Telematics and Informatics 31 (2014) 184193

Contents lists available at ScienceDirect

Telematics and Informatics

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

When lifestyle becomes behavior: A closer look at the situational context of mobile communication
Veronika Karnowski a,, Olaf Jandura b,1
a b

Institut fr Kommunikationswissenschaft, und Medienforschung, LMU Mnchen, Oettingenstr., 67, 80538 Mnchen, Germany Kommunikations- und Medienwissenschaft, Heinrich-Heine-Universitt Dsseldorf, Universittsstr. 1, 40225 Dsseldorf, Germany

a r t i c l e

i n f o

a b s t r a c t
The web is going mobile, and the scope of mobile communication is widening tremendously, thus paving the way for a wide array of new forms of mobile device use. However, not every user is necessarily all the time taking advantage of the expanded affordances of mobile devices. Texting and phoning are still the predominant services in mobile communication. Previous research has argued that different styles of mobile communication are related to different user lifestyles. Thus, a remapping and matching of the landscape of mobile communication in relation to user lifestyles seems necessary. In this paper, we take one step back and rst consider the instances in which lifestyles become behavior; i.e. actual usage situations of mobile communication. We outline three empirically deduced types of mobile communication usage situations, as well as three types of mobile web usage situations, to shed light on the instant at which lifestyle becomes behavior; i.e. at which specic usage situations of mobile communication actually occur. 2013 Elsevier Ltd. All rights reserved.

Article history: Received 30 April 2012 Received in revised form 20 February 2013 Accepted 5 November 2013 Available online 11 November 2013 Keywords: Mobile communication Mobile web Media use Quantitative survey Classication Situational approach

1. Introduction The web is going mobilea trend which cannot be denied any longer. The scope of mobile communication has widened tremendously, thus paving the way for a wide array of new forms of use. However, not every user is necessarily taking advantage of these expanded possibilities of mobile phone use. Texting and phoning are still the predominant services in mobile communication (e.g. Boase and Ling, 2011). Previous researchers have argued that different styles of mobile communication are related to the different lifestyles of users (e.g. Bouwmann et al., 2012). Thus a remapping and matching of the landscape of mobile communication in relation to user lifestyles seems necessary. In this paper, we take a step back and rst consider instances in which lifestyle patterns become communication behaviors; i.e. we examine actual usage patterns of mobile communication and their specic contexts. On the one hand, research in the tradition of communication studies has been mostly blind to the situational contexts of new media usage; on the other hand, research in information systems and computing has integrated context factors, but not precisely at the situational level, and mostly from the viewpoint of technical artifacts rather than user standpoints. We would like to help close this gap by integrating situational context factors in a user-centered analysis of mobile communication behavior. In this rst step we outline a classication of mobile communication usage situations based on situational contexts, and examine differences among the services used in relation to these situational contexts.

Corresponding author. Tel.: +49 89 2180 9495; fax: +49 89 2180 9429.
1

E-mail addresses: karnowski@ifkw.lmu.de (V. Karnowski), jandura@phil-fak.uni-duesseldorf.de (O. Jandura). Tel.: +49 211 81 10660; fax: +49 211 81 15212.

0736-5853/$ - see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tele.2013.11.001

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

185

2. Situational contexts and usage of new media services Researchers commonly refer to the Theory of Planned Behavior (TPB; Ajzen, 1985) or various permutations thereof, to explain the adoption and usage of new media services (e.g. see Bouwman et al., 2007; Wang et al., 2008). The TPB (Ajzen, 1985) developed as an offshoot of the Theory of Reasoned Action (Fishbein and Ajzen, 1975), which takes into account the inuence of social norms on the adoption decision. According to the TPB, behavior is inuenced not only by attitudes towards the behavior in question, but also by subjective norms and perceived behavioral control. Attitudes towards a behavior consist of two interacting components: an individuals expectations regarding the consequences of the behavior in question, and his/her positive or negative evaluations of these consequences. Subjective norms refer to the pressure exerted by the social surroundings of an individual, which inuence the individual to execute or not to execute the behavior in question. Social norms also consist of two components: the individuals appraisal of what behavior is expected by his/her peers, and his/her evaluation of these expectations. Perceived behavioral control refers to the extent to which an individual feels able to execute his/her behavior; it consists of both situational and internal dimensions. The situational dimension describes the extent to which an individual objectively can execute a given behavior, while the internal dimension refers to whether the individual subjectively feels that he/she is able to execute the behavior (see Ajzen, 2005). The most prominent extensions of TPB in the eld of new media include the Technology Acceptance Model (TAM; Davis, 1989), the Unied Theory of Acceptance and Use of Technology (UTAUT; Venkatesh et al., 2003), and the Mobile Phone Appropriation model (MPA-model; Wirth et al., 2008). These theories elaborate on the TPB in terms of differentiating both the factors that inuence new media behaviors (TAM, UTAUT, MPA-model) and the actual forms of the new media behaviors (MPA-model). However, these theories have one prominent shortcoming: they do not include situational contexts. This limitation becomes especially poignant when analyzing mobile communication. As mobile services are ubiquitous, the situations that are involved, and their respective requirements, are virtually unlimited. Situational contexts of usage must, therefore, be integrated into our analysis. The information systems literature has discussed similar factors for over a decade, although they have generally been labeled as context factors. As early as Kristoffersen and Ljungberg (1999) pointed out that the use context of mobile handsets in work life varies widely across different professions. Similarly, Perry et al. (2001) noted the inuence of social and infrastructural factors on mobile computing. In subsequent years, several studies in the eld of information systems drew upon context factors when analyzing the adoption and intended usage of various innovations, clearly stating the impact of these context factors on adoption and usage patterns (e.g. Mallat et al., 2009; Bouwman et al., 2012; Turner et al., 2008; van de Wijngaert and Bouwman, 2009). However, these studies did not focus on the usage situation per se, but were still operating under the assumptions of mixed and/or broader levels of analysis. The analysis of single usage situations requires that we probe deeper, and ask what exactly constitutes a single usage situation. Belk (1975) dened the environment of consumer behavior on the basis of ve categories: physical, social, temporal, task, and antecedent states. While not completely based on a specic situation, these categories provide us with insights into relevant dimensions of situational contexts; i.e. physical and social. Similarly, Lee et al. (2005) and Vartiainen (2006) split situational contexts into physical and social (human) factors. The situational model of Zhang and Zhang (2012) follows similar lines, but focuses on media behaviors. Zhang and Zhangs model distinguishes between two interdependent factors inuencing new media behaviors: personal psychologies on the one hand, and location-related conditions on the other (see Fig. 1). Personal psychologies are conceptualized alike gratications sought, as in the Uses and Gratications Approach (see Katz et al., 1974; Palmgreen and Rayburn, 1985). Regarding location-related conditions, Zhang and Zhang differentiate

Fig. 1. The integrated model of computer multitasking (Zhang and Zhang, 2012, p. 1886).

186

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

the dimensions of physical environments, media access, and social dynamics with respect to multitasking on and with mobile phones. Location-related conditions include the physical environment, media access, and social dynamics. Physical environments consist of a users familiarity with his/her surroundings; i.e. whether he/she is at home or in public, and whether he/she is familiar with his/her environment. Media access refers to whether the user has other opportunities for media access in addition to a mobile phone, and if so, which? Media access is postulated to inuence new media behaviors. Finally, the specic social dimensions in a situation form part of the location-related conditions. Is the user amongst peers or is he/she alone; i.e. who is in the users actual surroundings? Social dimensions also inuence new media behavior. By combining the TPB and its extensions with the Situational Theory of New Media Behaviors, we can conceptualize mobile web usage as inuenced by user-related factors, location-related factors, and restrictions (see Fig. 2). User-related factors refer to the actual emotional state of a user, according to Bradley and Lang (1994). Location-related factors are differentiated into physical environments, media access, and social dynamics, according to Zhang and Zhang (2012; see above). Restrictions are differentiated into nancial, technical, temporal, and cognitive restrictions, according to Wirth et al. (2008). Based on the theoretical framework outlined above, our research questions are as follows: RQ1: In which situations, dened by physical environments, media access and social dynamics, do people use their mobile devices? RQ2: What different types of situations can be distinguished? RQ3: Can we distinguish between a specic set of usage situations for mobile web usage as compared with other mobile communication services? 3. Method and measures To answer the above research questions, we conducted an online survey in cooperation with Tomorrow Focus AG, one of Germanys leading publicly traded internet companies. The survey was conducted during the period 25 November to 23 December 2010, using a quantitative online questionnaire. To achieve our goal of collecting not a representative but a heterogeneous sample, the questionnaire was linked to several websites which, in combination, reach two thirds of all German internet users. The websites included news sites and special interest sites (e.g. pages for women, men, people interested in fashion, etc.). However, because of privacy regulations, our survey results did not include the website that the survey participants had visited to take part in the survey. The nal sample consisted of 1400 individuals. The composition of the sample resembled that of average German daily internet users (ALLBUS, 2010): 72% of the respondents were male, 28% were female, the average age was 41 years, 55% reported a low to moderate degree of education, 36% reported a high degree of education, and 31% (432 respondents) browsed the internet via a mobile device. Our data are consistent with the ndings that in Germany, men tend to go online more often than women, indicating that there is still a gender gap in internet usage in Germany (Zillien, 2009). In terms of patterns of online usage with mobile devices, we found that: the mobile online group was a little younger on average (average age, 37), and in this group the percentage of higher educated respondents was higher (43%), the percentage of respondents with a higher income was greater (15% vs. 9%, respectively), and the percentage of respondents with professional lifestyles was greater (84% vs. 72%, respectively). Because of the substantial similarity of the sample to representative samples of German daily online users, we refrained from weighing the sample using sociodemographic variables and the frequency of the internet use.

Fig. 2. Factors inuencing mobile web usage.

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

187

The questionnaire contained questions that provided us with information on four online usage dimensions. The rst dimension focused on the prerequisites of mobile internet usage. Respondents were asked whether or not they possessed a mobile device capable of online access; if so, the respondents were asked if they actually used the device to access the internet, and if so, when was the last time that they had used the device in this way. The second dimension was used to characterize the last usage episode. Because mobile internet use was in an early stage at the time of the survey, we asked about the last time that the respondent had actually gone online via a mobile device. In this way, we gained information only on usage situations relevant to our study. By initially addressing users with a sudden popup window during actual internet sessions, we were assured that: (1) The times of day that online users were asked to take part in the survey varied, as did the last mobile internet usage situation; 29% of the questionnaires were completed in the morning (6 amnoon), 15% in the afternoon (noon5 pm), 16% in the early evening (5 pm8 pm), 21% in the late evening (8 pmmidnight), and 21% at night (midnight6 am). (2) Participants took part in the survey only once, thus maintaining the statistical independence between personal and situational variables. (3) The questionnaire was not completed only on weekends, as internet usage patterns on weekends typically differ from those on weekdays. Because our goal was to analyze the last usage episode in detail, we asked about aspects related to: the online sites that were visited, the respondents location (nominal scale: at home/at work/en route), the presence of other people (nominal scale: family/friends/colleagues/strangers/nobody), whether the location they were in when they accessed the internet was familiar (nominal scale: yes/no), and whether alternative media services were available at the time (newspapers, magazines, TV, radio, internet via desktop/laptop). We also asked about the respondents mood during online access, measured by the self-assessment manikin (SAM), a non-verbal pictorial assessment technique (Bradley and Lang, 1994). Finally, we asked about possible restrictions on access (temporal, cognitive, technical, nancial; Wirth et al., 2008). Each dimension was explored using two items, and respondents were asked about the extent to which they agreed or disagreed with each item, using a 7-point Likert scale. Third, we measured motives and gratications. In line with previous studies, we asked respondents about their motives and gratication in relation to their online experiences using a mobile device. We included the following dimensions according to the previous studies of Leung and Wei (2000), LaRose and Eastin (2004) and Wei (2008): information, entertainment, maintaining relationships, status, and availability/access. A total of 21 gratication items were measured. Each dimension was explored using four to ve items. A 5-point Likert scale was used to determine the extent to which respondents agreed or disagreed with each item. 3.1. Control measures We measured demographic variables such as age, gender, region (nominal scale: village/small city/big city), income (3-point scale: low/middle/high), formal education (3-point scale: low/middle/high), and currently employed (nominal scale: yes/no). In addition, we asked how often, typically, the respondent was in an unknown location (5-point Likert scale, ranging from very often to not at all) and how often he/she spends time traveling (5-point Likert scale, ranging from very often to not at all). 4. Results 4.1. Mobile communication in general To identify different classes of mobile communication usage situations, we used the clustering technique of latent class analysis (LCA). Latent class analysis has a variety of advantages over traditional cluster analysis, as it allows for the classication of variables at each level of measurement, and different levels of measurement can be integrated into the analysis. In contrast to traditional cluster analysis, LCA does not necessarily result in a cluster solution, and it can also reject clustering of the data (see Fraley and Raftery, 1998). Latent class analysis provides statistical tests to identify the exact number of clusters. Accordingly, it is less arbitrary than traditional cluster analysis. On account of its probabilistic conception, LCA also takes into account the possibility that the clustered variables might not be completely reliable or completely valid. As outlined above, usage situations consist of user-related and location-related aspects, as well as restrictions. Unfortunately, the variance of the emotional state of our respondents, as well as that of perceived restrictions, was low throughout our sample. Thus, only location-related aspects of mobile web usage situations, as well as the availability of data services on the mobile device in a specic situation, could be integrated into the LCA. To identify the number of clusters, we rst calculated and compared one- to ten-cluster solutions. All but the one-cluster solution show a non-signicant p-value of the likelihood-ratio test and the CressieRead test; thus, the model predictions did not differ signicantly from the observed data. A check of the likelihood-ratio test by both the bootstrapping method and Pearsons v2 also indicates a signicant p-value for the two-cluster solution, which can thus be eliminated.

188

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

Consequently only the three-, four-, and six- to ten-cluster solutions were retained in the analysis (see Table 1). Generally, the solution including the fewest parameters to be estimated, and thus the lowest BIC value, is considered the most suitable. The three-cluster solution meets this criterion. Based on the LCA, there are specic probabilities of the different parameter values integrated into the analysis for each cluster. The different clusters can be described on the basis of these probabilities (see Table 2). Cluster 1: Mobile@home In the evening, Sarah is watching TV at home. All of a sudden, she remembers that it is her cousins birthday today. As it is already too late to call, she quickly sends a text to her cousin saying, Happy Birthday!. This is by far the largest cluster, containing 56.5% of all usage situations. These usage situations most likely occur in a xed and highly familiar location (such as at home). Cluster 2: En route Mr. Smith is on his way to meet a former class mate. Unfortunately, he left the exact address at home. So, he tries to reach some of his other former class mates to ask for the address. As he does not have a at rate plan, he wants to keep the conversation as short as possible. Usage situations in this cluster represent mobile communications occurring while the user is on his/her way. The social surroundings are mostly unknown to the user, and the familiarity with the location varies widely. The availability of online services is least probable in this cluster. This cluster is far smaller than cluster 1 (including only 23.8% of all usage situations). Cluster 3: Hanging out with peers Tom and some of his mates are sitting in a bar. They all have at rate plans for their cell phones. Throughout the evening, they are constantly posting photographs and comments on Facebook. This usage cluster, which includes 19.8% of all usage situations, is the smallest usage cluster. Here, usage most probably occurs en route, but could also be in a xed location. The social surroundings are well known to the user, but the familiarity
Table 1 Likelihood-ratio test (including bootstrapping), CressieRead test, Pearsons v2, and the Bayesian information criterion (BIC) for the one- to ten-cluster solutions for mobile communication usage situations.

Table 2 Average probabilities of parameter values, explained variance of classied variables, and relative cluster size of the three-cluster solution for the mobile communication usage situations. Cluster 1 (%) Online services available on mobile device Fixed location On the road Known social surroundings Financial restrictions low middle high Familiarity with location low middle high Relative size 67 100 0 53 56 33 12 2 13 85 56.5 Cluster 2 (%) 56 0 99 5 52 34 14 17 30 53 23.8 Cluster 3 (%) 75 17 99 78 61 30 9 20 31 49 19.8 R2 (%) 1.9 87.7 97.4 25.2 0.5

15.4

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

189

with the location is low. The probability of the availability of online services is higher in this cluster than in any of the other clusters. We now compare the classes of mobile communication usage situations based on sociodemographic parameters, gratications sought, and services used. To do so, each case is attributed to the cluster it most probably belongs to. The classication error in this procedure (i.e. the percentage of cases which are classied incorrectly) is 5.1%. Despite the overall predominance of males in our sample, the usage cluster Hanging out with peers occurs more often for male than for female users (p < 0.05); the same is true for users with a lower education level; that is, Hanging out with peers occurs more often amongst those with low than those with medium or high education levels (p < 0.05). In addition, users in the usage cluster En route are signicantly older than are all other users in our sample (see Table 3). Overall, telephony is the dominant usage of mobile phones. This service is used signicantly more often in En route situations than in Mobile@home and Hanging out with peers situations. Text messaging occurs most often in Mobile@home situations. Other services are most often used when Hanging out with peers. For the other mobile phone services, there are no signicant differences in usage between the three types of mobile communication situations (see Table 4). Items pertaining to gratication of mobile communication were subsumed to ve indices: status, maintaining relationships, entertainment, access, and information. According to Cronbachs alpha values, the internal reliability of these indices is satisfactory (see Table 5).

Table 3 Sociodemographic parameters by mobile communication usage situations. Mobile@home(n = 483) (%) Sex Male Female 70a 30a En route(n = 381) (%) 74a 26a 15a 25 50 44.8b Hanging out with peers(n = 87) (%) 81b 18b 25b 19 44 38.9a F-Value 3.8

g2 (%)
0.6

Educational level Low 18a Medium 25 High 44 Age

3.0 1.1 3.0 20.6

0.5 0.2 0.5 3.1

38.9a

p < 0.05,

p < 0.01,

p < 0.001; means marked by different characters differ signicantly.

Table 4 Percentage of services used by mobile communication usage situations. Service used Telephony SMS/MMS Radio/MP3 E-mail Internet via browser Internet via app Games Other (camera, calendar, alarm clock, . . .)

Mobile@home (n = 483) (%) 45 27a 4 5 6 3 3 9a

a

En route (n = 381) (%) 59 14b 4 5 7 4 2 6a

b

Hanging out with peers (n = 87) (%) 43 17b 3 5 12 4 2 15b

a

FValue 13.7 16.1 0.2 0.1 2.6 0.5 0.5 6.2

g2
(%) 2.1 2.5 0.0 0.0 0.4 0.1 0.1 1.0

p < 0.05,

p < 0.01,

p < 0.001; means marked by different characters differ signicantly.

Table 5 Means and Cronbachs alpha values for gratication indices of the mobile communication usage situations. Gratication indices Status Maintaining relationships Entertainment Access Information Number of items 4 5 4 4 4 Cronbachs alpha 0.74 0.77 0.87 0.76 0.70 Mobile@home (n = 483) 4.5a 3.6a 3.0 3.6a 3.0a En route (n = 381) 4.7a 3.7a 3.3 3.7a 3.0a Hanging out with peers (n = 87) 4.0b 2.7b 3.0 3.0b 2.2b FValue 10.1 11.5 2.1 5.5 5.9

Agreement on a 5-point scale, from 1 = strongly agree to 5 = strongly disagree. p < 0.05, p < 0.01, p < 0.001, means marked by different characters differ signicantly.

190

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

Signicant differences occur in all but one dimension of gratications sought between the clusters Mobile@home and En route and the cluster Hanging out with peers. In situations of Hanging out with peers, users assign higher importance to status, maintaining relationships, access, and information, than do users in other usage situations.

4.2. Mobile web use In this section, we concentrate on mobile web use, which is a specic sub-set of mobile communications. In our sample, 432 respondents were mobile online users, and each of these respondents reported their last mobile web usage situation. To identify specic classes of mobile web usage situations, we used the LCA clustering technique. To identify the number of clusters, we rst calculated and compared the one- to ten-cluster solutions. All but the onecluster solution show non-signicant p-values for the likelihood-ratio test and the CressieRead test; the same results were obtained using the bootstrapping method. Pearsons v2, on the other hand, indicates signicant p-values for the two-cluster and ve-cluster solutions, which can thus be eliminated. As the three-cluster solution has the lowest BIC value, this solution is the most suitable (see Table 6). We now describe these three clusters based on the specic probabilities of the different parameter-values integrated in the analysis for each cluster (see Table 7).

Cluster 1: On the way Each morning, Mr. B is sitting in the metro on his way to work. Sometimes his neighbor is sitting beside him, as they are both commuting. Some years ago, he used to buy a newspaper to read on the train, but he does not do so any longer. Each morning while commuting, he is checking his favorite news sites via his smartphone.
Table 6 Likelihood-ratio (incl. Bootstrapping), CressieRead, Pearsons v2, and Bayesian information criterion (BIC) tests of the one- to ten-cluster solutions for mobile web usage situations.

Table 7 Average probabilities of parameter values, explained variance of classied variables, and relative cluster size of the three-cluster solution for mobile web usage situations.

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

191

On the way is the largest cluster of mobile web usage situations. Usage situations in this cluster occur while the user is in transit. In these circumstances, the user has no, or nearly no, competing media access to mobile web usage. As compared with the circumstances in the other two clusters, the social surroundings are rather unfamiliar to the user, and the probability of a rather unfamiliar location is highest in this cluster. Cluster 2: Homezone Heather is at home watching a new serial on TV. During the commercial break, she pulls out her smartphone and browses through her friends Facebook posts. She could check them using her PC, but she is too lazy to get up, go to her desk, and start the PC. These usage situations most probably occur in private and very familiar locations. Possibilities of alternative media access are rather high. The probability that known individuals are present in the users social surroundings is average; i.e. the user might be in the companionship of others, but he/she might also be alone. Cluster 3: Work or friends Mark and Kate are spending an evening at their favorite bar together with their friends. Mark tells them about a book he is currently reading, but he can not remember the authors name. Thus, he takes out his smartphone and Googles the authors details. The probability of being together with peers or family is higher in this cluster than in the other clusters. The usage situations occur within xed and quite familiar locations, which are not necessarily in the private realm. Additional media access varies within these usage situations, but might be rather low. We compared the usage clusters using the exogenous variables of age, gender, service used, and gratications sought. To do so, each case was attributed to the cluster it most probably belongs to, which resulted in a classication error of 8.1%. No differences were detected among the different types of mobile web usage situations with respect to users gender or educational level. Users in the cluster On the way were signicantly older than other mobile online users (see Table 8). News and looking up information are the most frequent activities when using an online service with the mobile phone. Looking up information occurs most often in Work or friends situations, next most often in On the way situations, and least often in (but still in 29% of) situations in which the user is in his/her Homezone. Signicant differences occur with respect to mobile use of TV and video platforms, which occur signicantly more often when with Work or friends than On the way or in the Homezone. The same is true for the mobile use of web radio (see Table 9). Signicant differences occur between the three clusters in terms of the three gratication dimensions: maintaining relationships, entertainment, and information. Maintaining relationships is of lower relevance when using mobile web services in situations of Work and Friends, as compared with the other two situations. Because users are likely already in the companionship of peers in these usage situations, additional social contact via the mobile web is of no importance in such situations.
Table 8 Sociodemographics by mobile web usage situations. On the way (n = 193) (%) Sex Male Female 80 20 Homezone (n = 174) (%) 83 17 11 24 50 35.1b Work or friends (n = 65) (%) 85 15 12 23 51 34.9b F-Value 0.4

g2 (%)
0.2

Educational level Low 12 Medium 25 High 56 Age

0.0 0.1 0.8 4.1

0.0 0.1 0.4 1.9

39.1a

p < 0.05,

p < 0.01,

p < 0.001, means marked by different characters differ signicantly.

Table 9 Percentage of services used by mobile web usage situations. Service used News E-commerce Looking up information Social networks Online gaming Chat/instant messaging TV or video platforms Web radio Other

On the way (n = 193) (%) 32 7 42a 20 3 12 4a 4a 2

Homezone (n = 174) (%) 32 6 29b 27 3 10 9a 4a 3

Work or friends (n = 65) (%) 37 15 55c 32 5 15 18b 12b 0

F-Value 0.4 2.8 7.8 2.3 0.3 0.7 7.7 3.7 1.2

g2 (%)
0.2 1.3 3.5 1.1 0.2 0.3 3.5 1.7 0.5

p < 0.05,

p < 0.01,

p < 0.001, means marked by different characters differ signicantly.

192

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

Table 10 Gratication indices by mobile web usage clusters. On the way (n = 192) Status Maintaining relationships Entertainment Access Information 4.3 3.5a 3.2a 3.5 2.7a Homezone (n = 170) 4.2 3.3a 2.8b 3.3 2.7a Work or friends (n = 64) 4.0 2.9b 3.0ab 3.1 2.3b F-Value 2.40 7.27 5.47 2.19 4.51

g2 (%)
1.2 3.3 2.5 1.1 2.1

Agreement on a 5-point scale, from 1 = strongly agree to 5 = strongly disagree. p < 0.05, p < 0.01, p < 0.001, means marked by different characters differ signicantly.

On the way and Homezone are signicantly different with respect to entertainment. Entertainment is more relevant when On the way than when at home, where one has access to a variety of other distractions. Information is also of least relevance in usage situations involving Work or friends; i.e. when in the presence of others, but most probably deprived of other media access (see Table 10). 5. Discussion, limitations and conclusion The Situational Theory of New Media Behaviors provides a promising approach for the analysis of mobile communication situations generally, and of mobile communication for online usage specically. Our online survey identied three main situations in which mobile communication occurs: the rst is when users are among known peers and in familiar locations, a situation which occurs mostly at home (Mobile@home); the second is when users are on the way and in unknown surroundings and among unknown people (En route); and the third is when users are with peers but in unknown locations, such as in a restaurant or bar (Hanging out with peers). The occurrence of these mobile communication usage situations varies according to age, gender, and educational level. In addition, services used and gratications sought are different among the different usage clusters. A closer look at mobile internet use reveals a subset of similar mobile internet usage patterns. Not surprisingly, mobile phones are used to log onto the internet when people are in unfamiliar locations and in public situations (On the way). Just as predictable is the nding that respondents used their mobiles to go online when they were in xed and familiar locations, such as in bars or plazas, when access to other helpful media were limited or unavailable (Work or friends). A rather unexpected nding, however, was the occurrence of many situations in which users, despite opportunities to use conventional PCs or laptops, went online using their mobile devices while at home (Homezone). This nd underscores the importance of mobile communication via mobile devices in general, and internet usage in particular, in xed locations. In terms of sociodemographic factors, the variance among these clusters is very low; however, the clusters do differ in terms of mobile web services used; i.e. entertainment-related aspects are of greater importance when at work or with friends (Work or friends), whereas information and news services are most frequently accessed when On the way or in the Homezone. Although we successfully identied three main mobile communication usage situations, as well as three main mobile internet usage situations, we feel that further research is required in this area on account of the theoretical and methodological constraints of our study. First, the question arises as to whether the high percentage of situations in which people use cell phones at home is an artifact produced by the questionnaire. The respondents encountered our survey and were asked about their last mobile communication use in a situation in which they were already accessing the internet. Often, people will have done so in a xed location, which, in the majority of cases, is probably in the so-called Homezone. Thus, we cannot rule out the possibility that our study overestimated the number of mobile usage situations at home. Second, the measurements of the emotional state of the participants can be questioned. It appears that, despite the use of a well-tested scale, we ended up with rather nonspecic data because respondents were required to assess their moods retrospectively. We assume that our participants were unable to truly remember how they felt during their last mobile online usage situation. If we are correct about this shortcoming, then our method of mood measurement during online sessions is questionable; i.e. an in situ measurement such as the experience sampling method might be required (see Hektner et al., 2007; Karnowski and Doedens, 2010; Larson and Csikszentmihalyi, 1983). Third, we suggest another reason for the fact that we obtained rather nonspecic responses to the questions dealing with temporal, cognitive, technical, and nancial restrictions on mobile communication. On account of technical improvements in mobile devices, it is possible that Wirth et al.s (2008) questionnaire items, which were developed four years before the present study was implemented, were already outdated at the time of the study, and that the relevant restrictions cannot be measured validly at this time. Fourth, our data characterize the situation in a western European nation prior to the widespread use of mobile internet devices, when online usage was still an activity that was very much linked to traditional desktops and laptops. Finally, we must bear in mind that our survey is cross-sectional. Therefore, we were only able to gather information on one single usage situation. Because of the one-shot nature of the survey, and our decision to deal with actual online sessions rather than self-assessment summaries of individuals mobile internet use, we were only able to collect data for one situation; i.e. the two concepts are confounded. We therefore recommend that future research produce longitudinal data for the analysis of mobile usage patterns.

V. Karnowski, O. Jandura / Telematics and Informatics 31 (2014) 184193

193

References
Ajzen, I., 1985. From intentions to actions: a theory of planned behavior. In: Kuhl, J., Beckman, J. (Eds.), Action-control: From Cognition to Behavior. Springer, Heidelberg, pp. 1139. Ajzen, I., 2005. Attitudes, Personality and Behavior, second ed. Open University Press, New York. Belk, R.W., 1975. Situational variables and consumer behavior. Journal of Consumer Research 2, 157164. Boase, J., Ling, R., 2011. Measuring mobile phone use: self-report versus log data. ICA Pre-Conference 2011 Seamlessly Mobile? Mobile Communication @ a Crossroads, 25./26.5.2011, Boston, USA. Bouwman, H., Bejar, A., Nikou, S., 2012. Mobile services put in context: a Q-sort analysis. Telematics and Informatics 29, 6681. Bouwman, H., Carlsson, C., Molina-Castillo, F.J., Walden, P., 2007. Barriers and drivers in the adoption of current and future mobile services in Finland. Telematics and Informatics 24, 145160. Bouwmann, H., Lopz-Nicols, C., Molina-Castillo, F.-J., van Hattum, P., 2012. Consumer lifestyles: alternative adoption patterns for advanced mobile services. International Journal of Mobile Communications 10, 169189. Bradley, M.M., Lang, P.J., 1994. Measuring emotion: the self-assessment manikin and the semantic differential. Journal of Behavioral Therapy and Experimental Psychiatry 25, 4959. Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 13, 319340. Fishbein, M., Ajzen, I., 1975. Believe, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addison-Wesley, Reading. Fraley, C., Raftery, A., 1998. How many clusters? Which clustering method? Answers via model-based cluster analysis. The Computer Journal 41, 578588. Hektner, J.M., Schmidt, J.A., Csikszentmihalyi, M., 2007. Experience sampling method: measuring the quality of everyday life. Sage, Thousand Oaks. Karnowski, V., Doedens, S., 2010. Mobile experience sampling. Eine Methode zur Untersuchung mobilen Mediennutzungsverhaltens [Mobile Experience Sampling. A method for analyzing mobile media usage]. In: Jackob, N., Zerback, T., Jandura, O., Maurer, M. (Eds.), Methoden der Online-Forschung: Das Internet als Forschungsinstrument und -gegenstand der Kommunikationswissenschaft. Verlag Herbert von Halem, Kln, pp. 211226. Lee, I., Kim, J., Kim, J., 2005. Use contexts for the mobile internet: a longitudinal study monitoring actual use of mobile internet services. International Journal of HumanComputer Interaction 18, 269292. Katz, E., Blumler, J.G., Gurevitch, M., 1974. Utilization of mass communication by the individual. In: Blumler, J.G., Katz, E. (Eds.), The Uses of Mass Communications. Current Perspectives in Gratications Research. Sage, Beverly Hills, pp. 168249. Kristoffersen, S., Ljungberg, F., 1999. Making Place to Make IT Work: Empirical Explorations of HCI for Mobile CSCW. GROUP 99 Proceedings of the international ACM SIGGROUP conference on Supporting group work, 276285. LaRose, R., Eastin, M., 2004. A social cognitive theory of internet uses and gratications: toward a new model of media attendance. Journal of Broadcasting & Electronic Media 48, 358377. Larson, R., Csikszentmihalyi, M., 1983. The experience sampling method. In: Reis, H.T. (Ed.), Paperback Sourcebooks in the Jossey-Bass Social and Behavioral Science Series. Naturalistic Approaches to Studying Social Interaction. Jossey-Bass, San Francisco, pp. 4156. Leung, L., Wei, R., 2000. More than just talk on the move: uses & gratications of the cellular phone. Journalism and Mass Communication Quarterly 77, 308320. Mallat, N., Rossi, M., Tuunainen, V.K., rni, A., 2009. The impact of use context on mobile services acceptance: the case of mobile ticketing. Information & Management 45, 190195. Palmgreen, P., Rayburn, J.D., 1985. An expectancy-value approach to media gratications. In: Rosengren, K.E., Wenner, A., Palmgreen, P. (Eds.), Media Gratications Research. Current Perspectives. Sage, Beverly Hills, pp. 6172. Perry, M., Ohara, K., Sellen, A., Brown, B., Harper, R., 2001. Dealing with mobility: understanding access anytime, anywhere. ACM Transactions on ComputerHuman Interaction 8, 323347. Turner, M., Love, S., Howell, M., 2008. Understanding emotions experienced when using a mobile phone in public: the social usability of mobile (cellular) telephones. Telematics and Informatics 25, 201215. van de Wijngaert, L., Bouwman, H., 2009. Would you share? Predicting the potential use of a new technology. Telematics and Informatics 26, 85102. Vartiainen, M., 2006. Mobile virtual work concepts, outcomes and challenges. In: Andriessen, E.J.H., Vartiainen, M. (Eds.), Mobile Virtual Work. A new paradigm? Springer, Berlin, pp. 1344. Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F., 2003. User acceptance of information technology: toward a unied model. MIS Quarterly 27, 425478. Wang, C.C., Lo, S.K., Fang, W., 2008. Extending the technology acceptance model to mobile telecommunication innovation: the existence of network externalities. Journal of Consumer Behaviour 7, 101110. Wei, R., 2008. Motivations for using the mobile phone for mass communications and entertainment. Telematics and Informatics 25, 3646. Wirth, W., Von Pape, T., Karnowski, V., 2008. An integrative model of mobile phone appropriation. Journal of Computer-mediated Communication 13, 593 617. Zhang, W., Zhang, L., 2012. Explicating multitasking with computers: gratications and situations. Computers in Human Behavior 28, 18831891. Zillien, N., 2009. Digitale Ungleichheit: Neue Technologien und alte Ungleichheiten in der Informations- und Wissensgesellschaft [Digital inequality: new technologies and old inequalities in an information and knowledge society]. Vs Verlag, Wiesbaden.