social media and other Apps. The users’ choice order reflects the et al., 2017). One interpretation of this difference may be the contextual
dominance of social media with e-commerce, communications and other influence on UX and user choice (Hartmann et al., 2009), and facilitating
categories being more common third and fourth choices. Higher fre­ conditions on product acceptance (Magni et al., 2010; Williams et al.,
quencies of other Apps were abandoned Apps, suggesting users sample 2015). The Apps chosen by our users were not orientated towards

8
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

aesthetic considerations, apart from games in leisure Apps where a few reporting, which we attempted to counteract by probing questions in the
hedonic comments, e.g. “fun, excitement and stimulation” were interviews. The cognitive process results are tentative since the mea­
recorded. sures to differentiate slow/fast-path processing were indirect, and relied
The factors influencing users’ choice (third research question) were on assumptions about the salience in memory of factors affecting de­
consistent across App types with usefulness and usability dominating. cisions. More direct concurrent protocol studies are necessary to explore
This was also apparent from the minor differences in qualitative data users’ decision making in app choice. Finally, the lack of standardisation
according to App choice order, where social media were more frequent in User experience and TAM measures (Hornbaek and Hertzum, 2017)
as first and second choices, with other App types in third and fourth hinders interpretation of results. While usefulness and usability
positions. This is consistent with utility and ease of use being a constant approximate to perceived utility and ease of use in TAM models (Ven­
across a range of products in TAM studies (Gefen et al., 2003; Maragunić katesh et al., 2012), UX constructs of hedonics and beauty reflect aes­
and Granić, 2015; King and He, 2006). However, the diversity of other thetics, while pragmatics (Hassenzahl, 2004) are an amalgam of utility
reasons reveals several contextual influences on choice, such as change and usability. Indeed the utility of usability (ISO 2018) as a concept has
in job, location or users’ needs. Other influences on users’ choice, such been challenged by Tractinsky (2020) as confounding user experience,
as comments on reliability, security, lack of alternatives, influence of effectiveness, utility and many other constructs.
others, although infrequent also map to ‘facilitating conditions’ recor­
ded in later generations of TAM models (Magni et al., 2010; Williams 5. Conclusions
et al., 2015). Our findings broadly agree Ickin et al. (2017) survey on
app choice across products, which reported App reviews as the main The main contributions of our study are first to confirm the dominant
motivator for downloading, and no longer useful/ poor usability as the role of usefulness (utility) and usability in determining users’ choice,
main reasons for abandoning apps; although their survey found poor indicated by TAM studies (Venkatesh et al., 2012; Marangunić and
reliability and greedy memory as further reasons for rejection, which Granić, 2015) and longitudinal studies of UX (Kujala et al., 2017;
were not frequent comments for our survey. Mendoza et al., 2006). Utility, effectiveness, and efficiency were the
Our findings on the fourth research question indicate that most App most frequently reported attributes found in Weichbroth’s (2020) re­
choices are fast-path decisions rather than more thorough slow-path view of usability in mobile applications. The second contribution is
evaluation of alternatives (Cacioppo et al., 1986; Kahneman, 2011). constrasting with previous studies to illustrate that aesthetics and he­
This may reflect the App market where most products are free; hence donics are minor influences on App choice, although we note that
decisions are driven either by peer pressure and social norms for social pleasure was an important global variable. The third contribution, while
media, or simple matching to users’ needs and, serendipity for other App confirming the dominant role of social media Apps, is to demonstrate
categories. However, more systematic choice was evident when other that competition is more open to fill niches for e-commerce, leisure,
(non-social media) Apps were considered. financing management and communication domains. The fourth
Overall, the influences we found were consistent with Lee et al. contribution is to indicate that the rationale for user App choice reflects
(2012) Theory of Acceptance, apart from personalisation. We did not fast-path / system 1 decision making (Kahneman, 2011) where only a
measure personalisation in the survey, although there was opportunity small number of attributes are evaluated in the decision, such as utility,
for users to comment on this aspect in the ‘reasons’ part of the survey external influences (social media) and convenience. Few alternatives
and interviews. No personalisation comments were recorded. External were explored and even where choice was more diverse because more
influences were present, if not frequent, either in reasons for social Apps are free, a slow-path approach involving more complex reasoning
media use, e.g., networking effect and addiction comments or through (Payne et al., 1993) is infrequent. Further longitudinal studies of App
convenience and constraints on financial services and banks, use of selection and use are needed in a variety of cultural settings to elaborate
video communication Apps by family and friends. Facilitating condi­ App ecosystem models and choice related to user experience and tech­
tions only appear in some of the other reasons for choice, such as nology adoption.
operating system/platform compatibility. Trust and integrity issues
were not cited by our respondents, although these factors may have been Authors’ contribution
implicit in security reasons, especially for financial applications.
The limitations of our study arise from the limited sample of re­ Ons Al-Shamaileh, Outline study design, Design of the survey, con­
spondents to the survey and interviews, who represent users in a ducting survey, collation of results, Design of the interviews, conducting
particular regional setting and a diaspora community, whose choice may interviews, transcription, preliminary analysis, First draft write up of MS
have been influenced by friends and family in their countries of origin. Alistair Sutcliffe, Quantitative data analysis, statistical analysis and
The user interface scales we used were taken from reliable sources, so descriptive statistics. Revised qualitative data analysis, Several revisions
those measures are robust. The questionnaire design included many of MS, improvement of discussion, related work and introduction,
free-format responses and, combined with the interviews, provided Revision in response to reviewers’ comments.
considerable qualitative data to avoid the bias towards a small number
of measures. The mixed methods approach provided frequency and Declaration of Competing Interest
valency data on comment categories that could be combined with the­
matic coding of users’ reasons for choice, thus facilitating triangulation We have no conflicts of interest to declare as a consequence of con­
between quantitative measures and qualitative reasons for users’ choice. ducting this study.
Combination of quantitative user experience measures with qualitative The research was not funded by any external agencies.
interview data helped counter limitations in the selection of measures in
the survey, although our range of measures and relatively small sample Data availability
size limits generalisation of our findings. Our respondents’ nationalities
were diverse, reflecting the ex-patriot and diaspora communities in the Data will be made available on request.
United Arab Emirates and Jordan, and this may have influenced the
diversity of the selected Apps. Further study of the factors influencing
users’ choice to adopt/abandon Apps with users located in different Supplementary materials
countries is needed. The study was a cross-sectional snapshot of users’
choice and their rationale, and relied on retrospective memory of used Supplementary material associated with this article can be found, in
and abandoned Apps. There may have been memory bias in users’ the online version, at doi:10.1016/j.ijhcs.2022.102965.

9
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

References ISO (2018). Ergonomics of human-system interaction — Part 11: usability: definitions
and concepts. https://www.iso.org/obp/ui/#iso:std:iso:9241:-11:ed-2:v1:en</
bib>.
Addonizio, G., 2017. The Privacy Risks Surrounding Consumer Health and Fitness apps,
Johnson-Laird, P.N., 1983. Mental models: Towards a cognitive Science of language,
Associated Wearable devices, and HIPAA’s Limitations. Law School Student
Inference and Consciousness. Cambridge University Press, Cambridge.
Scholarship, p. 861.
Kahneman, D., 2011. Thinking, Fast and Slow. Penguin Books, London.
Adhikari, R., Richards, D., Scott, K., 2014. Security and privacy issues related to the use
King, W.R., He, J., 2006. A meta-analysis of the Technology Acceptance Model, (6),
of mobile health apps. In: Proceedings of the 25th Australasian Conference on
740–755.
Information Systems. Auckland, New Zealand.
Kujala, S., Mugge, R., Miron-Shatz, T., 2017. The role of expectations in service
Ajzen, I., 1991. The Theory of Planned Behaviour. Organ. Behav. Hum. Decis. Process.
evaluation: a longitudinal study of a proximity mobile payment service. Int. J. Hum.
50, 179–211.
Comput. Stud. 98, 51–61.
Alinejad, D., 2019. Careful co-presence: the transnational mediation of emotional
Kujala, S., Roto, V., et al., 2011. UX Curve: a method for evaluating long-term user
intimacy. Social Media & Society 5 (2), 1–19. https://doi.org/10.1177/
experience. Interact. Comput. 473–483.
2056305119854222.
Lavie, T., Tractinsky, N., 2004. Assessing dimensions of perceived visual aesthetics of
Aloudat, A., Al-Shamaileh, O., Michael, K., 2019. Why some people do not use Facebook.
web sites. Int. J. Hum. Comput. Stud. 60 (3), 269–298.
Soc. New. Anal. Min 9, 19. https://doi.org/10.1007/s13278-019-0564-z.
Lee, H.S., Kim, T.G., Choi, J.Y., 2012. A study on the factors affecting smart phone
Al-Shamaileh, O., 2018. I have issues with Facebook: but I will keep using it. IEEE
application acceptance. In: 3rd International Conference on e-Education, e-Business,
Technol. Soc. Mag. 37 (2), 40–45.
e-Management and e-Learning, pp. 27–34.
Al-Shamaileh, O., Sutcliffe, A.G., 2012a. The effect of website interactivity and repeated
Lindgaard, G., Dudek, C., et al., 2011. An exploration of relations between visual appeal,
exposure on user experience. In: Proceedings of the 4th Mexican Conference on
trustworthiness and perceived usability of homepages. ACM Transactions on
Human-Computer Interaction, pp. 1–8.
Computer-Human Interaction 18 (1), 1–30.
Al-Shamaileh, O., Sutcliffe, A.G., 2012b. Investigating a multi-faceted view of user
Liu, D., Yang, C.C., 2016. Media niche of electronic communication channels in
experience. In: Proceedings of the 24th Australian Computer-Human Interaction
friendship: a meta- analysis. Journal of Computer-Mediated Communication 21,
Conference, pp. 9–18.
451–466. https://doi.org/10.1111/jcc4.12175.
Al-Shamaileh, O., Sutcliffe, A.G., De Angeli, A., 2011. The effect of religious identity on
Magni, M., Taylor, M.S., Venkatesh, V., 2010. To play or not to play’: a cross-temporal
user judgment of website quality. In: In Proceedings of INTERACT 2011, IFIP-
investigation using hedonic and instrumental perspectives to explain user intentions
Springer LNCS 6949. Berlin, Heidelberg. Springer, pp. 620–623.
to explore a technology. Int. J. Hum. Comput. Stud. 68 (9), 572–588.
Auxier, B., Anderson, M., 2021. Social media in use 2021. Pew Research Center Report.
Malik, A., Suresh, S., Sharma, S., 2017. Factors influencing consumers’ attitude towards
https://www.pewresearch.org/internet/2021/04/07/social-media-use-in-2021/
adoption and continuous use of mobile applications: a conceptual model. Procedia
(accessed12 March 2022).
Comput Sci 122, 106–113.
Baldwin, J.L., Singh, H., Sittig, D.F., Giardina, T.D., 2017. Patient portals and health
Marangunić, N., Granić, A., 2015. Technology Acceptance Model: a literature review
apps: pitfalls, promises, and what one might learn from the other. Healthcare 5 (3),
from 1986 to 2013. Universal Access in the Information Society 81–95.
81–85.
Maruping, L.M., Bala, H., Venkatesh, V., Brown, S.A., 2017. Going beyond intention:
Bargas-Avila, J., Hornbæk, K., 2011. Old wine in new bottles or novel challenges: a
integrating behavioral expectation into the Unified Theory of Acceptance and Use of
critical analysis of empirical studies of user experience. In: Proceedings of CHI-2011,
Technology. J Assoc Inf Sci Technol 68 (3), 623–637.
pp. 2689–2698.
Mendoza, V., Novick, D.G., Paso, E., 2005. Usability over time. In: Proceedings SIG DOC
Cacioppo, J.T., Petty, R.E., Kao, C.F., Rodriguez, R., 1986. Central and peripheral routes
’05, pp. 151–158.
to persuasion: an individual difference perspective. J. Pers. Soc. Psychol. 51,
Morosan, C., DeFranco, A., 2016. Modeling guests’ intentions to use mobile apps in
1032–1043.
hotels. International Journal of Contemporary Hospitality Management 28 (9),
Creswell, J.W., Plano Clark, V.L., Gutmann, M.L., Hanson, W.E. (2003). Advanced mixed
1968–1991.
methods research designs, in: handbook of mixed methods in social and behavioral
Noh, M.J., Lee, K.T., 2016. An analysis of the relationship between quality and user
research, pp 209–240.
acceptance in smartphone apps. Information Systems and e-Business Management 2,
Cyr, D., Head, M., Ivanov, A., 2009. Perceived interactivity leading to e-loyalty:
273–291.
development of a model for cognitive-affective user responses. Int. J. Hum. Comput.
O’Brien, H.L., 2010. The influence of hedonic and utilitarian motivations on user
Stud. 850–869.
engagement: the case of online shopping experiences. Interact. Comput. 22,
Dai-yon, C., Hyun-jung, K., 2008. Analysis of trust in internet and mobile commerce
344–352.
adoption. Korea Internet e-Commerce Research 8 (2), 151–182.
Pan, A., Zhao, F., 2018. User acceptance factors for mHealth. In: International
Davis, L.D., Bagozzi, R.P., Warshaw, P.R., 1989. User acceptance of computer
Conference on Human-Computer Interaction Berlin. Springer, pp. 173–184.
technology: a comparison of two theoretical models. Manage Sci 35 (8), 982–1002.
Payne, J.W., Bettman, J.R., Johnson, E.J., 1993. The Adaptive Decision Maker.
De Angeli, A., Sutcliffe, A.G., Hartmann, J., 2006. Interaction, usability and aesthetics:
Cambridge University Press, Cambridge.
what influences users’ preferences?. In: Proceedings of the 6th Conference on
Phua, J., Jin, S.V., Kim, J.J., 2017. Uses and gratifications of social networking sites for
Designing Interactive Systems. New York. ACM Press, pp. 271–280.
bridging and bonding social capital: a comparison of Facebook, Twitter, Instagram,
Diefenbach, S., Hassenzahl, M., 2009. The Beauty Dilemma: beauty is valued but
and Snapchat. Comput Human Behav 72, 115–122.
discounted in product choice. In: Proceedings of SIGCHI Conference on Human
Porat, T., Tractinsky, N., 2012. It’s a pleasure buying here: the effects of web-store design
Factors in Computing Systems, pp. 1419–1426.
on consumers’ emotions and attitudes. Hum-Comput. Interact. 235–276.
Ding, Y., Chai, K.H., 2015. Emotions and continued usage of mobile applications.
Rozzell, B., Piercy, C., Carr, C.T., et al., 2014. Notification pending: online social support
Industrial Management & Data Systems 833–852.
from close and nonclose relational ties via Facebook. Comput Human Behav 38,
Dong-Wook, J., Sang-Hoon, K., Chang-Kyu, K., 2008. The influence of use purpose in
272–280. https://doi.org/10.1016/j.chb.2014.06.006.
mobile internet service on loyalty. e-Business Studies 9 (3), 129–157.
Sang-Hyeon, K., Sang-Hyun, O., 2002. The effects of internet shopping mall
Gefen, D., Karahanna, E., Straub, D.W., 2003. Trust and TAM in online shopping: an
characteristics on satisfaction, trust, and loyalty. Asia Pacific Journal of Small
integrated model. MIS Quarterly 27 (1), 51–90.
Business 24 (2), 237–271.
Gye-Soo, K., 2002. A study of service quality strategy on the internet portal site. Korean
Sloman, S.A., 1996. The empirical case for two systems of reasoning. Psychol. Bull. 119,
Management Review 31 (1), 191–209.
3–22.
Harris, M.A., Brookshire, R., Chin, A.G., 2016. Identifying factors influencing consumers’
Sutcliffe, A.G., 2009. Designing for user engagement: aesthetic and attractive user
intent to install mobile applications. Int. J. Inf. Manage. 36 (3), 441–450.
interfaces. In: Carroll, J.M. (Ed.), Synthesis Lectures on Human Centred Informatics.
Hart, J., Sutcliffe, A.G., 2013. The influence of interactivity and user types in user
Morgan Claypool, San Rafael CA.
experience. In: Proceedings of CHI 2013. ACM Press /Digital Library.
Sutcliffe, A.G., El Jarn, H., Dunbar, R.I.M., 2022. Investigating the use of social media in
Hart, J., Sutcliffe, A.G., 2019. Is it all about the apps or the device? User experience and
supporting intimate social relationships. Behav Inf Technol. https://doi.org/
technology acceptance among iPad users. Int. J. Hum. Comput. Stud. 130, 93–112.
10.1080/0144929X.2021.2023634 published online.
Hartmann, J., Sutcliffe, A.G., De Angeli, A., 2007. Investigating attractiveness in web
Teo, H.-.H., Oh, L.-.B., Liu, C., Wei, K.-.K., 2003. An empirical study of the effects of
user interfaces. In: Proceedings of the SIGCHI Conference on Human Factors in
interactivity on web user attitude. Int. J. Hum. Comput. Stud. 2003, 281–305.
Computing Systems. New York. ACM Press, pp. 387–396.
Tractinsky, N., 2020. The Usability Construct: a concern for both theory and practice.
Hartmann, J., Sutcliffe, A.G., De Angeli, A., 2009. Towards a theory of user judgment of
Hum-Comput. Interact. 35 (4), 338–353.
aesthetics and user interface quality. ACM Transactions on Computer-Human
Tunney, P., 2018. 5 apps that hospitality professionals should have on their phone. https
Interaction 15 (4), 15–30.
://www.hosco.com/en/advice/article/apps-for-hospitality-professionals (accessed:
Hassenzahl, M., 2004. The interplay of beauty, goodness and usability in interactive
20 March 2020).
products. Hum-Comput. Interact. 19, 319–349.
Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of
Holton, J.A., 2007. The coding process and its challenges. The Sage handbook of
information technology: toward a unified view. MIS Quarterly 27 (3), 425–478.
grounded theory 3, 265–289.
Venkatesh, V., Thong, J.Y., Xu, X., 2012. Consumer acceptance and use of information
Hornbæk, K., Hertzum, M., 2017. Technology acceptance and user experience: a review
technology: extending the Unified Theory of Acceptance and Use of Technology. MIS
of the experiential component in HCI. ACM Transactions on Computer-Human
Quarterly, 157–178157-178.
Interaction 24 (5), 1–30.
Wang, Y.S., Lin, H.H., Luarn, P., 2006. Predicting consumer intention to use mobile
Ickin, S., Petersen, K., Gonzalez-Huerta, J., 2017. Why do users install and delete Apps? A
service. Information Systems Journal 16 (2), 157–179.
survey study, in: international Conference of Software Business (pp. 186-191).
Wei, P.S., Lee, S.Y., Lu, H.P., Tzou, J.C., Weng, C.I., 2015. Why do people abandon
Springer, Cham.
mobile social games? Using Candy Crush Saga as an example. International Journal
Islam, R., Islam, R., Mazumder, T., 2010. Mobile application and its global impact.
of Social, Education, Economics and Management Engineering 9 (1), 13–18.
International Journal of Engineering & Technology 10 (6), 72–78.

10
O. Al-Shamaileh and A. Sutcliffe International Journal of Human - Computer Studies 170 (2023) 102965

Weichbroth, P., 2020. Usability of mobile applications: a systematic literature study. Williams, M.D., Rana, N.P., Dwivedi, Y.K., 2015. The Unified Theory of Acceptance and
IEEE Access 8, 55563–55577. Use of Technology (UTAUT): a literature review. Journal of Enterprise Information
Wikipedia (2020), List of most popular smartphone apps. https://en.wikipedia.org/wiki/ Management 28, 443–488.
List_of_most_popular_smartphone_apps. Yousafzai, S.Y., Foxall, G.R., Pallister, J.G., 2007. Technology acceptance: a meta
analysis of the TAM: part 1. Journal of Modelling in Management 2, 251–280.

11