Keywords

1 Introduction

The past decade has witnessed a remarkable and unprecedented increase in e-commerce marketplace. This increase is represented by not only the skyrocketing sales numbers but also the surge in feature/function innovation and format/channel richness that are driven by the constant advancement in technology. It gives consumers more choice, freedom, and better experience. And it also opens more research opportunities for us to better understand consumer behaviors. As such, multiple aspects of the design of the online shopping environment have been extensively examined. They predominantly focus on the interaction between consumers and the online content. Motivated by the recent trending imagination of the near-future virtual reality shopping design, this study takes a rather novel perspective by looking at the influence of users’ actual physical behavior (bodily movement) during online shopping. We focus on the interaction between consumers and the technical devices. We believe that it is an easily-ignored but rather important aspect of online consumer behaviors.

It has been widely acknowledged that human bodily movement can reflect what people think as well as how they feel [1]. For example, a relaxed and open gesture delivers the signal of liking others [2], yet a stance of leaning backward indicates a negative feeling towards the object [3]. There also exists a stream of research describing attitude as a collection of motor activities [4]. Most researchers believe that postures capture human subjective evaluations of objects, and there exists a strong linkage between bodily movement and evaluation.

Further researches have also revealed that not only does this linkage exist but also that it is bi-directional, which means bodily movement can also in return influence the formation of evaluation and attitude. For instance, Casasanto [5] has found that right-handed individuals are more likely to associate rightward space with positivity but leftward space with negativity and vice versa for left-handed individuals. Such finding contains useful implications for marketers, with the intensive exposure to performing bodily movement using the dominant hand, right-handed (left-handed) consumers will be more likely to render a more positive evaluation when the product is placed in the rightward (leftward) space than leftward (rightward).

Following this branch of research, this study aims to investigate how bodily movement influences consumers’ behaviors. Most existing literatures address the relationship between bodily movement and consumer behaviors in offline settings. For instance, arm flexion and extension have differential effects on attitude as arm flexion is more closely coupled with acquisition or consumption of the desired stimuli, leading to the association between arm flexion and positive attitude [6]. However, understanding how the influence of bodily movement transfer to the online context could also generate rich theoretical and practical implications. Without the real touch feelings and detailed assessment of product quality, how to let consumers have positive evaluation and deep impression is crucial. Despite product presentation formats, content details, reviews, and recommendations online, the interaction between consumers and the technological device that presents the production may also play an important role. Thus, the scope of this study is narrowed down to only investigating how the bodily movement in IT devices will influence consumers’ behaviors.

More specifically, in this study, we are interested in how does finger movement influence online consumer behaviors when browsing shopping website using a touch device. According to Pew Research Center [7], a fact bank that provides information on social issues, computer ownership percentage among younger adults below 30-year-old has reached 78%. And for more convenient carry and use, consumers begin to abandon the habit of using mouse and turn to use trackpad embedded in the laptop. There are two types of trackpad orientation: natural and reversed. The natural scrolling orientation is that the website content flows in the same way as the movement direction of fingers, while the reversed scrolling orientation is inverted. Therefore, when the consumer scrolls the website from top to bottom to view the product information online, the conditions of figures are different. In the natural orientation, scrolling down the website makes the figures in the states of extension, yet flexion for reversed orientation.

Hence, this study aims to specifically investigate the impacts of the two trackpad orientations on consumers’ evaluation and memory towards recommended products in social media. This study sheds lights into the interaction between physical movement and product evaluation, and fills in the research gap of lacking attention on the interaction between online consumers and technical devices. In order to establish the relationship between trackpad orientation and product evaluation as well as memory, information processing fluency is introduced as a mediator. We argue that natural orientation will produce a higher level of fluency, as a result of the fit between cognition and motor activity. Then the higher level of fluency will lead to a more positive evaluation but worse memory towards the product. Moreover, this study examines the role of shopping motivation, i.e. browsing-oriented shopping and goal-oriented shopping in moderating the effect of trackpad orientations. Thus, we further hypothesize that the superior (inferior) effect of fluency on evaluation (memory) for natural orientation over reversed orientation will be weakened in the condition of goal-oriented shopping.

The remainder of this study is organized as following. The next section reviews prior literature and provides theoretical foundations. The research model as well as hypotheses are developed, followed by experimental design procedures. Last but not least, conclusions including contributions and future work will be provided.

2 Theoretical Development

2.1 Embodied Cognition

When mind meets body, evaluations and judgments are subjected to be changed. For example, physical needs such as hunger can influence the person and transfer the person’s attention to food; a smile may trigger the mental image of happiness and make the person to feel delightful when upset. Mind and body are tightly connected, in which way that cognition influences body movement, and body movement influences cognition [8]. Thus, instead of only assuming that perception leads to motor priming, researchers in embodied cognition (EC) argue a reversed causal chain – motor simulation actives and influences the perception [9]. In EC literature, there are two streams of research: mild embodiment and radical embodiment, which we will only focus on the former one.

One representative theory proposed in mild embodiment is perceptual symbols theory (PST) [10]. PST states that knowledge and perceptions are not developed from vacuum. Rather, they are originated from sensory and motor contexts of their occurrences. It is the sensorimotor information that helps to establish the conceptual representations. Thus, during the formation of conceptual representations, multimodal sensing including perceptual, motor and introspective states are recorded. Later when similar sensing is processed, the conceptual representations are reactivated.

2.2 Embodied Cognition and Processing Fluency

Researchers have demonstrated that actions are builders of how knowledge is represented, thus the motor cues will facilitate or inhibit decision making during information processing [11, 12]. The processing of information can be affected by many factors such as processing speed, accuracy and consistency [13]. All these factors combined lead to an experience of processing ease, which is defined as fluency [14].

When viewing social media websites, the multimode of sensing inputs are finger and the visual input from computer screen. In embodied research, a typical study is about how visual up and down affects the emotional valence. Implicit learning from daily associations between physical experiences and emotional states co-occurred, such as people are more likely to stand straight and tall when feeling proud but slouch when feeling depressed, it is easily to conclude that positive is up and negative is down [5]. Applying this metaphor to social media context, when consumers view the website from top to bottom, they will scroll the screen to let the screen move upwards. The upward motion may induce positivity for all consumers. However, different types of finger movements may cause distinct emotional valence, which will be congruent (incongruent) with the feelings caused by visual inputs, resulting in different processing fluencies. There is a lack of research in embodied cognition of finger, but we can refer to the arm motions to have a similar inference. When the fingers are in the state of extension, the state is more associated with daily experiences of approaching objects. On the contrary, when the fingers are in the state of flexion, it may remind consumers of being afraid of losing something. Thus, the finger extension may induce positivity, while negativity for finger flexion.

In the natural orientation, on the one hand, visual upward movement of website produces positivity; on the other hand, finger extension produces the positive feeling as well. Whereas, in the reversed orientation, finger flexion causes negative feeling, which is contracted with the positivity of visual sensing. The processing of information is considered as more fluent if two cognitions are congruent. In addition, natural orientation enables the movement directions of finger and visual input to be consistent, which may also enhance the feeling of fluency. Hence, we propose the following hypothesis:

Hypothesis 1.

Compared to reversed orientation, natural orientation will produce a higher level of processing fluency.

2.3 Processing Fluency, Attitude and Memory

Multiple theoretical notions review that high fluency is reliably associated with more positive evaluations [13]. Regardless of the exposure repetition number, other variables that facilitate fluent processing also tend to increase liking. When the trackpad orientation is natural, processing fluency is expected to be higher, which will generate a more positive evaluation towards the product displayed online. Thus, we hypothesize as the following:

Hypothesis 2.1.

The natural trackpad orientation will lead to a more positive evaluation towards the product via a higher level of processing fluency.

A more interesting relationship happens between processing fluency and memory. We argue that there is a twist here: though natural orientation gives rise to a positive attitude, it leaves a worse memory. Traditionally, researchers believe that a higher level of fluency will yield a better memory. However, in the social media context, this may not always apply. Usually the presentation content of product is not difficult to understand, i.e., it does not require a high cognitive load to fully digest the information, a relative lower fluency will not induce problems with understanding and memorizing the product. Instead, a lower level of processing fluency is possible to result in a better memory. Dual-process theories [15, 16] describe that there are two reasoning systems: system one is quick, effortless, and intuitive, yet system two is slow, effortful and analytical. The two systems are all related to human judgments, but which system dominates the process largely depends on the ease of difficult with which information comes to mind [17]. Alter et al. [18] verified that when the processing fluency was higher, system one would guide judgment; System two was operated to guide judgment when participants were experiencing disfluency. Therefore, the feeling of disfluency will activate analytical reasoning, pushing consumers to focus on the details of product and put more cognitive efforts to evaluate. When consumers viewing website using natural scrolling trackpad, the fluency may activate system one in a way that consumers tend to adopt holistic and quick viewing of the product information. As for reversed trackpad, the disfluency will trigger analytical viewing, which will leave a deeper impression of the product to consumers. Hence, we hypothesize the following:

Hypothesis 2.2.

The reversed trackpad orientation will lead to a better memory towards the product via a higher level of processing fluency.

2.4 The Moderating Role of Shopping Motivation

Shopping motivations are categorized as two types: goal-oriented and browsing-oriented. Goal-oriented consumers have a targeted object in mind, and are more likely to process information of product in details in order to match their goal. Therefore, systematic processing is activated. Contrarily, browsing-oriented consumers have no specific goals in their minds. For example, large group of people in weibo like viewing products recommended by stars in their spare time. The main goal of browsing websites is to pursuit enjoyment [19]. Hence, they view product information more quickly and pay less attention to the details of product, employing heuristic processing strategy. In the previous section, we hypothesized that natural orientation will lead to a positive evaluation but worse memory due to fluency. But when consumers are goal-oriented, the systematic processing will neutralize the tendency of holistic viewing caused by fluency. Essentially, even with natural trackpad orientation, consumers will still activate systematic information processing to evaluate product. In such case, evaluations towards the product will more depend on the assessment of the product itself, rather than the feeling of fluency. Hence, the positive attitude may be attenuated if the product viewed does not match consumer’s goal. In terms of memory, since consumers will analyze the information of product naturally when they are using systematic processing strategy, so they will have a deeper impression regardless of the orientation of trackpad. In short, when consumers are goal-oriented when shopping online, systematic processing will dominate and moderate the effect of fluency. Stated formally, we propose the following hypotheses:

Hypothesis 3.1.

The superior effect of fluency on product evaluation for natural orientation over reserved orientation will be weakened in the condition of goal oriented shopping motivation.

Hypothesis 3.2.

The inferior effect of fluency on memory for natural orientation over reserved orientation will be weakened in the condition of goal oriented shopping motivation.

In summary, we argue that there is a fit between the emotion induced by finger motion and visual information, which helps to form a sense of processing fluency when the trackpad orientation is natural. Consequently, the processing fluency leads to a more positive product evaluation but worse memory towards the product. Finally, the shopping motivation is expected to moderate the effects between fluency and evaluation as well as memory. The theoretical framework is shown in Fig. 1.

Fig. 1.
figure 1

Theoretical framework

Full size image

3 Research Methodology

The relationship between trackpad orientation and product evaluation as well as memory is designed to be tested in three laboratory experiments. Study 1 tests the main effect of trackpad orientation on evaluation and memory. Study 2 tests the underlying processes and is expected to reveal that it is processing fluency that mediates the effect between trackpad orientation and attitude and memory. Combining the results of study 1 and 2, it should be possible to give some insights to the chain from trackpad orientation to fluency, and then to attitude and memory. Finally, study 3 examines the moderating effect of shopping motivation. Across the experiments, various products were used to generalize the results, ruling out the possibility of results validity being affected by product types.

The goal of study 1 is to test the main idea that different trackpad orientations have differential influences on product evaluation and memory towards the product viewed in social media. It is expected that natural orientation, compared to reversed orientation, participants will have a better evaluation towards the product, but a worse memory. When being recruited, participants are required to complete a short survey with demographic information, the laptop brand and experiences with using trackpad. Only participants using Apple laptops will be accepted. The laptop from Apple is chosen because most students who are using this device prefer to use trackpad than mouse. In addition, the design of trackpad of Apple laptops is more sensitive and comfortable. Also, the orientation of trackpad can be freely set as natural or reversed based on users’ own preferences. Upon arriving at the lab, participants are randomly assigned to two groups and instructed to perform two tasks. All participants are provided with MacBook. In the natural condition, the trackpad orientation is set as natural, and similarly in the reversed group, the trackpad orientation is set as reversed. As participants may have their own habits of using trackpad, obvious sense of disfluency will occur if the trackpad orientation in experiment is contradicted with their habits, which will contaminate the interval validity of this study. In order to rule out this possibility, the first task aims to let participants get used to trackpad orientation in experiments. Basically the first task is to keep participants viewing website, using trackpad as much as possible. For example, the task can be viewing news website and providing feedbacks on the content or design. After an hour usage of trackpad, participants are given five minutes to relax. After finishing the first task, participants are required to proceed to the second task. A simple social media website with product recommendation is shown and there are five products with detailed information shown for each participant (Appendix 1). All five products are selected from various categories, and the sequences for five products displaying on the screen are counter balanced. The format of Amazon is chosen, and the product pages are usually in the waterfall format, meaning that the page lasts long, which guarantees that consumers will scroll down several times in order to see all the information listed. After this, participants are instructed to complete a post-experiment survey (Appendix 2). The dependent variable product evaluation is measured by a six-item and nine-point scale [20]. The participants will be presented with the image of the first product, and then rate from six aspects on nine-point scales (1 = bad/dislike/unfavorable/not useful/undesirable/low-quality, 9 = good/like/favorable/useful/desirable/high-quality). Then participants will be presented the second product image and rated again, until the last one. The sequences of products presented will be counter balanced as well. For memory test, we have six memory questions for each product, asking its detailed information presented in the websites but not revealed in the images shown to participants.

Study 1 has several limitations that should be addressed by study 2. First, study 1 does not reveal the underlying mechanism between trackpad orientation and attitude and memory. Thus, study 2 is carried out to confirm the mediating effect of processing fluency. Second, mood will also influence the evaluation towards product, and the earlier exposure of certain product will affect the accuracy of memory results. Thus in study 2, we will measure the general mood state to ensure the validity of product, and ask participants whether they have seen same or similar product information before. After the first two tasks, participants are instructed to fill in a survey. Apart from the original questions in the survey, participants are asked to rate how fluent they perceived the website viewing procedure with the use of trackpad. Moreover, it is better to capture a more objective measure of processing fluency – the amount of time each participant takes to finish the website viewing for three products. Thus, another question is added to know whether there is a product that participants are particular interested in, which may interfere with the total duration on websites (Appendix 3). Besides the exact same two tasks (we choose different products displayed to prevent same participant memory or information contamination from Study 1), participants have two additional seemingly independent tasks after two weeks. The aim of the third and fourth tasks is to capture the emotion valance induced by finger movement and website movement as the processing fluency is predicted to be caused by emotion congruence. Two weeks later in the third task, participants are asked to rate their emotion valance after doing figure flexion and extension respectively. Similarly, in the fourth task, participants rate after seeing website moving up and down. The temporal separation is to prevent participants linking figure movements to the processing fluency questions in the survey, and then give fake answers due to demanding effect. They will be asked whether they have linked the two experiments together and guessed the true aim, and data will be removed if the answer is yes.

Study 3 is a 2 (trackpad orientation: natural vs. reversed) * 2 (shopping motivation: goal-oriented vs. browsing-oriented) experiment design. Same as in Study 1 and 2, participants are randomly distributed into one of the four conditions. The shopping motivation is manipulated by giving participants different tasks. For goal-oriented groups, participants are given a task of finding a birthday gift for friend, by knowing some function requirements from friend; in browsing-oriented groups, participants are told to browse websites naturally as usual to spend their leisure time. The experiment procedures are same as study 1: participants have same two tasks and fill in similar post-experiment surveys.

4 Conclusion

The study makes several theoretical contributions to different streams of research. First, it contributes to the knowledge body of embodied cognition literature. Prior research mainly focuses on the embodied cognition in offline setting. For example, researcher has studied how gestures changed thought [21]. However, few literatures really give insights into how gestures change behavior in online context. This study extends understanding of the relationship between motor activity and cognition by demonstrating how trackpad orientation and finger movement influence consumers’ product evaluation and memory when browsing social media websites. It is shown that different finger movement can lead to different levels of processing fluency, which will affect evaluations of the product.

The study also contributes to the research stream of e-commerce. With most research stressing on online reviews, presentation formats and marketing strategies, this proposal creates a new angle of improving consumers’ attitude towards the products. The research also examines the moderating effect of shopping motivation on attitude and memory. Thus, under the e-commerce context, although finger movement will influence consumers’ evaluations, how much the evaluations and memory will be influences depends on shopping motivations. It is argued here that when consumers are browsing-oriented, natural trackpad orientation will result in positive attitude but reversed trackpad orientation will lead to a better memory. When consumers are goal-oriented, the differences will be weakened.

This research idea also has practical implications. It suggests that when designing the future highly interactive user viewing experience, social media networks should not only focus on designing the content but also designing how consumer physically interact with the content. With the current trend, future technologies and innovations should only make the physical device more and more interactive.

To point some future work directions for this research idea, first, one limitation of this study is that the emotions or feelings of fluency are measured by self-report surveys. Although viewing duration is added in order to support the fluency results, there is still a possibility that the results are biased. Hence fMRI is suggested to detect the differences between finger extension and flexion, as the brain activity level should be different for the two movement. Second, a more popular way to do online shopping now is to shop on mobile devices such as phones and pads. The most common interface of phones and pads are touchscreens, and there are different finger movements or gestures that can be recognized on the touchscreens. Thus, one possible future direction is to study how finger movements on touchscreen will change consumers’ evaluations.