24
However, what is also evident from the reviewed literature is that it suffers from a lack of conceptual clarification which a) hinders the effective accumulation of knowledge b) neglects important nuances, which make it hard to compare findings as well as set guidance for future research. Further, extant empirical findings at the individual-specific level lack insight into the impact when incorporated into a product (rather than disclosed on an external platform such as Facebook). Finally, extant research is dominated by large-scale experiments all of which apply a business-perspective. Building on Freelon's (2014) recent call for more research that applies a user-perspective on such digital traces of behavior, I argue that this lack of user-perspective risks faulty design and untapped opportunities.
25
(Watson, 1913), and that rewards and punishments are important factors for learning through operant conditioning (Skinner, 1963). Put simply, the focus was on learning as an effect of one’s own direct experiences (Bandura, 1977). Bandura’s stance on learning differs from that of behaviorism in the sense that he recognizes the important role of the social environment.
Specifically, using the illustrative examples of teaching children to swim or adolescents to drive cars he has famously claimed that learning would be a hazardous affair if all learning was to take place through the direct, personal experience of trial and error (Bandura, 1977). Rather, he posits that humans are capable of learning through the observations of others’ behaviors.
Although Bandura in this manner differs from the behaviorists of the early 20th century, he does not dismiss the learning impact of rewards and punishments as emphasized in operant conditioning (Skinner, 1963). In fact, in Bandura’s view, learning through the observation of others is strengthened if the outcome of the modeled behavior is also observable. For example, observing that one’s siblings are rewarded for good manners can increase the observer’s learning of good manners. Bandura’s theory of observational learning is in essence a theory of learning. However, given its foundational emphasis on the social environment in making changes in individual behavior, we can also view it as a theory of social influence with high relevance for design.
From an observational learning perspective products whose use are high in observability (e.g.
clothing, cars), are generally more prone to influence others than those products whose use cannot be observed that easily (e.g. shampoo, bed linen) (Bandura, 1986; Rogers, 2003). This is a crucial point for this dissertation, as digitization transforms a wide range of products, or more precisely their usage, from being low in observability to being highly observable, at least in an indirect way (e.g. not observing actual music listening behavior, but rather digitally observing a story or trace of a person listening to music). Accordingly, the main theoretical perspective applied in the extant literature about digitally observable behaviors has been that of
observational learning (e.g. Chen et al., 2011) and the related phenomenon of informational cascades (Duan et al., 2009; Thies et al. 2016). This PhD dissertation is no exception. The main theoretical motivation for embarking on this project has been the premise that observing the behaviors of others can influence one’s own attitudes and behaviors.
Consequently, in this project, and specifically for the online experiment set to answer RQ 2, I hypothesize that the inclusion of social information in online products and services – i.e.
information about other users’ product-related behaviors and opinions - can positively affect
26
potential users’ attitude towards an online service. To give further nuance to this hypothesis, building upon Bandura’s (1986) observational learning theory and social impact theory (Latané, 1981), this research recognizes that a number of factors impact how effective that influence is. These are: 1) who the observed person or the ‘model’ is 2) the number of people modeling, and 3) whether the outcome of the modeled behavior is observable. These are expanded below.
First, observing the actions of or receiving recommendations from people who are
knowledgeable about a specific topic, or in some way opinion leaders, generally has a stronger impact than observing random people’s behavior (Bandura, 1986; Bikhchandani et al., 1998, Latané, 1981). Consequently, I hypothesize that the product-category relevance of the
modelling person matters, and that the behaviors of category-specific influentials will have a stronger impact on observers than those of random individuals. Second, social impact theory posits that a greater number of influencers increases the likelihood that an individual will be influenced (Latané, 1981). However, from the field of economics we find theoretical arguments that explain how observing the actions of as few as two other people can serve as a turning point where one starts disregarding their own information and follows the behavior of others (Bikhchandani et al., 1998). Consequently, I hypothesize that the inclusion of social information from two friends is enough to create an impact but that a higher number of friends will have a larger impact on observers. Third, Bandura (1986) argues that the effect is generally amplified when the observer can observe the consequence of a behavior. For example, in the case of a behavior where the model is rewarded for that behavior there will be more impact on the observer than without seeing the reward outcome. Turning to the online sphere, we find that people perform a vast number of product-related behaviors every single day, booking a hotel, ordering groceries online, streaming music and the like, many of which can be made
observable online to others. In Bandura’s terminology, these can be regarded as mere behaviors without observable outcome. On the contrary, if the user provides some kind of evaluation, such as a review, adding a ‘like’, a rating, or some other indication of how the user experiences that event, one must assume that this opinion was formed on the basis of an experience (a behavior) with the product/service. Accordingly, the evaluation implicitly tells us about a person’s behavior with that product/service, and this behavior can thus be regarded as observable behavior with observable outcome. Contrary to the popular saying that “actions speak louder than words” I thus hypothesize that opinions will have stronger impact than behaviors. On a final note, it is worth mentioning that Bandura also recognized the importance
27
of observing one’s own behaviors to improve performance, such as with improving in sport (Bandura, 1986). This is an important point, particularly in terms of Paper 5 where the impact on the individual when faced with their own music listening behaviors is also discussed.
2.3.2 Self-determination theory
As demonstrated in the literature review, extant literature has been focused on identifying the impact of behavior-based information on consumer decision-making and behavior and have done so mainly from an observational learning perspective, where the point of departure is the business and the impact on the business. Only few studies take a user perspective, for example Luarn, Yang, and Chiu (2015) who identified user motivations to share their behavioral
information on location-based services. Moreover, the current knowledge base has primarily taken a functional view of behavior-based information. Extant research assumes that the user interprets and uses the behavior-based information as cues to what is popular (hence the term popularity information) and thus make either better or more effortless decisions in line with observational learning theory. While this approach has its benefits and valid theoretical rationale, I argue that in order to truly shed light on how to best make use of behavior-based information in product design, there is a need to go beyond this purely functional view and seek knowledge from the point of the user: how users actually interpret this rather sterilized information, and how it is used in practice. Simply put, design of solutions with behavior-based information must also take into account the user needs as well as concerns in order to take full advantage of the potential of behavior-based information. To support that end, the self-determination theory (SDT) is introduced. Specifically, I draw on a sub-theory of SDT, namely the theory of basic psychological needs. It posits that all human beings possess three basic psychological needs (Ryan & Deci, 2017). Each of these must be nurtured and satisfied for human beings to thrive. If deprived of this need satisfaction, the well-being of an individual will decrease, just as with a human being who is deprived of basic physiological needs such as sleep, food, and water. The three needs are: autonomy, competence and belonging.
First, autonomy refers to circumstances where “one’s behaviors are self-endorsed, or congruent with one’s authentic interests and values” (Ryan & Deci, 2017, p. 10). It has to do with freedom of decision and the personal meaningfulness of the task at hand (Sailer, Hense, Mayr, & Mandl, 2017). Competence refers to the basic need to master something, be it sports, cooking, or math, and generally operate efficiently within important life contexts (Ryan & Deci, 2017). It is what drives the golf player to practice small details again and again and the toddler to insist on
28
doing things themselves instead of having their parents help. Lastly, the need for relatedness describes human being’s need for feeling cared for by others and to belong to social groups. An important component of relatedness is to give and contribute to significant others (Ryan &
Deci, 2017). This explains, for example, the joy of giving personals gifts to family members for Christmas, the gifts becoming a signifier of close relationships.
SDT, and its three basic psychological needs, is especially well-suited for design topics as it emphasizes the active role of the environment in satisfying these basic psychological needs (Ryan & Deci, 2017; Sheldon, Abad, & Hinsch, 2011). Accordingly, I argue that by shedding light on how behavior-based information within a digital service can satisfy basic psychological needs, designers will be better equipped to design services that retain and engage users
through satisfaction of these human needs. As such, SDT is used to discuss the empirical findings in Paper 5 based on a hypothesis that behavior-based information affords users more than the functional affordance of guidance towards popular content, as prescribed by extant literature. To the best of my knowledge, SDT has not previously been applied to the case of behavior-based information in digital products and services. However, it has been successfully applied to related areas such as online gaming (Ryan, Rigby, & Przybylski, 2006), Facebook use (Sheldon et al., 2011), the use of gamification design elements (Sailer et al., 2017) and in Q&A communities (Li, Huang, & Cavusoglu, 2012). Furthermore, while most often applied in
quantitative – often experimental – research, it is has also proved to be a well-suited theoretical lens for qualitative research (Garn, Matthews, & Jolly, 2010; Ryan & Niemiec, 2009).