Mobile Instant Messaging for Customer Service Interaction

Preparation of a Model-based Approach Exploring Behavioral Intention

Hannes Schimmele, Stephan Schl

ogl and Aleksander Groth

Interaction Lab, Department Management, Communication & IT,

MCI Management Center Innsbruck, Austria

Keywords:

Mobile Instant Messaging, Customer Service Interaction, Theory of Planned Behavior, Technology Readiness

Index.

Abstract:

Mobile Instant Messaging (IM) has become a main tool for communicating with friends and family. Recent

efforts are now exploring IM use in business-to-customer communication. To this end, particularly customer

service settings are given a high priority. A clear understanding of the relevant customer perspective is, how-

ever, still missing. Consequently, this paper reports on a study evaluating people’s behavioral intention of

adopting mobile IM for customer service interaction. We used an integrated research model based on Ajzen’s

Theory of Planned Behavior (TPB) and Parasuraman’s Technology Readiness Index. A total of 154 ques-

tionnaires were analyzed. Results indicate only small effects of technology-related personality traits on the

cognitive dimensions of the TPB. Concerning TPB-internal relations, Attitude and Subjective Norm were found

to have signiﬁcant inﬂuences on Behavioral Intention. In summary, however, the study indicates that people’s

intention to adopt mobile IM for customer service interaction is only slightly positive.

1 INTRODUCTION

On February 1

2016 WhatsApp announced their ef-

forts in exploring alternative application areas for mo-

bile Instant Messaging (IM)

. In particular, they high-

lighted their goal to offer users a new way of commu-

nicating with businesses and organizations. In other

words, individuals and businesses should be able to

exchange important information in a modern, more

convenient and less complicated manner. WhatsApp’s

intentions may be seen as an attempt to open up a new

chapter in the ﬁeld of customer-business interaction.

Since then, ﬁrst prototypical implementations of us-

ing IM for customer communication have been put in

place. Yet, so far, little is know about the customers’

perspective and consequent attitude towards these ser-

vices. The goal of the work presented in this paper is

thus to shed some light on the potential use of IM in

customer service settings. Employing a model-based

research approach we aim to contribute to a better un-

derstanding of the users’ underlying motives and their

possible reluctance factors, which play an essential

role in the adoption and/or refusal of said technology

for Customer Service Interaction (CSI).

https://blog.whatsapp.com/ [retrieved: May 15 2016]

2 THEORETICAL BACKGROUND

With the ongoing propagation of smartphones the

availability of mobile IM is constantly increasing. In

contrast to the use of traditional SMS-based text mes-

saging, the use of mobile IM is not limited to a given

number of characters and is usually free of charge

– provided the user has access to an internet con-

nection via data plan or a wireless network connec-

tion (Church and de Oliveira, 2013). Furthermore,

IM includes a number of positive characteristics, like

different types of media content such as for exam-

ple videos, photos, documents, contacts, audio ﬁles,

the user’s current location as well as his/her voice

mails (Rennecker and Godwin, 2003). In addition,

most mobile IM tools feature information about a

communication partner’s status, such as him/her be-

ing online/ofﬂine, busy, typing, idle or free, which

mimics important characteristics of a real-time physi-

cal person-to-person conversation (Deng et al., 2010).

Moreover, mobile IM applications provide numer-

ous emoticons to convey emotions (Lancaster et al.,

2007). Thus, given the variety of different transfer-

able media types combined with the ‘emotional’ ef-

fect of emoticons, studies have already shown that

users perceive (mobile) IM as more personalized than

other means of digital communication such as for ex-

Schimmele, H., Schlögl, S. and Groth, A.

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention.

DOI: 10.5220/0006490100390049

In Proceedings of the International Conference on Computer-Human Interaction Research and Applications (CHIRA 2017), pages 39-49

ISBN: 978-989-758-267-7

ample email (Church and de Oliveira, 2013; Huang

et al., 2008).

However, perceived (technological) advantages

alone do usually not sufﬁce to ensure the adoption

of technology in a given application area. Motiva-

tional, behavioral as well as social factors also play

their part. To this end, technology acceptance and/or

adoption, being “conceptualized as an outcome vari-

able in a psychological process that users go through

in making decisions about technology” (Dillon and

Morris, 1996), is perceived as a crucial determinant

inﬂuencing the success or failure of a technology

brought to market. Consequently, its underlying the-

oretical models have been subject to many previous

studies (Ghyas et al., 2012; Venkatesh et al., 2003).

2.1 Theory of Reasoned Action

The Theory of Reasoned Action (TRA) developed by

Fishbein & Ajzen (Fishbein and Ajzen, 1975) was

one of the ﬁrst widely applied theoretical models to

explain human behavior. Until today, it is one of

the most applied theories and has received substantial

empirical support (Dabholkar, 1994; P

uschel et al.,

2010). The model is based on the assumption that a

person’s actual behavior is inﬂuenced by his/her in-

tention to execute said behavior. Intention to perform

behavior X can thus be characterized as an indicator

“of how hard people are willing to try, of how much

of an effort they are planning to exert, in order to per-

form the behavior” (Ajzen, 1991). Consequently, a

strong behavioral intention should also increase the

likelihood of actually performing the respective be-

havior (Ajzen, 1985; Ajzen, 1991). Intention to per-

form behavior X is further mutually determined by

the individual’s Attitude towards said behavior and

his/her Subjective Norm concerning it (Fishbein and

Ajzen, 1975). These two dimensions reﬂect on the

one hand the personal impact and on the other hand

the social inﬂuences on the intention to perform a be-

havior (Ajzen, 1985). That is, Attitude towards be-

havior X “refers to the degree to which a person has a

favorable or unfavorable evaluation or appraisal of the

behavior in question” (Ajzen, 1991). Subjective Norm

on the other hand is deﬁned as “the person’s per-

ception of the social pressure put on him/her to per-

form or not perform the behavior in question” (Ajzen,

1985). At a deeper level, one may thus argue that

people’s salient beliefs determine their Attitude and

Subjective Norm concerning behavior X. In summary

the TRA therefore proposes that, individuals intend

to carry out a behavior if they, for themselves, pos-

itively evaluate the behavior in question and if they

think that people of their social environment would

support them performing this behavior (Ajzen, 1985;

Ajzen, 1991).

2.2 Technology Acceptance Model

While TRA was designed to explain human behav-

ior and its antecedents in general, it also served as

the foundation for other models aiming to explain be-

havior in more speciﬁc contexts. In case of technol-

ogy, such particularly concerns the Technology Ac-

ceptance Model (TAM) developed by Davis (Davis,

1989). TAM focuses on explaining user acceptance

of technology in the workplace by replacing the TRA

element Subjective Norm with the dimensions Per-

ceived Usefulness and Perceived Ease of Use (Davis

et al., 1989; Davis, 1989). Perceived Usefulness is

deﬁned as “the degree to which a person believes

that using a particular system would enhance his/her

job performance” (Davis, 1989). Here the term use-

ful is based on the assumption that a technology

can be used advantageously (Davis, 1989). Con-

sequently, a system which is characterized by high

Perceived Usefulness “is one for which a user be-

lieves in the existence of a positive use-performance

relationship” (Davis, 1989). As such it is directly

connected to both Attitude and Behavioral Intention

and grounded in the argument that, in a work set-

ting, the intention to use a system may also stem

directly from its assumed impact on the overall job

performance, independent of the overall attitude to-

wards using the system (Davis, 1989). Perceived Ease

of Use, on the other hand, describes “the degree to

which a person believes that using a particular system

would be free of effort” (Davis, 1989), where the term

ease implies the “freedom from difﬁculty or great ef-

fort” (Davis, 1989). In contrast to TRA, TAM has

been widely applied in studies on mobile services us-

age; e.g. (Basg

oze, 2015; Glass and Li, 2010; Godoe

and Johansen, 2012; Gombachika and Khangamwa,

2013; Guhr et al., 2013; Jin, 2014; Li et al., 2005; Lu

et al., 2009).

2.3 Theory of Planned Behavior

Next to TAM also the Theory of Planned Behavior

(TPB) strongly builds on the basis of the TRA. Instead

of replacing the component Subjective Norm, how-

ever, it extends the model by a third dimension. Origi-

nally TRA was created to explain volitional behavior,

i.e. behavior where an individual voluntarily chooses

to perform an action (Ajzen, 1985). This has been a

major limitation to the TRA model, as it is not appli-

cable to behavior which is not under full perceived or

actual control (Ajzen, 1985; Ajzen, 1991). To over-

CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications

come this limitation, the TPB thus incorporates the

dimension Perceived Behavioral Control, which is de-

ﬁned as the “perceived ease or difﬁculty of perform-

ing the behavior and it is assumed to reﬂect past ex-

perience as well as anticipated impediments and ob-

stacles” (Ajzen, 1991). Except for the amendment of

Perceived Behavioral Control and its direct impact on

the actual behavior, the TPB construct is, however,

identical to the TRA. In other words, the TPB essen-

tially assumes “the more favorable the Attitude and

Subjective Norm with respect to a behavior, and the

greater the Perceived Behavioral Control, the stronger

should be an individual’s intention to perform the be-

havior under consideration” (Ajzen, 1991).

3 ACCEPTANCE OF MOBILE

INSTANT MESSAGING

Lu et al. (Lu et al., 2009) as well as Glass & Li (Glass

and Li, 2010) aimed at generally explaining mo-

bile IM acceptance with an integrated TAM. Shih &

Fan (Shih and Fan, 2013), on the other hand, explored

IM in a more work-related context. In their case, the

TPB was extended by a construct measuring technol-

ogy readiness. The results showed that an optimistic

view of technology also fosters a positive attitude to-

wards applying IM for a professional purpose (Shih

and Fan, 2013). Building on these results, both inter-

net and mobile banking have been explored through

applying the TRA (Nor, 2015) as well as a Decom-

posed version of the Theory of Planned Behavior

(DTPB) (P

uschel et al., 2010; Shih and Fang, 2004).

Whereas the DTPB studies applied the original model

almost in its pure form, Nor (Nor, 2015) integrated a

technology readiness construct into the TRA. Similar

study settings comparing the TRA and the DTPB are

found in the educational sector. For example, Ghyas

et al. (Ghyas et al., 2012) evaluated the acceptance

of e-readers using DTPT whereas Jin (Jin, 2014) in-

vestigated the technology through an integrated and

adapted TAM. E-learning, in particular, was subject

to TAM studies with technology readiness dimension

to them (Yen and Chen, 2010).

Another relevant comparison may be found in

acceptance studies focusing on mobile commerce.

Basg

oze, for example investigated mobile shop-

ping using TAM and a technology readiness exten-

sion (Basg

oze, 2015). Similar approaches were taken

to explore the acceptance of mobile payment (Guhr

et al., 2013; Shin and Lee, 2014). On a more general

level, the inﬂuence of gender for the acceptance of in-

formation and communication technologies was eval-

uated by Gombachika and colleagues (Gombachika

and Khangamwa, 2013). The study strongly focused

on technology readiness and again measured ICT ac-

ceptance through different TAM constructs. Finally,

we have also seen the acceptance of e-service and

self-service technologies being studied by integrating

technology readiness into the TPB (Chen and Chen,

2008; Chen and Li, 2010).

In summary, however, we see two common char-

acteristics inherent with previous studies. First, very

few studies employ the original models such as a pure

version of TRA, TPB or TAM. Instead, researchers

keep integrating other behavioral theories or seem-

ingly foreign models so as to potentially improve their

research results. Second, the extension of original

models often concerns an additional construct that

aims at measuring people’s technology readiness - an

approach which is further discussed by the following

section.

3.1 Technology Readiness

With technology increasingly becoming the main re-

quirement for services and products being transported

(not to say offered) to customers, technology readi-

ness has evolved as an important facet of behavioral

research. It allows to draw conclusions about people’s

emotional reaction to newly introduced technologies.

Mick & Fournier (Mick and Fournier, 1998) found

eight paradoxes which users are faced with when be-

ing confronted with a new technology. This shows

certain similarities to the underlying assumptions put

forward by the TPB. Dabholkar (Dabholkar, 1996),

for example, highlights that people have different

feelings about technology which inﬂuence their indi-

vidual behavioral intention to use it. Parasuraman, on

the other hand, proposes that “a combination of posi-

tive and negative feelings about technology underlies

the domain of technology readiness” (Parasuraman,

2000). Furthermore, he states that “whereas posi-

tive feelings propel people towards new technologies,

the negative feelings may hold them back” (Parasur-

aman, 2000). Building on this theory Parasuraman

developed the Technology Readiness Index (TRI) as

a construct which “refers to people’s propensity to

embrace and use new technologies for accomplish-

ing goals in home life and at work” (Parasuraman,

2000). The TRI theory in its version from 2000 in-

corporates 36 items which can be assigned to the

four categories Optimism, Innovativeness, Discom-

fort and Insecurity. They reﬂect the positive (Op-

timism and Innovativeness) and negative (Discom-

fort and Insecurity) perceptions of technology accord-

ing to Mick & Fournier (Mick and Fournier, 1998).

Accordingly, they can be interpreted as an individ-

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention

ual’s technology-related personality traits, which in-

ﬂuence his/her technology readiness. In other words,

Optimism and Innovativeness are deﬁned as drivers

of technology readiness whilst Discomfort and In-

security inhibit technology readiness (Parasuraman,

2000).

In 2014, Parasuraman & Colby updated the ini-

tial TRI version by introducing TRI 2.0 (Parasura-

man and Colby, 2014). Like the original model (TRI

1.0), TRI 2.0 consists of the four personality traits

Optimism, Innovativeness, Discomfort and Insecurity,

with the ﬁrst two being motivators and the latter two

being inhibitors of technology readiness. Also the

terms and deﬁnitions of TRI 1.0 were transferred to

TRI 2.0. However, the items were re-worded, so as

to be less speciﬁc to one particular technology, and

condensed (Parasuraman and Colby, 2014). The ﬁ-

nal TRI 2.0, which should be applicable to a wider

ﬁeld of technology explorations, now consists of four

questions per trait, leading to a total of 16 items to

be put on a questionnaire. According to Parasuraman

& Colby (Parasuraman and Colby, 2014), the newly

integrated TRI 2.0 may be particularity used to in-

vestigate a “moderating variable in studies involving

multivariate frameworks”. Following we conclude

the necessary background analysis by discussing sev-

eral approaches that have integrated this type of tech-

nology readiness into technology acceptance studies,

leading over to the research agenda perused by our

work.

3.2 Integrating TRI with TRA, TAM

and TPB

Ever since Ajzen proposed that besides Attitude also

other personality traits have a major inﬂuence on

human behavior (Ajzen, 1991; Ajzen, 2005), re-

searchers have worked on integrating technology

readiness into behavioral models. As for an integra-

tion with technology acceptance, particularly TAM,

we have seen a number of application scenarios;

e.g. (Chen and Chen, 2008; Chen and Li, 2010; Shih

and Fan, 2013). Hence, while for TAM fundamen-

tal work such as the Technology Readiness and Ac-

ceptance Model (TRAM) by Lin et al. (Lin et al.,

2007) exists, the integration of TRI with TPB is less

researched. Chen & Chen (Chen and Chen, 2008)

were among the ﬁrst to examine the effect of technol-

ogy readiness on consumer behavior and may conse-

quently be regarded as the pioneers in integrating TRI

(in version 1.0) with TPB. Doing so, they examined

the inﬂuence of personality traits on the antecedents

of Behavioral Intention, ﬁnding that personality traits

do indeed play an important role in the information

technology adoption (Chen and Chen, 2008). Another

approach to integrate TRI with TPB was undertaken

when examining the adoption of e-services (Chen and

Li, 2010). It was theorized that technology readiness

can be interpreted as a belief and would therefore pre-

cede Attitude, Subjective Norm and Perceived Behav-

ioral Control, as explained by the TPB (Ajzen, 1985;

Ajzen, 1991; Chen and Li, 2010). Finally, a third

study, set in the context of IM adoption by Taiwanese

travel agency workers, concentrated on the inﬂuence

technology-related personality traits have on the TPB

element Attitude alone (Shih and Fan, 2013).

In summary, we have seen that in the past primar-

ily two conceptions integrating technology readiness

with acceptance have emerged. On the one hand, re-

searchers have studied how technology readiness as a

whole inﬂuences the antecedents of TAM and TPB.

On the other hand, studies were aimed at a diverse set

of personality traits and how they inﬂuence technol-

ogy adoption, particularly focusing on relating single

TRI components to TAM/TPB elements; e.g. (Chen

and Chen, 2008; Godoe and Johansen, 2012; Wal-

czuch et al., 2007). Our study aims to add to this

speciﬁc research domain by investigating the accept-

ability/adoption of mobile IM in customer service set-

tings. The applied research model, which builds on

the work of Chen & Chen (Chen and Chen, 2008),

but replaces TRI 1.0 by its successor TRI 2.0, is dis-

cussed below.

4 RESEARCH MODEL AND

HYPOTHESIS

The discussion above highlights that TRA, TPB as

well as TAM are viable constructs capable of evalu-

ating technology adoption/acceptance. Also, all three

have been integrated with the TRI 1.0 measurement

scale. However, with respect to the exploration of be-

havioral intention, TPB seems to be a better ﬁt than

TRA and TAM (and their respective derivatives). One

reason for this is because TPB provides “the possibil-

ity of making further distinctions among additional

kinds of beliefs and related dispositions” (Ajzen,

1991). In addition, personality traits do play an im-

portant role in human behavior (Ajzen, 1985; Ajzen,

2005). Most previous studies using a TRI-TPB in-

tegration are, however, grounded in the initial model

proposed by Chen & Chen (Chen and Chen, 2008).

The main reason why this initial model is preferred

over other TRI-TPB integrated approaches, such as

the one by Chen & Li (Chen and Li, 2010), lies in

its focus on each single technology readiness dimen-

sion which does not aggregate all of the items into

CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications

one measure. Empirical results conﬁrmed the valid-

ity of this approach as it was found that not all TRI

components have an equal inﬂuence on the TPB el-

ements Attitude, Subjective Norm and Perceived Be-

havioral Control (Chen and Chen, 2008). Shih &

Fan’s results (Shih and Fan, 2013) support this ar-

gumentation as well, and Godoe & Johansen (Godoe

and Johansen, 2012), who employed a TRI-TAM con-

struct also recommend the decomposition of technol-

ogy readiness into its underlying technology-related

personality traits instead of relating them to the TPB

elements as one single measure.

The presented study aims to add to the body

of knowledge in this ﬁeld by integrating TPB with

the updated version of TRI, i.e. TRI 2.0. As per

the TPB framework, behavior is dependent on In-

tention, and Intention in turn is explained by its an-

tecedents Attitude towards the behavior, Subjective

Norm and Perceived Behavioral Control, we may the-

orize that technology-related personality traits can di-

rectly be related to the TPB elements. Here Opti-

mism and Innovativeness are characterized as drivers,

and Insecurity and Discomfort as inhibitors of tech-

nology readiness (Parasuraman and Colby, 2014;

Parasuraman, 2000). People who hold an optimistic

and innovative position towards technology are there-

fore expected to also have a positive attitude towards

the adoption of mobile IM for Customer Service In-

teraction (CSI). In contrast, people who feel insecure

and dis-comfortable with respect to technology are

predicted to have a rather negative attitude towards

the adoption of mobile IM for CSI. Consequently one

may hypothesize that:

• H1(a): Optimism is positively related to Attitude.

• H1(b): Innovativeness is positively related to Atti-

tude.

• H1(c): Discomfort is negatively related to Atti-

tude.

• H1(d): Insecurity is negatively related to Attitude.

It is further expected that optimistic and innova-

tive character traits make people expect support from

their social environment concerning the adoption of

mobile IM for CSI. Yet, feelings of insecurity and dis-

comfort towards technology make people think that

their social environment would rather constrain such

ambitions. Therefore one may further hypothesize

that:

• H2(a): Optimism is positively related to Subjective

Norm.

• H2(b): Innovativeness is positively related to Sub-

jective Norm.

• H2(c): Discomfort is negatively related to Subjec-

tive Norm.

• H2(d): Insecurity is negatively related to Subjec-

tive Norm.

Similar arguments should account for peo-

ple’s Perceived Behavioral Control. That is, people

who have an optimistic and innovative view of tech-

nology are expected to be conﬁdent in operating mo-

bile IM for CSI. In contrast, feelings of insecurity

and discomfort towards technology should result in

a low conﬁdence concerning these operations. Con-

sequently one may argue that:

• H3(a): Optimism is positively related to Perceived

Behavioral Control.

• H3(b): Innovativeness is positively related to Per-

ceived Behavioral Control.

• H3(c): Discomfort is negatively related to Per-

ceived Behavioral Control.

• H3(d): Insecurity is negatively related to Per-

ceived Behavioral Control.

Finally, according to the original TPB model, Atti-

tude, Subjective Norm and Perceived Behavioral Con-

trol inﬂuence Behavioral Intention (Ajzen, 1985;

Ajzen, 1991). Therefore, one may conclude that a

positive attitude towards mobile IM for CSI results

in a stronger behavioral intention to actually employ

it. Accordingly, it is also expected that people’s be-

liefs in a supporting social environment would trigger

a stronger behavioral intention. The same accounts

for people’s conﬁdence in their ability to use mobile

IM for CSI, wherefore it may be hypothesized that:

• H4(a): Attitude is positively related to Behavioral

Intention.

• H4(b): Subjective Norm is positively related

to Behavioral Intention.

• H4(c): Perceived Behavioral Control is positively

related to Behavioral Intention.

4.1 Questionnaire Design

Our study design tackling the above hypotheses is

strongly model-based, building on prior research con-

ducted in technology adoption of e-services. The

work by Chen & Chen (Chen and Chen, 2008) served

as the main guideline, however, we focused speciﬁ-

cally on the use of mobile IM for CSI and further re-

placed the original TRI model with its successor TRI

2.0.

Following the example of previous surveys (Chen

and Chen, 2008; Dabholkar, 1996), we designed a

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention

Table 1: Questionnaire survey completed by 154 respondents.

Name Item

ATT1 Using WhatsApp for customer service interaction is a good idea.

ATT2 I like the idea of using WhatsApp for customer service interaction.

ATT3 Using WhatsApp for customer service interaction is a wise idea.

ATT4 Using WhatsApp for customer service interaction would be pleasant.

SN1 People who are important to me (e.g. friends, family) would think that I should use WhatsApp

for customer service interaction.

SN2 People who are important to me (e.g. friends, family) would think that using WhatsApp for

customer service interaction is a good idea.

SN3 People who inﬂuence me (e.g. bloggers, role models) would think that I should use WhatsApp

for customer service interaction.

SN4 People who inﬂuence me (e.g. bloggers, role models) would think that using WhatsApp for

customer service interaction is a good idea.

PBC1 I would be able to use WhatsApp for customer service interaction.

PBC2 I have the resources (e.g. Smartphone with WhatsApp installed, mobile data plan) to use What-

sApp.

PBC3 I have the knowledge to operate WhatsApp for customer service interaction.

PBC4 I have the ability how to operate WhatsApp.

BI1 I would plan to use WhatsApp for customer service interaction.

BI2 I would intend to use WhatsApp for customer service interaction within the next three months.

BI3 I would recommend WhatsApp for customer service interaction to others.

OPT1 New technologies contribute to a better quality of life.

OPT2 Technology gives me more freedom of mobility.

OPT3 Technology gives people more control over their daily lives.

OPT4 Technology makes me more productive in my personal life.

INN1 Other people come to me for advice on new technologies.

INN2 In general, I am among the ﬁrst in my circle of friends to acquire new technology when it

appears.

INN3 I can usually ﬁgure out new high-tech products and services without help from others.

INN4 I keep up with the latest technological developments in my areas of interest.

DIS1 When I get technical support from a provider of a high-tech product or service, I sometimes feel

as if I am being taken advantage of by someone who knows more than I do.

DIS2 Technical support lines are not helpful because they don’t explain things in terms I understand.

DIS3 Sometimes, I think that technology systems are not designed for use by ordinary people.

DIS4 There is no such thing as a manual for a high-tech product or service that’s written in plain

language.

INS1 People are too dependent on technology to do things for them.

INS2 Too much technology distracts people to a point that is harmful.

INS3 Technology lowers the quality of relationships by reducing personal interaction.

INS4 I do not feel conﬁdent doing business with a place that can only be reached online.

GENDER What’s your gender?

AGE How old are you (in years)?

COUNTRY My home country is:

WAEXP I have used WhatsApp for:

SEXP I have owned a smartphone (or several smartphones) for:

WADAILY The time I spend every day in WhatsApp is:

CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications

questionnaire study which included a short introduc-

tion to the purpose of the research, a potential usage

scenario which let respondents imagine a concrete ap-

plication of mobile IM for CSI, and the actual ques-

tion constructs, covering TPB, TRI 2.0, and demo-

graphic items.

With respect to the TPB, items developed by Tay-

lor & Todd (Taylor and Todd, 1995) and validated

by several follow-up studies (Ghyas et al., 2012; Lu

et al., 2009; P

uschel et al., 2010; Shih and Fang,

2004; Teo and Pok, 2003) were adjusted to the con-

text of mobile IM for CSI. The term mobile IM was

thereby substituted by the (today) more familiar term

WhatsApp. Consequently, the ﬁrst part of the actual

questionnaire focusing on TPB included four Attitude

items developed according to Taylor & Todd (Taylor

and Todd, 1995). Additional four Subjective Norm

items were taken from work by Shih & Fang (Shih

and Fang, 2004) who also provide the four items

for Perceived Behavioral Control as well as two items

for Behavioral Intention. The third (and ﬁnal) Behav-

ioral Intention item, which particularly incorporated

the recommendations of WhatsApp for CSI to others,

was inspired by Lu and colleagues (Lu et al., 2009).

The second part of the questionnaire incorporated

the 16 TRI 2.0 measurement items (Parasuraman and

Colby, 2014). To this end no adaptation was re-

quired as TRI measures are technology and context

independent. All variables used a ﬁve-point Likert

scale differential ranging from “strongly disagree” to

“strongly agree”, complying with recommendations

for both TPB as well as TRI 2.0 (Cariﬁo and Perla,

2007; Parasuraman and Colby, 2014). The ﬁnal part

of the questionnaire comprised six demographic vari-

ables including gender, age, WhatsApp experience,

smartphone experience and daily WhatsApp usage,

as well as one question concerning the participant’s

country of residence. The complete survey (cf. Ta-

ble 1) was distributed online using a convenience

sample of WhatsApp users in Central Europe, particu-

larly focusing on Germany, Austria and Switzerland,

and led to 154 completed responses within 29 days.

4.2 Questionnaire Reliability and

Validity

The returned data shows an equal gender distribution

(i.e. 72 male and 78 female participants) with the

respondents’ age ranging from 18 to 62 years (note:

66.7 percent were between 21 and 26 years old). The

majority (87.3%) has owned a smartphone for at least

three years. As for the WhatsApp usage, 37.3% re-

ported to use the app between 31 and 60 minutes

per day, while 38.7% would use it shorter and 24.0%

longer. Calculating Cronbach’s a the questionnaire

items investigating Attitude (ATT), Subjective Norm

(SN), Perceived Behavioral Control (PBC), Behav-

ioral Intention (BI) as well as Innovativeness (INN)

had coefﬁcients greater than 0.7, showing high relia-

bility. The reliability of the items investigating Opti-

mism (OPT), Discomfort (DIS) and Insecurity (INS)

had coefﬁcients ranging from 0.6 to 0.7, deeming

them slightly less reliable.

Next, similar to previous studies (Gombachika

and Khangamwa, 2013; Shih and Fan, 2013) we

used an Exploratory Factor Analysis (EFA) to en-

sure that all the TRI 2.0 and TPB items would load

to their associated construct. This process started

with a variable selection and then continued with

the consequent factor extraction. Here, the Kaiser-

Meyer-Olkin (KMO) criterion measures whether a set

of variables qualiﬁes for an EFA (Kaiser and Rice,

1974). It is based on the anti-image covariance ma-

trix and lies between 0 and 1, where a value lower

than 0.5 indicates that a given variable set is not suit-

able. Applying the KMO criterion to the model-based

variable constructs used in the above survey (i.e. the

TRI 2.0 and the TPB questionnaire items) produced a

value of 0.848, classifying our variable set as “merito-

riously” suitable for an EFA (Kaiser and Rice, 1974).

The Measure of Sampling Adequacy (MSA) follows

an approach similar to the one pursued by the KMO

criterion but serves as an instrument for eliminating

those variables which do not qualify for the EFA.

Based on this analysis we found that he majority, i.e.

22 of our 31 model-based questionnaire items, qualify

as “meritoriously” suitable (MSA>0.8). The remain-

ing 9 values, i.e. SN1, SN2, PBC2, PBC3, PBC4,

DIS1, DIS2, DIS3 and DIS4, qualify as “mediocre”

(MSA>0.6), but still did not have to be excluded for

the EFA.

The consequent principle component analysis

pointed to an eight-factor model with each factor hav-

ing an Eigenvalue >1. As this model is able to explain

66.402% of the total variance implied by the rele-

vant data, the validity of our underlying research con-

structs composed of four TPB components and four

TRI 2.0 components, could be conﬁrmed.

5 MULTIVARIATE STATISTICS

Previous work proposes to deﬁne the mean value of 3

as a cut-off point for measurements based on the Lik-

ert response format (Gombachika and Khangamwa,

2013). The argument derives from 3 being inter-

preted as “undecided” or “neutral”. Participants’ re-

sponses to TPB questionnaire items all exceeded this

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention

middle point, indicating an overall positive attitude

(ATT mean=3.46) and favorable inﬂuences from their

social environment (SN mean=3.12) towards using

mobile IM (i.e WhatsApp) for CSI. A strong agree-

ment with the Perceived Behavioral Control (PBC

mean=4.43) further shows that people belief that the

use of WhatsApp for CSI would be easy and that they

feel conﬁdent applying it. Consequently, it is not

surprising that the Behavioral Intention to use What-

sApp for CSI also surpassed an average rating (BI

mean=3.20). Finally, concerning technology-related

personality traits, the data shows that respondents

were rather Optimistic (OPT mean=3.74) as well as

Innovative (INN mean=3.43) but felt Discomfortable

(DIS mean=2.91) and Insecure (INS mean=3.35).

In order to further investigate the connection be-

tween TRI 2.0 and TPB elements, and consequently

tackle the hypotheses put forward earlier, a Multiple

Linear Regression (MRA) analysis was performed. A

step-wise method was applied, as it supports the ex-

clusion of those independent variables, which do not

meet a predeﬁned level of signiﬁcance (in our case

p<0.05). Consequently, only variables which make

a signiﬁcant contribution are included into the ﬁnal

model. According to our hypotheses, we tested the

inﬂuence of technology-related personality traits on

ATT, SN and PBC. Finally, one additional analysis

was necessary so as to evaluate (and potentially con-

ﬁrm) the relations within the original TPB construct.

The results of the MRA show no signiﬁcant inﬂu-

ence of INN, DIS and INS on ATT (p>0.050), lead-

ing to the rejection of hypotheses H1(b), H1(c) as

well as H1(d). OPT, however, had an inﬂuence on

ATT (p=0.000), although it was only able to explain

7.9% of its variance (R

=0.079). A positive connec-

tion was conﬁrmed by the respective regression coef-

ﬁcient (B=0.369; β=0.281). Consequently, H1(a) was

not rejected.

As for the relation between the TRI 2.0 compo-

nents and SN the data showed, similar to the previous

analysis concerning ATT, no signiﬁcance with respect

to INN, DIS and INS (p>0.050). The connection be-

tween OPT and SN was, however, again signiﬁcant

(p=0.002; R

=0.066) producing a regression coefﬁ-

cient that conﬁrms the positive inﬂuence of OPT on

SN and thus supports H2(a) (B=0.240; β=0.256).

Evaluating the construct with PBC as the depen-

dent variable - i.e. focusing on hypotheses H3(a),

H3(b), H3(c), and H3(d) - identiﬁed INN as the only

inﬂuencing construct (p=0.000). OPT, INS and DIS

did not show any effects for which H3(a), H3(c) as

well as H3(d) had to be rejected. With INN we saw

that it was able to explain 8.2% of the variance ex-

hibited by PBC (R

=0.082). The resulting regression

coefﬁcient led to the assumption that INN positively

inﬂuences PBC and thus supports H3(b) (B=0.201;

β=0.286).

The ﬁnal analysis evaluated the relations within

the original TPB construct. Both ATT and SN were

found to have an impact on BI, whereas PBC was

excluded from the construct. With two independent

variables being signiﬁcant (p<0.05) we were able to

formulate two different models. Model 1, integrat-

ing only ATT as an independent variable, was able to

explain 76.7% of its variance (R

=0.767; p=0.000).

Adding SN as a second independent variable we were

able to push the rate to 78.1% (R

=0.781). This

second model also showed positive regression coefﬁ-

cients for ATT (B=0.868; β=0.793; p=0.000) as well

as SN (B=0.220; β=0.144; p=0.003) consequently

supporting H4(a) and H4(b). A much higher regres-

sion coefﬁcient of ATT, however, shows that its in-

ﬂuence is much stronger than the inﬂuence of SN.

Figure 1 depicts the resulting research model with its

standardized path coefﬁcients and variances of depen-

dent variables.

In summary we may argue that our data did not

support the existence of a signiﬁcant relations be-

tween INN and ATT, DIS and ATT or INS and ATT.

Neither did we ﬁnd a connection between INN and

SN, DIS and SN, and INS and SN, or an effect of

OPT, INS or DIS on PBC. Effects between OPT and

ATT, ATT and BI, SN and BI, as well as between OPT

and SN and between INN and PBC were, however,

signiﬁcant.

6 INTERPRETATION OF

RESULTS

Looking at the above results from the perspective of

Attitude being the dependent variable, only an op-

timistic view of technology leads to a positive atti-

tude towards using mobile IM (i.e. WhatsApp) for

CSI. This complies with the ﬁndings of Chen &

Chen (Chen and Chen, 2008) and Shih & Fan (Shih

and Fan, 2013), who also found support for a positive

inﬂuence of Optimism on Attitude. Yet, while Chen &

Chen in addition found support for an INN-ATT rela-

tion, our study was unable to show that people’s Inno-

vativeness would lead to them having a more positive

Attitude towards using mobile IM for CSI. Similarly,

feelings of Discomfort and Insecurity did not show

any inﬂuence on our participants Attitude.

Concerning the inﬂuencers of Subjective Norm

we found a signiﬁcant connection with Optimism,

whereas a connection to Innovativeness highlighted

by Chen and colleagues (Chen and Chen, 2008) was

CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications

Figure 1: Results of the computed research model.

not inherent. However, we found that being open to

new technologies makes people feel that their social

environment would support the use of mobile IM for

CSI. Again feelings of Insecurity and Discomfort had

no signiﬁcant impact on people’s Subjective Norm.

Innovativeness was the only signiﬁcant inﬂuence on

Perceived Behavioral Control. One may thus argue

that, people considering themselves as technology pi-

oneers makes them feel comfortable and easily able to

perform mobile IM for CSI. Surprisingly, however, an

optimistic view of technology was not found to signif-

icantly contribute to this Perceived Behavioral Con-

trol, which contradicts Chen & Chen’s results, who

also identify INN to be the most important inﬂuencer

of PBC but further highlighted a signiﬁcant relation-

ship between OPT and PBC (Chen and Chen, 2008).

With respect to the actual TPB model, our study

found that a positive attitude towards the technology

is the strongest driver for people’s Behavioral Inten-

tion to use IM for CSI. Furthermore, we found that

perceived social pressure signiﬁcantly inﬂuences peo-

ple’s BI, although with a much lower impact than their

positive attitude. In contrast, people’s conﬁdence in

their ability to perform does not impact on their BI.

These results conﬁrm ﬁndings from previous studies

which show that “intentions are often formed without

subjective norms playing a major role” (Dabholkar,

1994). The lacking inﬂuence of PBC on BI is also

supported by the ﬁndings of Teo & Pok (Teo and Pok,

2003) who furthermore found signiﬁcant effects for

the relations ATT-BI as well as SN-BI.

In summary we may thus argue that our survey

results are generally in line with previous work, al-

though from a mobile IM perspective we expected

certain deviations. Social pressure, for example, was

expected to have less impact, as the actual communi-

cation would occur only between a customer and the

company, and does not include any peers. In contrast,

the perceived ability to use mobile IM was expected to

be the main requirement for BI, yet did not show any

signiﬁcant effect. Consequently we may also state

that our study, which essentially explored people’s IM

use for CSI and a respective connection between the

TRI 2.0 and the TPB, conﬁrms previous results re-

ported by Chen & Chen (Chen and Chen, 2008). In

doing so, it is particularly the TPB part which shows

high power in explaining people’s Behavioral Inten-

tion to use IM for CSI, accounting for 78.1% of the re-

spondent’s variance. To this end, previous studies fo-

cusing on other technologies were often signiﬁcantly

lower, e.g. Teo & Pok (Teo and Pok, 2003) were able

to explain 17.2%, Lu and colleagues (Lu et al., 2009)

reported 61% and P

uschel et al. (P

uschel et al., 2010)

68.8%.

7 CONCLUSION AND FUTURE

WORK

The presented work explores the acceptance of mo-

bile IM (i.e. WhatsApp) for CSI settings, using a

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention

model-based questionnaire survey that integrates pre-

vious work on human behavior and technology readi-

ness. Its results may be summarized by three core

ﬁndings: (1) although we were not able to deduce a

reliable statement as to whether people would actually

like to use mobile IM for CSI, our data shows a slight

tendency towards a positive adoption; (2) the drivers

that may potentially lead to the adoption of mobile IM

for CSI are mostly found in people’s attitude towards

the technology (Note: while attitude has the strongest

inﬂuence, also people’s subjective norms play a sig-

niﬁcant role. One may therefore argue that friends or

family members as well as other inﬂuencers such as

role models or tech bloggers have a positive impact

on people’s intention to adopt mobile IM for CSI);

(3) technology-related personality traits did not inﬂu-

ence technology adoption, i.e. the personality traits

measured by the TRI 2.0 added little to explain the

adoption of IM for CSI.

Future studies will aim at deepening our un-

derstanding in this application area. In particular

we want to focus on other models integrating tech-

nology readiness and technology-related personality

traits. Using the model by Chen & Li (Chen and Li,

2010) may, for example, show that not the underly-

ing technology-related personality traits but technol-

ogy readiness as a whole inﬂuences the adoption of

mobile IM. Furthermore, the direct comparison of an

integrated TRI-TPB model with an integrated TRI-

TAM model might deliver valuable insights. Our

study was not able to show that technology-related

personality traits have a meaningful inﬂuence on peo-

ple’s behavioral intention to adopt mobile IM for CSI.

Such may imply that for this speciﬁc application area

technology-related personality traits are of little im-

portance. On the other hand, this could also mean

that the research model needs to be re-calibrated.

Also, an exploratory approach based on qualita-

tive research methods may be an interesting path for

further investigation, as it may add to a better discov-

ery of underlying reasons (Hyde, 2000). In particu-

lar privacy aspects could be of high relevance here.

Finally, the emerging market of mobile IM for busi-

nesses shows that artiﬁcial intelligence enabled mes-

saging bots receive a lot of attention in the tech-

community. Next to convenience tasks, also cus-

tomer service belongs to the targeted application ar-

eas, wherefore a more general exploration of messag-

ing adoption and its inﬂuencers may also be an inter-

esting future research direction.

REFERENCES

Ajzen, I. (1985). From intentions to action: A theory of

planned behavior.

Ajzen, I. (1991). The theory of planned behavior. Orgniza-

tional Behavior and Human Decision Processes,

50:179–211.

Ajzen, I. (2005). Attitudes, Personality and Behaviour.

In Manstead, T., editor, Mapping Social Psychology,

pages 1–178. McGraw-Hill Education, New York, 2

edition.

Basg

oze, P. (2015). Integration of Technology Readiness

(TR) Into the Technology Acceptance Model (TAM)

for M-Shopping. International Journal of Scientiﬁc

Research and Innovative Technology, 2(3):26–35.

Cariﬁo, J. and Perla, R. J. (2007). Ten Common Misunder-

standings, Misconceptions, Persistent Myths and Ur-

ban Legends about Likert Scales and Likert Response

Formats and their Antidotes. Journal of Social Sci-

ences, 3(3):106–116.

Chen, S.-C. and Chen, H.-H. (2008). The inﬂuence of

technology readiness on the theory of planned behav-

ior with self-service technologies. WMSCI 2008 -

The 12th World Multi-Conference on Systemics, Cy-

bernetics and Informatics, Jointly with the 14th Inter-

national Conference on Information Systems Analysis

and Synthesis, ISAS 2008 - Proc., 6(40):85–90.

Chen, S.-C. and Li, S.-H. (2010). Consumer adoption of

e-service: Integrating technology readiness with the

theory of planned behavior. African Journal of Busi-

ness Management, 4(16):3556–3563.

Church, K. and de Oliveira, R. (2013). What’s up with

whatsapp?: comparing mobile instant messaging be-

haviors with traditional sms. In Proceedings of the

15th international conference on Human-computer

interaction with mobile devices and services, pages

352–361, Munich, Germany. ACM.

Dabholkar, P. (1994). Incorporating Choice into an Attitu-

dinal Framework: Analyzing Models of Mental Com-

parison Processes. Journal of Consumer Research,

21(1):100–118.

Dabholkar, P. (1996). Consumer evaluations of new

technology-based self-service options : An investiga-

tion of alternative models of service quality. Intern. J.

of Research in Marketing, 13(1):29–51.

Davis, F. (1989). Perceived Usefulness, Perceived East of

Use, and User Acceptance of Information Technology.

MIS Quarterly, 13(3):319–340.

Davis, F., Bagozzi, R., and Warshaw, P. (1989). User

Acceptance of Computer Technology: a Comparison

of Two Theoretical Models. Management Science.

Aug1989, 35(8):982–1003.

Deng, Z., Lu, Y., Kee, K., and Zhang, J. (2010). Under-

standing customer satisfaction and loyalty: An em-

pirical study of mobile instant messages in China.

International Journal of Information Management,

30(4):289–300.

Dillon, A. and Morris, M. G. (1996). User acceptance

of new information technology: theories and models.

CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications

Annual Review of Information Science and Technol-

ogy Volume 31, Vol. 31:3–32.

Fishbein, M. and Ajzen, I. (1975). Belief, Attitude, Inten-

tion and Behavior: An Introduction to Theory and Re-

search. Addison-Wesley, Reading, MA, USA.

Ghyas, Q. M., Sugiura, H., and Kondo, F. N. (2012). Ex-

amining User Intention Behaviour Towards e-Readers

in Japan Using the Decomposed Theory of Planned

Behaviour. In MOBILITY 2012, The Second Interna-

tional Conference on Mobile Services, Resources, and

Users, pages 38–45, Venice, Italy. IARIA XPS Press.

Glass, R. and Li, S. (2010). Social Inﬂuence and Instant

Messaging Adoption. Journal of Computer Informa-

tion Systems, 51(2):24–30.

Godoe, P. and Johansen, T. S. (2012). Understanding adop-

tion of new technologies: Technology readiness and

technoloy acceptance as an integrated concept. Jour-

nal of European Psychology Students, 3(1):38–53.

Gombachika, H. S. and Khangamwa, G. (2013). ICT readi-

ness and acceptance among TEVT students in Univer-

sity of Malawi. Campus-Wide Information Systems,

30(1):35–43.

Guhr, N., Loi, T., Wiegard, R., and Breitner, M. H.

(2013). Technology Readiness in Customers ’ Per-

ception and Acceptance of M(obile ) -Payment : An

Empirical Study in Finland , Germany , the USA and

Japan. International Conference on Wirtschaftsinfor-

matik, March:119–133.

Huang, A. H., Yen, D. C., and Zhang, X. (2008). Explor-

ing the potential effects of emoticons. Information &

Management, 45(7):466–473.

Hyde, K. F. (2000). Recognising deductive processes in

qualitative research. Qualitative Market Research: An

International Journal, 3(2):82–90.

Jin, C. H. (2014). Adoption of e-book among college stu-

dents: The perspective of an integrated TAM. Com-

puters in Human Behavior, 41(1):471–477.

Kaiser, H. F. and Rice, J. (1974). Little Jiffy, Mark IV. Edu-

cational and psychological measurement, 34(1):111–

117.

Lancaster, S., Yen, D. C., Huang, A. H., and Hung, S.-Y.

(2007). The selection of instant messaging or e-mail:

College students’ perspective for computer communi-

cation. Information Management & Computer Secu-

rity, 15(1):5–22.

Li, D., Chau, P. Y., and Lou, H. (2005). Understanding In-

dividual Adoption of Instant Messaging: An Emprical

Investigation. Journal of the Association for Informa-

tion Systems, 6(4):102 – 129.

Lin, C.-H., Shih, H.-Y., and Sher, P. J. (2007). Integrat-

ing Technology Readiness into Technology Accep-

tance: The TRAM Model. Psychology & Marketing,

24(7):641–657.

Lu, Y., Zhou, T., and Wang, B. (2009). Exploring Chinese

users’ acceptance of instant messaging using the the-

ory of planned behavior, the technology acceptance

model, and the ﬂow theory. Computers in Human Be-

havior, 25(1):29–39.

Mick, D. G. and Fournier, S. (1998). Paradoxes of

Technology: Consumer Cognizance, Emotions, and

Coping Strategies. Journal of Consumer Research,

25(2):123–143.

Nor, A. S. A.-A. K. M. (2015). Challenges of adoption

of internet banking service in Yemen. International

Journal of Bank Marketing, 33(2):178–194.

Parasuraman, A. (2000). Technology Readiness Index

(TRI) - A Multiple-Item Scale to Embrace New Tech-

nologies. Journal of Service Research, 2(4):307–320.

Parasuraman, A. and Colby, C. L. (2014). An updated

and streamlined technology readiness index: TRI 2.0.

Journal of service research, 18(1):1–16.

uschel, J., Mazzon, J. A., and Hernandez, J. M. C. (2010).

Mobile banking: proposition of an integrated adoption

intention framework. International Journal of Bank

Marketing, 28:389–409.

Rennecker, J. and Godwin, L. (2003). Theorizing the Un-

intended Consequences of Instant Messaging (IM) for

Worker Productivity. Sprouts: Working Papers on In-

formation Environments, Systems and Organizations,

3(3):137–168.

Shih, Y.-C. and Fan, S.-t. (2013). Adoption of Instant

Messaging By Travel Agency Workers in Taiwan: In-

tegrating Technology Readiness with the Theory of

Planned Behavior. International Journal of Business

and Information, 8(1):120–136.

Shih, Y.-Y. and Fang, K. (2004). The use of a decomposed

theory of planned behavior to study Internet banking

in Taiwan. Internet Research, 14(3):213–223.

Shin, S. and Lee, W. J. (2014). The effects of technology

readiness and technology acceptance on NFC mobile

payment services in Korea. Journal of Applied Busi-

ness Research, 30(6):1615–1626.

Taylor, S. and Todd, P. A. (1995). Understanding informa-

tion technology usage: A test of competing models.

Teo, T. S. H. and Pok, S. H. (2003). Adoption of WAP-

enabled mobile phones among Internet users. Omega,

31(6):483–498.

Venkatesh, V., Morris, M. G., Davis, G. B., and Davis, F. D.

(2003). User Acceptance of Information Technology:

Toward a Uniﬁed View. MIS Quarterly, 27(3):pp. 425

– 478.

Walczuch, R., Lemmink, J., and Streukens, S. (2007). The

effect of service employees’ technology readiness on

technology acceptance. Information and Manage-

ment, 44(2):206–215.

Yen, J.-y. and Chen, S.-c. (2010). Perspectives from the

TRAM Model on Adopting e-Learning: An Analysis

of the Chain and Franchise Industry in Taiwan.

Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention