Apps, activities and travel: an conceptual exploration based on activity theory

Over the past decades many researchers have addressed the question what the rise of ICTs implies for travel behaviour (e.g. Mokhtarian 2002; Farag et al. 2003; Wang and Law 2007; Zhang et al. 2007). A typical characteristic of this field is that it is rapidly evolving in response to technological development (Lyons 2014). Whereas about a decade ago, most advanced ICT applications were based on communication using land lines (e.g. teleworking or teleshopping from home), nowadays large segments of the population own smartphones, which allow them to be connected to the internet anywhere and anytime, and apply a wide variety of software (apps). It goes without saying that as a result, the variety of ICT tools available to travellers has increased exponentially (Schwanen 2015).

The central tenet of this paper is that this dramatic development in the context of ICTs (and apps in particular) has profound bearing on the discussion about the impact of ICTs on travel, calling for complementary frameworks to the ones used to day. Whereas many studies to date tend to describe the effects of ICTs on travel in terms of the amount of increase/decrease of travel, the increasing diversity of ICT tools calls for more detailed analysis, which not only focuses on specific ICT tools rather than on broad types of ICTs, but also focuses in much more detail on the practices of using ICT tools and the contextual factors that influence their use and effect, as well as their further development. In the context of apps, Schwanen (2015) has argued that rather than focusing on the traveller as a decision maker, it is helpful to focus on the App as the object of investigation as it can be regarded as a more or less autonomous agent entangled in a web of relationships with humans and technological devices.

In agreement with this reasoning, this paper proposes an alternative approach to analysing the effect of apps on travel. In particular, we propose activity theory (AT) (Engeström 2014) as a framework for analyzing the effects of apps on travel behaviour. This paper aims to explore how AT can be used to further our understanding of the impact of apps on travel, and complement the existing approaches focusing at travellers’ behaviour. To this end, we will use AT to systematically explore the mechanisms by which a number of different apps influence travel behaviour. We will assess the added value of AT as well as its limitations in the context of apps and travel.

The paper is organized as follows. “Existing approach to studying ICT and travel” section discusses existing approaches used to date to study the relationship between ICTs and travel. “Apps, activities and travel: a taxonomy” section then proposes a taxonomy of apps, which serves as a starting point for more detailed analyses using AT. “Activity theory” section introduces the principles of AT. “The effect of apps on activities and travel according to activity theory” section describes the application of AT to two smartphone apps (WhatsApp) and Travel Feedback apps. “Conclusions and discussion” section, finally, draws conclusions regarding the applicability of AT to the study of apps and their impact on travel, the mechanisms by which the two apps influence travel behaviour, and discusses possible avenues for further research.

At the rise of ICT technology (such as the Internet), expectations were high that ICTs would offer a substitute for on-site activities and thereby strongly reduce travel (Salomon 1998). This hypothesis spurred a strand of research testing the effect that ICTs have on the amount of travel. Research in this realm acknowledged that substitution might not be the only effect (Mokhtarian and Tal 2013). Alternatively, ICTs may stimulate travel, for instance because time is saved that can be used for travel, allows travel time to be used more efficiently and allows social networks to be maintained over longer distances. Other options are that ICT modifies travel (e.g. choosing a different destination because of online information) or has no effect on travel at all (neutrality).

Empirical studies in this stream have mostly investigated the amount of travel (number of trips by purpose, travel distance by purpose) as a function of a rather general definition of ICT use. For instance, Wang and Law (2007) define ICT use as a binary variable indicating experience with any of a set of ICT tools, and investigate its effect on time use, travel time and number of trips. Zhang et al. (2007) investigate the effect of a set of ICT indicators (frequency of internet use, number of cell phones in the household) on VMT and the number of walking and total trips. In an overview article Mokhtarian (2009) reports that generation is the dominant effect of ICT on travel.

In addition, papers discussing the positive utility of travel (e.g. Mokhtarian et al. 2001) note that portable ICTs may influence travel decisions by making travel more productive [e.g. by working on a laptop (see Lyons et al. 2013)] or more pleasant and less boring (e.g. by gaming, reading or communicating: (see Lyons et al. 2013; Ettema et al. 2012). By making travel time more pleasant of productive, these ICTs may influence travellers’ willingness to travel as well as travel time valuations.

Mokhtarian and Tal (2013) provide an overview of mechanisms increasing, decreasing or modifying travel which include a variety of ICT tools (phone, cell phone, email, web browsing etc.) applied for many different purposes. These examples imply that, while investigating the effect of general ICT indicators (e.g. frequency of Internet use) leads in itself to useful indications of ICT effects, they might be complemented by analyses looking more closely into the effects of specific ICT tools/applications, used for a specific purpose, since these might have specific outcomes. For instance, while using the Internet to buy groceries might have a travel reducing effect, browsing the Internet to find an outdoor activity on Sunday afternoon may have a travel generating effect. The need to focus in more detail on specific tools becomes even stronger in the context of multifaceted devices such as smartphones, which offer opportunities to call, text, browse the internet and use dedicated apps within one portable device, with each function and each app having potentially different implications for travel behaviour. This calls for an approach that distinguishes between specific ICT devices and tools, and the different practices of using them. In the remainder of the paper, we introduce AT as an alternative approach to studying the impact of ICTs, by focusing on smartphone apps as a specific category of ICTs.

A first category includes apps with functionalities not in a direct sense related to travel or physical out-of-home activities. Examples are apps offering access to online entertainment or games. In such cases, using the App is an activity in itself. Such apps may impact on out-of-home activities (and associated travel) by substituting for out-of-home activities (e.g. in-home gaming or entertainment instead of going to a cinema) or by competing for time. However, since apps are footloose, such entertainment based apps may influence travel and out-of-home activities also in a qualitative sense. Travelling while watching a movie or gaming may make travel more pleasant and less problematic, while using these apps during work, social and recreational activities fundamentally changes the value and character of these activities (Mokhtarian et al. 2001; Ettema et al. 2012; Lyons et al. 2013). The indirect effects of these qualitative changes on participation in out-of-home activities and travel remains open for discussion and further research. For instance, using entertainment apps may make people decide to accept longer travel times in order to reach more remote destinations.

A second category of apps in some way facilitates the participation in activities and travel. Examples are apps for wayfinding and public transport trip planning, but also apps for finding (and ex ante evaluating) activity locations such as restaurants, entertainment centres, museums etc. On the longer term, in the context of repeated behaviours, we can also regard travel and activity feedback apps as falling in this category. For instance, apps providing feedback on daily levels of physical activity (step counters), may influence travel mode choice (active vs. passive) or invoke activities with more physical activity. The direct effect on activity participation and travel is straightforward, in the sense that information about travel and activity conditions, as well as feedback on travel and activities, may influence activity choice, mode choice and location choice. To our knowledge, however, limited knowledge has been accumulated about the quantitative effect. Apps from this category may have an additional effect on activities and travel as they (like any App) can be used during activities and travel. For instance, one may look for the best bus connection to the final destination while making a train ride, or find the best restaurant while attending a meeting. Such App use may change the character of and attention given to travel and activities and affect their pleasantness and value. Again, such effects remain open for further research.

A third category of apps has a direct impact on activities and travel by substituting for activities on location. Common examples include banking or online shopping via a smartphone App. In line with the first hypotheses about the effect of ICTs on travel, the effect may be that trips to physical locations are suppressed. However, it is likely that the time saved by not visiting a physical facility, is used for additional trips and activities, that would not be made otherwise. Again, such apps can be used anywhere and anytime, including during travel and other activities, which may change the character of travel and out-of-home activities, affecting how they are appreciated.

A final category of apps provide their users with tools for communication with other App users. Such communication includes voice, photo, video and instant messaging. While such communication may include issues also addressed by other App categories (entertainment, information gathering), the set of communication apps available on current smartphone provides a unique situation of ‘connected presence’ (Licoppe and Smoreda 2006), implying that everyone in one’s social network can be contacted anywhere and anytime at almost no costs, to communicate about practical, important and trivial issues by a variety of modes. Communication may also invoke face-to-face contacts, leading to out-of-home activities and travel. Obviously, this state of connected presence extends to travel and out-of-home activities, with impact on how time is spent during travel and activities and how they are experienced.

The taxonomy bears implications for how the relationship between apps, travel and activities is conceptualized theoretically. Traditional travel theory suggests that activities and their associated travel are invoked by specific motives or needs. Apps may interfere with this process in multiple ways. In relation to motives and needs, a general need (e.g. for entertainment) may not only lead to an activity in a specific facility (e.g. a cinema), but also to using an App to fulfil the need (e.g. gaming). However, App use may also invoke a need leading to activity/travel, for instance if an App provides information about real life places worth a visit (e.g. an advertisement for an event via social media). Apps may facilitate travel and activities by supporting planning and logistics as discussed before. Furthermore, in line with studies of positive utility of travel time, apps may change the travel/activity experience if they are used during travel and activities, with implications for future activity and travel engagement. Finally, apps may be used to obtain objective feedback about one’s travel or activity behaviour.

Activity theory (AT) dates back to the work of Soviet psychologists and Marxist philosophers (see Vygotsky 1978), and was introduced to the western world in the 1980s by the Finnish psychologist Engeström (2014). AT aims at describing and understanding individuals’ mental processes (such as decision making) in relation to their physical environment by analysing practices in their real life context. The central notion is that consciousness (including decision making, classification and remembering) is not disconnected from everyday practice, but intrinsically tied to it. According to AT, our mental processes are determined by our daily practices and vice versa. Hence, if we are to understand human behaviour, we need to study it, according to AT, in connection with the tools associated with that behaviour, since these not only provide the means for engaging in the behaviour, but will also influence peoples’ attitudes and decision making preceding the behaviour. Based on these principles, AT has been applied frequently to the fields of for instance ICT supported learning (e.g. Lim and Hang 2003) and human computer interfaces (HCI) (e.g. Nardi 1996a, b). In the remainder of this section we will further introduce AT.

Activity systems

AT describes activities in a very general sense as a subject (a person) working on an object, mediated by tools, and aimed at a certain outcome. In addition, the activity is embedded in a social context, and is subject to certain rules. The activity together with the structure in which it is embedded is called the activity system (see Fig. 1). The subject is the person acting on the world, in our case the traveller using a smartphone App. The activity implies that the subject uses a tool to bring about a change to an object. Tools can be mental (e.g. using arithmetic to solve a maths problem), physical (e.g. a hammer) or virtual (e.g. software). Hasan and Kazlauskas (2014) distinguish between primary tools (physical), secondary tools (language, ideas, models) or tertiary (communities, contexts or environments). An important notion in AT is that tools mediate but are also mediated by their surroundings (Fjeld et al. 2002). That is, the tool allows the user to materialize internal mental activity into material artefacts (exteriorization). In reverse, contact with the physical surroundings and the social context stimulate mental activity (internalization), for instance handling increasingly more abstract concepts and objects.

Open image in new window

Tools are used to work on an object, that is changed in order to achieve an intended outcome. Like tools, an object can be material (e.g. a nail being hammered) or non-tangible (e.g. a Word document being modified, a bank account that is increased or a public opinion being influenced). The application of the tool to an object results in an outcome, which, if the activity system works right, was intended by the subject. AT supposes that activities are undertaken with a certain motive, which is to satisfy a need (Kaptelinin and Nardi 2006). However, activities may in AT have multiple motives. For instance, a cyclist cycling to work may do so to address various motives, such as getting to work timely, getting exercise and enjoying the landscape, which relate to different needs such as self-realisation, restoration and achievement. According to Kaptelinin and Nardi (2006), the combination of motives determines the way the activity is carried out. For instance, if the only motive of a traveller is to get to work on time, a trip by car may be the preferred option. However, if maintaining health and fitness is another motive, the car is not an appropriate tool, and the bicycle offers better opportunities for satisfying the underlying needs. The differences in needs and motives between people in relation to the same ‘objective’ activity (in this case commuting) implies that objects in AT are defined at an individual level and in different ways (Hasan and Kazluaskas 2014).

AT furthermore holds that activities are embedded in a social context, which involves rules, a community and division of labour. The community are other individuals, involved in activities that are related, but not identical to the activity carried out by the subject. The community may consist for instance of colleagues in a company with which one cooperates. The involvement of other people implies that activity systems are multi-voiced (Engestrom 1999): different people involved will have different interests, opinions and traditions, so that the object addressed in the activity system has different meanings for different involved people. This may be a source of tension, but also a driver of change in the activity system, leading to modifications and improvements. These different viewpoints by different parties in the activity system are related to the division of labour within the system. For instance, in a school teachers and pupils both work on the teaching activity (Lim and Hang 2003), but with clearly different tasks and responsibilities and interests.

Finally, the activity system is subject to rules. Rules can be more or less formal, and define how the subject should behave within the community. For instance, in the context of travel, formal rules could pertain to where and when travel modes are allowed to use (highways, cycling paths, pedestrian areas, use of an off-peak travel pass for public transport). Implicit rules are related to norms in the community regarding what is appropriate behaviour, such as using bicycle or public transport not being in accordance with a manager’s status.

Kwan (2007) argues that to understand the effects of ICTs on travel and activities, it is important to acknowledge that ICTs relieve spatio-temporal constraints and the role of distance on which traditional models of mobility behaviour (such as gravity models, four stage models or time geography) are based. In particular, the point based representation in these models of people and social networks is limiting when investigating the effects of ICTs on activities and travel. Kwan argues that alternative metaphors may help us to understand the effect of ICTs on travel. In this section we elaborate on the application of AT as such a metaphor to better understand the working and implications of apps. A particular feature of AT is that it is not based on notions of distance or proximity, and that it is not limited to activities and tools taking place in physical locations. As discussed, tools and objects in AT may be physical, mental or virtual. Therefore, AT is a potentially promising theory to study ICTs (here apps), acknowledging both the physical, mental and virtual levels in which apps work out.

The taxonomy outlined in “Apps, activities and travel: a taxonomy” section already provided first indications of the potential effects of smartphone apps. However, their exact effects depend on their use as tools as part of activities by subjects while working on a specific object, as suggested by AT. It should be noted that also within (sub)classes of this taxonomy, apps may have different effects, as their functionalities, but also their rules and communities and the motives of their use may differ. Below we present such specific analyses based on AT for apps from specific subclasses: a communication by instant messaging App (Whatsapp) and travel related (travel feedback) apps. As specific instances of their subclass, they illustrate how travel related and other effects of apps can be analysed using AT.

This paper introduced activity theory (AT) as a systematic way of investigating the effect of ICTs on (amongst others) travel behaviour. The basic assumption underlying AT is that tools (such as smartphone apps) are used by subjects (in this case the traveller) to perform an activity with a certain outcome, which is driven by motives that satisfy underlying needs. We propose AT as an alternate way of analysing the effect of ICTs on travel, which has as an important merit that it analyses in detail the practises of using a specific ICT device/tool, in relation to the user’s motives and embedded in a social context. Such a detailed analysis is considered a useful addition to the state of the art in research on ICTs and travel, which has often implicitly tended to treat ICT as a single, homogenous phenomenon, instead of recognizing the variety in ICT devices and tools, which has in recent years increased exponentially through the development of smartphone apps. In addition, studies of travel and ICTs have mostly focused on the quantitative effects of ICTs on travel, expressed in terms of substitution, complementarity, modification or neutrality, without focusing on the underlying mechanisms of such effects, the implications for both travel and other outcomes, and the embeddedness of travel and ICT use in people’s daily activities and social context (the latter of which may extend to the society as a whole). Hence, we agree with Schwanen (2015) that when studying the effect of apps, more attention is needed for the functioning of apps as more or less independent agents in networks of users, stakeholders and other (ICT) devices. AT offers a framework for carrying out such analyses, with an emphasis on the relationship between user and tool, embedded in a social context.

Exploratory analyses applying AT to specific apps (WhatsApp and Travel feedback apps) based on existing literature suggest that AT may be a useful tool for analysing apps (and ICTs in general) in a broader context. Relevant issues that came to the fore include the often polymotivational nature of using apps, implying that the travel related effect of apps partly depends on non-travel related motivations, such as financial or material gain or maintaining friendships. Another important implication of the analyses is that social structure differs strongly between apps. Whatsapp is used in an egalitarian community, with users facing and similar opportunities and constraints, and have more or less likeminded motivations. On the other hand, the community around travel feedback apps includes travellers as well as the consortium launching the travel feedback programme, which have different motives for using the App, which not necessarily coincide. Our AT based analysis also gave indications of contradictions in the AS of travel related apps. For instance, difference in motives of users and program consortia may lead to counter effective stimuli that make the App less attractive for users, and withhold non-users from adopting it. Finally, AT may offer a framework to understand the development in the practises of using tools as the outcome of contradictions, such as the strategies applied by some users to handle unwelcome incoming messages.

The reliance on existing studies in this analysis has been an obvious limitation. A particular disadvantage is that scientific literature may lag behind daily practice in a rapidly changing context such as smartphone apps. For instance, while the literature regarding Whatsapp focuses on the use of this App by groups of friends of younger age, which typically were early adopters of Whatsapp, our own observations and discussions with fellow researchers suggest that the use of this App is now much more widespread, and is used to coordinate family life, social ties and club and professional activities of many. Another limitation is the focus on dedicated travel feedback apps, thereby ignoring health apps that may also provide feedback on travel behaviour (e.g. Strava). As certain health apps are more widespread than dedicated travel feedback apps, and may be used with different motivations, it is important to address their use and effects in future research.

Based on this exploration, we feel that studies based on AT can be a useful complement to survey based quantitative studies of ICTs and travel. While in survey based studies, rather straightforward assumptions need to be made about the impact of ICTs on travel, AT allows one to investigate the relationships between actors in social networks who use ICTs/apps in much more detail and with more theoretical rigour. Importantly, studies based on AT can identify actors’ practices when using ICTs/apps and increase understanding of how the effects of ICT use on travel are dependent on more general objectives of using ICTs and how these are embedded in daily practices.

Studies of apps based on AT would typically be qualitative in nature, questioning subjects in activity systems about their motives and practices when using the App, the rules applying to the use of the App, the involved community and contradictions occurring in the activity system. Given the importance of apps as independent agents, research based on AT should also focus on the App, its linkages with other agents (either software, databases or humans), and the interaction (exchange of messages and information) with these agents. To address the relationship with travel and activities, such an approach may be complemented with more traditional approaches such as activity and travel diaries and diaries of App use.

In practice, studies based on AT could for instance be used to identify useful classifications of ICTs/apps, based on the objectives, practices and embeddedness in community, which could provide a basis for quantitative studies and may refine the taxonomy presented in this paper. In addition, studies based on AT could be used to identify differences in practices in the use of ICTs/apps, which might be related to differences in preferences, lifestyle or socio-economic position. Such findings may provide a basis for segmentation analyses in quantitative studies. Another implication of such an approach could be the identification of inequalities in the use of apps and the extent to which people have access to them and profit from them, which is indicative of equity issues. In the context of the relationship between social networks and travel, which has been an emerging area over the past decades, studies of ICTs/apps based on AT can provide important insights into the role of ICTs in maintaining social relationships and how travel and activities emerge in the context of social relationships.