A conceptual model for future wear-able augmented reality system: Influence user behaviour into choos-ing stairs over elevator

(1)

Examination Thesis, Computer science

15 Credits, Bachelor’s Degree

A conceptual model for future

weara-ble augmented reality system:

Influence user behaviour into

choos-ing stairs over elevator

En konceptuell modell för framtida bärbara augmented reality

system: Påverka användarens beteende till att välja trappor

över hiss

Bachelor of Information Architecture, 180 Credits 27 August 2019

Examiner: Alberto Enrique Alvarez Uribe Supervisor: Thomas Pederson

Authors: Martin Bui ( martin.bui@hotmail.com ) Marika Moreau ( 89marika.moreau@gmail.com )

(2)

Abstract

We set out to find connections between different areas of interest in a persuasive augmented reality system, in an effort to change behaviour. To limit the scope of the study we chose to create a scenario, namely persuading users to take the stairs and not the elevator when trying to reach higher floors of a building. During a thorough literature study, we found Cialdini’s work about influencing human behaviour and Fogg’s behavioural model, which describes why we make the choices as we do. A small semi-structured interview was con-ducted on respondents that was using the stairs and elevator in a building to verify these findings, which our analysis did. We then found information and methods from gamification and color theory that was applicable in order to change or support these behaviours. Our predetermined scenario and the theoretical framework were combined into a conceptual model. The conceptual model includes what we call “Vision scenario”; It is a collection of manipulated images, of how the methods could be visualized through the augmented reality glasses when used. Persuasive technology systems come with some ethical dilemmas, which are discussed during the latter part of our thesis. We concluded that there are connections to be made between these areas of interest. However, further research is needed in order to improve our conceptual model. We as Authors are aware that there is much to be explored and connections of other areas that could be made in our conceptual model, but the time and our previous knowledge was limited.

(3)

1. Introduction

In our modern society we as individuals are constantly bombarded with information and are exposed to information overload (Heylighen, 2002). The human brain can only process a limited amount of information at a time (Chen & Choi, 2019). Our attention is drawn to certain types of information and stimuli (Moran & Desimone, 1985). This kind of mecha-nism are evolutionary and we as individual has inherit those mechamecha-nism in order to recog-nize important and life-threatening situation (Öhman & Mineka, 2001). It has become an issue for individuals to focus on important aspect of life like following their goals such as living a more active lifestyle without getting disturbed by all the information surrounded by us. That’s why we as individual often use devices like wearables and smartphones that can support our goals that we have set up for ourselves and our behaviours (Hänsel, Wilde, Haddadi, & Alomainy, 2015).

It’s not possible to avoid all the information and stimuli around us, but we think that there are better ways to support individual’s goals and behaviour. In this paper we describe a potential solution that involves a future wearable Augmented Reality glasses (AR-glasses) that could be carried by individuals who need support and guidance in their everyday life. The reason is to aid the user in a way that makes it easier for the user to achieve their goals. In our case, the future AR-system will support user’s goals by visually persuading them into actions that leads them towards their goal that they have set up for themselves. In order to keep the scope, narrow as possible we decided to focus only on the human visual sense like visual cues and persuasion techniques. While conducting our literature study we found valuable information in psychology, influencing and persuasive design about human behaviour and motivation. Since the AR-technology can’t perform real-world manipulation in real-time that our future scenario demands, it became difficult to find usable methods. We therefore decided to try to find connections between the discovered techniques, meth-ods and human psychology, in the hope of creating our own conceptual model that could support developers when developing future wearable AR-technology. In order to do so we set out to investigate the areas of psychology, gamification, color theory, storytelling and gaze guidance.

The reason why we choose to focus on AR-wearables glasses was because, devices like mobile phones and wearables in general has become widely popular in fitness tracking and supporting human behaviour and establishing goals (Buchem, Merceron, Kreutel, Haesner, & Steinert, 2015). AR-glasses are getting more attention since this kind of way of interac-tion creates new possibilities in wide range of areas to be explored, and its predicted that AR-glasses & technology are expected to grow rapidly in sales by 31 billion USD until year 2023 (Market Research Future, 2019). The AR-glasses are also getting more advanced and the components are getting smaller, this can be seen at the end of this paper (appendix B)

(6)

2

where we highlight companies that are pushing the wearable AR-technology forward, see

(Figure 1.1, 1.2, 1.3).

In summary, we created a conceptual model for development of future AR-technology based on earlier research, theories, methods and frameworks from various areas that we think would work together. The motive behind our conceptual model is to persuade users into making decisions that guide and support them toward their goals. We will present pictures of a real-world environment that will represent our scenarios, then apply our conceptual model on these scenarios to realize our vision for our readers and developers. The scenarios together with our conceptual model will show how the future wearable AR-glasses interface may look like in the near future when the AR-system is trying to persuade the user.

2. Research questions

1. To what degree could X be used to help individuals achieve behaviour change, using an Augmented-Reality system, despite habitual everyday action where X is:

a. methods from influencing

b. methods from persuasive technology c. methods from color theory

2. What are the ethical concerns to inducing a behavioural change with augmented reality?

3. Method

In the beginning of our study we started with a broad literature study, using keywords regarding: “Persuasive technology, Influencing behaviour, Storytelling methods in AR, Wearables, AR glasses, Gaze-guidance, Subtle gaze-guidance, Visual storytelling methods, Gamification, Interactive storytelling, Motivation, Behaviourism, Mobile healthcare, Color and Persuasion“. This gave us a sense of what the current AR-technology is capable of. We then decided on a scenario (real-world system) we thought would be possible to visu-alize, namely persuading users to take the stairs instead of the elevator. The real-world system (the route and choices made by the user) was mapped out through observation and then visualized in photographs.

The literature study was an iterative process where we went back and forth between the scenario to find methods applicable to the scenario. All methods that were found and used in our conceptual model were gathered and written down in the section “theoretical frame-work” for future updates and iterations.

It was difficult to find concrete methods about how and what to visually alter the user’s field of view in real time, in order to influence their path or gaze-direction. Much of our

(7)

3

focus was drawn towards psychology to understand the reason behind human behaviour and why we perform certain task or action. Cialidni’s and Fogg’s models and methods are well established, proven and used by other researchers. Our conceptual model deals with visual methods only, and we do not want to distract future users, of these kinds of systems, so our techniques need to be non-disruptive as possible. Color theory seems to fit those criteria well, and so does gamification. We use techniques from subtle-gaze guidance in order to make sure that users are looking at the right direction. Since our paper is based on a future vision about how the AR-glasses may work in the future, we chose to gather knowledge by observing how people behave in a real-world environment that later became our scenario. We also choose to conduct semi-structured interview to gather in-depth knowledge from the 40 respondents that we observed. The answers provided us new knowledge of information that was not considered when designing the question-form for the interviews. One issue with this type of interview is that the quality of answers is correlated with the interviewers interviewing skills (Braun & Clarke, 2006). It is important for the interviewer to be versed in the subject, in order to avoid asking the wrong questions or in the wrong way. A thematic analysis was used to analyse the data from the interviews. With this method, all the data was coded and grouped into categories which makes the data manageable. This analysis method is widely used in qual-itative research and are not tied to a specific research area which makes it easy to use.

4. Methodology

Theoretical framework

A theoretical framework is a collection of previously written literature or conducted studies surrounding a specific topic, which are evaluated, analysed and validated through compar-ison to each other (Ranjit, 2011). Information from widely differencing areas of theory and from different perspectives is sorted into themes and theories and should all extend knowledge applicable to a given situation or problem. This process will highlight discrep-ancies and gaps of information and research that still needs to be conducted. Conducting a literature study and developing a theoretical framework is an iterative process that can be extended for an eternity. The process of iterations was stopped after we had found at least one appropriate method from each subject area.

Conceptual framework

A conceptual framework contains the section(s) of the theoretical framework that becomes the focus of your study (Ranjit, 2011). It works as a reminder to make sure that the correct aspects are discussed during and after the process by everyone involved.

(8)

4

Semi-structured interviews

Semi-structured interviews are widely used for data collection in social science (Bradford & Fin, 2012) and make it possible to gather in-depth information of peoples’ experience and understandings of the reality (Braun & Clarke, 2006) (Kumar, 2019).

Thematic Analysis

The thematic analysis method is used to identify patterns, meanings and value across big clustered dataset. The advantage of using this method is that it is flexible and not tied to a specific research area which makes it easy to use for a novice researcher to use. The method also allows us as writers to be flexible in our choices of theoretical framework Braun & Clarke (2006) describe thematic method in a six-phase guide, which are describe right below this section.

Braun & Clarke’s six phase guide for thematic analysis: Phase 1. Get familiarised with the data

- Reading and re-reading the entire data to get more familiar with the data and transcribe verbal data.

Phase 2. Generating codes

- This phase is all about grouping all similar pieces of the data into small chunks of codes. The code should represent a general meaning of the similar pieces of data.

Phase 3. Sort all code into sub-themes.

- The codes should generate patterns which will become sub-themes that gives the code a broader meaning.

Phase 4. Reviewing and refine themes

- Reviewing the themes by modify, develop, or remove themes. See if the data forms a coherent pattern and remove data that does not fits into the patterns. Remove overlapping themes and ensure that the themes make sense to the re-search questions.

Phase 5. Defining and naming themes

- Create main-themes and connect the sub-themes with the main-themes. The themes should fit into overall narrative. The goal is to identify what the main themes means and how they relate to each other.

(9)

5 Phase 6. Producing report

- Write a report that explains the gathered data in an interesting and understand-able way. Consider your audience when writhing the report and be logical, co-herent, short and concise.

5. Theoretical framework

5.1. Augmented reality glasses

Augmented reality glasses is described as a wearable visual interface that put computer generated information into the real world, in front of the user’s field of view (Peddie, 2017) (Rauschnabel & Ro, 2015). Today we receive a lot of support from devices like smartphones and wearables when achieving tasks. Instead of using a traditional computer or device, we as users become the interface; This is a revolutionary way of interacting with a computer system. All possibilities for this technology are far from known and developers are still finding new areas to explore where AR-technology can enhance the experience or function-ality. Augmented reality glasses can further reduce friction in our day-to-day tasks, because we do not need to interrupt our actions by taking out or phones or interacting with our wearables. Interacting with multiple devices at the same time have shown to be cumber-some, since switching between devices disturb the experience (Rashid, A, & Quigley, 2012). Also, when AR-glasses are used in combination with other devices like handheld-wearable or smartphone during real time, usage has also shown to be cumbersome for users (Quigley & Grubert, 2015).

One could argue that AR-glasses could be enhanced when sound and visualization are used together. In some scenarios this could be the case, but recent research has shown that visualization mixed together with sound could have a negative impact on the user’s expe-rience. Using sound together with visualization to instruct the user have a significant de-crease of task performance in comparison to graphic visualization only (Kim, Hong, & Kim, 2016).

Gaze-guidance

The user’s attention is drawn to visual features and goals or task has also an impact on user’s attention (Grogorick, Stengel, & Magnor, 2017). Gaze-guidance is all about support-ing the user’s visual sense by guidsupport-ing the gaze towards important objects in the environ-ment (Pomarjanschi, Dorr, & Barth, 2012). Research has shown that user’s attention and task of the user are strongly correlated since the user’s gaze are directed towards object that are involved in achieving the specific task of the user (Yarbus, 1967 ).

Subtle gaze-guidance

Subtle gaze-guidance is like gaze-guidance, but it works without the user being aware of it. The focus is to direct users gaze towards important objects in the environment, but in a

(10)

6

more discreet way. This of course means that this method uses a bit different approach from traditional gaze-guidance.

The peripheral vision is more sensitive to stimuli and movement and processes stimuli quicker than the human foveal vision (Ogden & Miller, 1966). The foveal vision is better than the peripheral vision at registering details (Bailey, McNamara, Sudarsanam, & Grimm, 2009). There is some evidence that the peripheral field requires movement in order to process a stimulus effectively (Finlay, 1982). After the peripheral vision has registered interesting objects or movement in the environment it triggers saccadic eye-movement. The saccadic movement are the involuntary eye movement that quickly scans the environ-ment, directing our foveal vison. (McNamara, Bailey, & Grimm, Improving search task performance using subtle gaze direction, 2008).

Subtle gaze guidance could be used to guide or highlight important objects in a static environment. For subtle gaze guidance to work efficiently in a dynamic and complex envi-ronment, it is recommended to use stronger stimuli (McNamara, o.a., 2012). This is because our eye fixation is strongly drawn to motion in the environment and to our task and goal (Chen & Choi, 2019). AR-technology can also raise user’s awareness of the surrounding more effectively than traditional methods like radio sound, maps or handheld displays. An important part of subtle gaze guidance is that the viewer shouldn’t be aware of the manip-ulations.

Bailey, McNamara, Sudarsanam, and Grimm developed a technique that takes advantage of the difference in visual acuity and stimuli detection time between the peripheral vision and the foveal vision, by subtly modifying images to direct viewer’s gaze. By first determening where the user is not looking and then luminating that spot in their peripheral vision, the authors managed to direct the subjects gaze very effectively. The lumination is terminated before it is in the foveal vision of the viewer, so the viewer isn’t made aware of the manipulation. This is possible because of the gap in perception time between the different vision types. During tests of the technique it was concluded that the viewers gaze did not stay on the wanted point, but after the saccadic movement the eyes are drawn to the closest salient feature.

The technique has been tested and proven by other studies. “Directing Gaze in narrative arts” by McNamra o.a. are one of the studies where they used this technique to direct in which order an image was viewed without noticeably disrupting the visual experience of the image.

Influencing automatic behaviour

It has been documented that people often responds automatically to information, without much thoughts and about the time to do a certain action (Cialdini, 2001). We are pro-grammed to react to triggers that sometimes lead to unwanted consequences. Cialdini calls this phenomenon “click, whirr” and think that we’ve developed this behaviour because we

(11)

7

can’t analyse all aspects of our complicated environment and need shortcuts or quick deci-sions. The automated responses are not always the best for the situation but in many cases, it will lead us to the most rational response. Cialdini mentions several different triggers for automatic behaviour in his book about influencing behaviour, the following are applicable to our scenario.

5.4.1 Power of reason

An example of this behaviour in humans is described in an experiment by Ellen Langer and her co-workers who proved that people will let you cut in line to a copying machine if you give them a reason, no matter it's validity. We are programmed to concede to a per-ceived reasoning, by others.

5.4.2 Power of contrast

Another type of automated behaviour in humans is the power of contrast. If we are intro-duced to two options, one after the other, with slight differences, we will perceive them as being more differential then they are. This is frequently exploited in sales where the sales-person will first show the customer something in order to sell the second thing showed easier.

5.4.3 Power of inconsistency

An inconsistent person is looked at as two-faced, fake or a liar in our society and on the flip side a consistent personality is regarded as a strong and reliable person. As humans we therefore have a need to be and look consistent in our actions, no matter if the result is in our best interest or not. Once we have decided on an issue it’s easy and doesn’t take cognitive effort to automatically act consistent to how we’ve acted before. This is a difficult reaction to get rid of and according to Cialdini we use it as shield from having to think. It can go so far that we knowingly make a bad decision and then stick to it, just so we don’t have to think anymore. Businesses often uses this behaviour against us to make us buy more of their wares.

5.4.4 Power of commitment

Since we don’t want to appear inconsistent, another good way of predicting our future behaviour is to get committed. It has been proven that if we commit to a task or goal, especially if it’s a written statement, it will make us much more likely to follow through a task. When we achieve in collecting a commitment, a person will act accordingly to not be perceived as inconsistent. We are far more likely to adhere to a commitment we’ve made publicly; Many weight watching programs therefore make their participants write down their goals and then share them public. It has also been proven that the more effort our commitment takes, the greater chance it has to influence us.

5.4.5 Power of social proof

Apart from these influencing methods there is another phenomena Cialdini calls “social proof”. This describes how we as humans gages our actions correctness by comparing it to

(12)

8

other people’s actions, either around us, through observations or through social inquiry. This is the reason to why sitcoms use canned laughter after jokes: We find it funny because someone else is laughing which creates a chain reaction that spreads among individuals.

Fogg Behaviour model

B.J. Fogg has created a behaviour model for persuasive design that identifies three im-portant factors in influencing users’ behaviour (Fogg, A Behavior Model for Persuasive Design, 2009). These three crucial factors are motivation, ability and triggers. All of them need to be fulfilled in order for a behaviour to happen. A user can’t lack either motivation or ability which means that it isn’t enough to motivate the user, more often we should be focusing on making the behaviour simpler. Motivation and ability can balance each other out, for example if a user with low motivation may perform a simple task and a difficult task might be performed if the motivation is high enough. As a counterpoint sometimes we want to influence a user to not perform an action. This is more difficult but can be achieved by lowering the motivation or scaling up the difficulty to make the user unable to perform the behaviour. Good persuasive design therefore enhances motivation or ability, or both. 5.5.1 Motivation

Fogg has reached the conclusion of three motivators, with two sides to each, which he feels accounts for what motivates human behaviour most effectively though other models exist. Designers should choose the appropriate one.

The first motivator is:

Pleasure & pain , which are powerful motivators. According to Fogg this motivator, espe-cially pain, might not be the ideal approach but it is a strong motivator. Pleasure and pain usually lead to an immediate result where people are responding to what is happening in the moment without much thinking or anticipating.

The second motivator is:

Hope & fear. With this motivator the expectation of a result is present, with hope it is a good result and with fear it is a bad result. They are powerful motivators in persuasive technology and Fogg believes hope to be the most ethical and empowering motivator. The third motivator is:

Social acceptance & rejection. As people we want to avoid the ridicule of others and are motivated to gain social acceptance. Which is why we conform to things like fashion and how we should talk, what words to use. With the social technology breakthrough this

motivator has become prominent.

5.5.2 Ability

Since people are usually not inclined to be taught or trained because it demands effort, designers should focus on simplifying their design and making the actions easier to perform.

(13)

9

Designers should try to find what factor our users lack the most of to perform a task easily and try to remove obstacles. According to Fogg there are six factors that determines sim-plicity and if one of them does not meet the requirements the simsim-plicity is not achieved. The first factor is “time”: A user needs time to perform a task and if they don’t the task is not simple.

The second factor is “money”: If the user need money to perform a task and doesn’t have enough or limited amount of money, the task is not simple.

The third factor is “physical effort”: What this means is that that a task that requires physical effort might not be easy.

The fourth factor is “brain cycles”: If the user is performing another task or if the targeted behaviour demands a lot of thinking, it might not be simple to perform it. As designers we often overestimate how much users wants to think.

The fifth factor is “social deviance”: which means that users don’t want to go against the norm or break social rules, and it is not considered to be easy.

The sixth and final factor is “called non-routine” : Behaviour we perform repeatedly is usually considered simple and people usually stick to their routine. To go against a routine

is therefore considered difficult.

5.5.3 Trigger

If the user is both motivated and has the ability to perform a task, all we need is a well-timed trigger for the behaviour to be performed, even on impulse. A trigger comes in many forms but no matter which, it has 3 facts in common. These are that first we need to notice the trigger, then we have to associate it with an intended behaviour and lastly it has to happen when we are motivated and able to perform. Timing is often the thing lacking from a non-successful persuasive design, even though it is so vital that the ancient Greek had a word for it: Kairos - the opportune moment to persuade. If the trigger is presented when the motivation and ability isn’t high enough, it will not lead to acting out the wanted behaviour. According to Fogg, if the user lack motivation a trigger will be distracting and if the user lacks the ability the trigger will be frustrating.

Gamification

The purpose of gamification is to increase attention, learning, motivation and involve user engagement of the user through game-based elements from the field of game-design into a non-game context (Deterding, Dixon, Khaled, & Nacke, 2011) (Kapp, 2019). These game based elements are taken from games, hence “gamification” and involves picking elements such as stories, feedback, goals, scoreboard, challenges, progress, levels, rewards and achievements (Buchem, Merceron, Kreutel, Haesner, & Steinert, 2015). The purpose why people like gameplay is that it is rewarding, challenging, socializing and gives a fulfilment. All of this could lead to enjoyment, motivation and engagement of the user, but could also have an opposite effect depending on the users preferences (Tuunanen & Hamari, 2012).

(14)

10

Since gamification is not about actively playing a game in a traditional way from start to finish it does not require too much involvement. Instead the games elements and activities are used to support users’ task or goal in variety of areas such as education, management, health and fitness. The gamification method is implemented in many fields and fits into wide range of people and ages.

There are core principals of gamification but still, gamification is designed in different ways (Francisco & Brangier, 2013). In order to implement effective gamification, there must be a good understanding of the context of use and the user using it (Nicholson, 2012). Based on Francisco & Brangier extensive literature review and experiments, the writers have found three categories of gamification design elements which are task support, motivation and attractiveness. Description about each of these three categories and subcategories are written down below:

 Task support: Support the users in achieving their task and goals by adapting in-teractions to the user to increase knowledge and ability to achieve the task. The adaption could be (levels, models, complexity, unlocking, goals self-creations)  Motivation: motivation through emotions and persuasive elements are divided into

four subcategories:

o Self-accomplishments : (rewards, feedback, points, collection, scarcity, punition) o Social accomplishments : (competition goals, leader-board, teams, groups) o Self-expressions : (profile, voting, creating, choosing, commenting, badges,

diplo-mas)

o Social relationship : (offerings, helping, trading, community, public information, newsfeed, mischief)

 Attractiveness : This element is designed to generate good interactions like (sensory-motor, vocabulary, narration, interactive avatar) and surprises (Easter egg, discov-ery, randomness) all of this provides good emotions and high attractiveness.

Influencing with colors

The emotional power of colors is well established in many industries and it’s known to be able to influence our actions and thought patterns. Many researchers have attributed a set of emotions to a particular set of feelings, but Kolenda’s study of color theory, it’s not as simple as that (Kolenda, 2016). However, the color’s value (brightness, as in light or dark) and chroma (how vivid or washed out the color looks) have much greater effect on our emotions than hue (type of color, such as blue or red) (Kolenda, 2016). Previous studies show that our color preferences depend on 3 factors: Evolution, Gender Schema and Ecol-ogy but Kolenda gives more weight to other reasons, namely: past experiences, culture, and context.

(15)

11 5.7.1 Association

We as humans associate everything we know to other things. If we like a color not is therefore connected to the context and what else we associate the color with, often depend-ant on previous experiences or culture. Kolenda uses an example of online dating when explaining this phenomenon. When a user is accessing a dating website, they’re, maybe unconsciously, thinking about passion and romance. This thought activates an association reaction in the user’s brain such as romance, love, sex and the color red. Since red is now active, the user is more susceptible to red stimuli and be able to process images with red more easily. When the user sees a picture of a person wearing a red shirt on this website, he/she will receive a pleasant sensation which makes the person in the picture look more attractive.

In a study (Meier, D’Agostino, Elliot, Maier, & Wilkowski, 2012) it was concluded that red makes people walk faster if they’re thinking about the context love. When the test subjects where thinking about the context of achievement, they walked slower. This is because red is associated with failure in the context of achievements (think a red correction pen while grading a test) and it is associated with passion and “success” in a romantic context. The authors of the study discussed the importance of attending to context when using color psychology.

5.7.2 Arousal

Color produces two types of reactions: Arousal or evaluative reactions. An arousal reaction means that the person is experiencing a greater sense of stimulation, is feeling activated and their heart rate will heighten as well as their adrenaline and blood pressure. According to the studies Kolenda researched, warm colors increase arousal, such as red or orange. The evaluative reaction tells us if a person “likes” the color, or not. Studies show that most people tend to like colors with shorter wavelengths, such as green or blue.

5.7.3 Correct color

In order to choose a “good” color, Kolenda explains that we must consider three things. First: Appropriateness , which in turn is divided into two parts: Emotion and Semantic meaning. Emotion just means we all associate colors to certain emotions, and we like the colors which expresses the appropriate emotion for the circumstance. Like painting your bedroom in a “calming color”. A color’s semantic meaning tells us what we associate the color within a given context. As an example, Kolenda tells us that you might like the colors brown and red but it’s not very likely you would like a red table or a brown sports car because it wouldn’t be semantically appropriate. The two other things to consider is Aes-thetic and Value. AesAes-thetic is simply that is important that we choose to use visually appealing colors. Value has two subsections: social and functional. A color has a social value that depends on if it is socially useful. For example, in 2019 coral was very much in fashion and had a high social value. The functional value depends on if the color is appro-priate for the objects function. White shoes have a low functional value since the ground is dirty and they won’t stay white for long.

(16)

12 5.7.4 Warm colors (red in particular)

When processing information, there are generally two ways to go about it, namely Heuris-tically (quick and simple-minded analysis) or SystemaHeuris-tically (Thorough and rational anal-ysis). When you want the user to take their time to make an informed choice, cool colors (like purple, green or blue) should be used to decrease arousal. And vice versa, stating that warm colors (like red, yellow or orange) increases arousal and therefore lowers cognical functions, forcing the users to make quick decisions. In a study discussing the power of the color red (Kuniecki, Pilarczyk, & Wichary, 2015), it was demonstrated that the color red captures attention and facilitates congruent motor response, when in an emotional context (following an emotional stimuli). This resulted in that participants were faster and more accurate in their responses when following a red-colored cue. Red facilitates motor output. However, there are some contexts where red is associated with stop, danger and avoidance (Rohr, Kamm, Koenigstorfer, Groeppel-Klein, & Wentura, 2015), a simple example of this is our traffic lights. It has been proven that in the context of achievement red is associated with danger and caution (Moller, Elliot, & Maier, 2009), which means that a trigger for association with achievement should give us the opportunity to stop users from unwanted actions.

6. Semi structured interview

In order to make sure that the users’ motivations and sense of abilities coincided with what other researchers had found about influencing and behaviourism we conducted a semi-structured interview (see Appendix A: Figure. 6.1., 6.2). We asked 40 people, using a survey, who were using either the stairs or an elevator, with no thought on their age, gender and ethnicity. The questions were about their energy levels, level of distraction, motivations for choosing stairs or elevator and their exercise routine. This study was conducted in the entrance of a modern building (Niagara) belonging to Malmö University, Sweden, May 2019. The entrance has easy and clear access to both stairs and elevators and had plenty of room for us to take photographs for later manipulation. Through the literature study we found that the intensity of visual stimuli needs to be greater with a lot of distractions. We decided to conduct part of the interviews when the entrance was almost vacant and the other part when it was busier. The qualitative answers where thematically analysed to give us themes. These themes were compared to and found similarities in Cialdini’s and Fogg’s work regarding behavioural studies (Cialdini, 2001), (Fogg, A Behavior Model for Persuasive Design, 2009).

Result

The survey form can be found at the end of this paper (see Appendix A: Figure. 6.1., 6.2) The first part of the survey, question 1-3 was answered by observing the behaviour of respondents. While reviewing the answers on question 1, we can tell that being occupied with other activities did not influence if the users chose to take the stairs or the elevator, since 50% of the questioned was determined to be multitasking (Figure. 6.3). Through

(17)

13

thematic analysis we determined 3 types of reason for distraction (Figure. 6.4), namely watching a screen, using headphones to listen to something and talking to friends and classmates.

(Figure. 6.3 Percentage of distracted users who were multitasking during our observation. “No” means the user wasn’t distracted. “Yes” means the user was distracted.)

(Figure. 6.4 By doing a thematic analysis of we found three main themes that represent user’s types of distractions.)

There was a slightly higher number of stairs users that were distracted by social reasons (Figure. 6.5), this could mean that social interactions could motivate users to take the stairs. It could also be a coincidence, since only a limited amount of people was asked. However, when questioning the respondents for reasons and situations when they would choose the other option (Figure. 6.6), 15% of elevator users and 10% of stairs users named social reasons. These reasons were extracted by sorting the users’ response into themes (see Figure 6.7). This coincides with what Cialdini tells us about the theory “social proof” (Cialdini, 2001) and what Fogg says about the third motivator being social acceptance/rid-icule (Fogg, A Behavior Model for Persuasive Design, 2009).

We also found that stairs users were not watching a screen or interacting with their phone while walking up the stairs (Figure. 6.5), from which we conclude that people want focus and to see where they’re going while walking up the stairs. This is supported by the fact that a few more of the stairs users where not distracted by another task and coincides with

50 % 50

%

(18)

14

Fogg’s behaviour model when discussing users’ ability being hindered by their brain cycles being too occupied to see the usage of stairs as simple (Fogg, A Behavior Model for Persuasive Design, 2009).

Figure. 6.6 and 6.7 also tells us that elevator users would only consider taking the stairs if there were barriers towards using the elevator, with a few exceptions claiming that friends could make them change their usual behaviour. Stairs users are more likely to change their choice to elevator for different reasons.

(Figure. 6.5. Some of the questioned users were distracted. The reasons of distractions were sorted into types. “None” means the user wasn’t distracted. “Social reasons” mean they were talking and/or walking with someone. “Using headphones” means they were listening to something using headphones. “Watching screen” means they were interacting with a mobile device with a screen. A) Represents the number of Elevator users, grouped by the type of distraction they were experiencing while taking the elevator. B) Represents the number of Stairs users, grouped by the type of distraction they were experiencing while taking the stairs.)

0 1 2 3 4 5 6 7 8 9 10

A

None Social reasons

Using headphones Watching screen 0

1 2 3 4 5 6 7 8 9 10 11 12

B

(19)

15

(Figure. 6.6. The users were asked if they would change their choice for any reason. The reasons were grouped into types. “Barrier” means the user would change their choice if their usual choice would be or appear unavailable. “Simplicity” means the user would change their choiceif the other option became the simpler choice. “Social” means the user would change their choice if the people they’re walking with would want to choose the other option. “None” means they would never change their decision. “Healthier life choices” means the user would change their choice if they decided to live healthier. A) Represents the number of elevator users, grouped by reason to why they would change their decision from elevator to stairs. B) Represents the number of Stairs users, grouped by reason to why they would change their decision from stairs to elevator)

(Figure. 6.7 Themes extracted from survey for reason why user would choose the other option than they did the day of the survey.)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

A

Barriers Simplicity Social

0 1 2 3 4 5 6 7 8 9

B

Barriers Healthier life choices

None Simplicity

(20)

16

The users’ answers were analysed and sorted into four categories, regardless of they chose the stairs or elevator (Figure. 6.8). Some users responded that they had no reason for their choice and are grouped as “no reason”. A total of five categories was therefore used in our diagrams depicting this. The elevator users were as suspected more likely to make their choice because of automated actions (Figure 6.9) and by barriers toward stairs like to far to walk, tiredness and soreness (Figure. 6.8) and some of them had no reason, showing another tendency towards automated behaviour. Elevator users main reason for them choosing the elevator over the stairs, was the simplicity like ease of use and availability. All the user’s answers were analysed and sorted into themes, regardless of their choice of choosing stairs or elevator. All elevator users have what they consider reasons for not taking the stairs.

(Figure. 6.8 Themes were extracted from survey and then processed to show reasons for users’ choices.)

(21)

17

(Figure. 6.9. “Automated action” means they didn’t make a clear decision it but could think of a reason to why they did it, such as walking with friends or force of habit. “Barriers” means there was, or they perceived it to be more difficult to take the stairs. “Simplicity” means they took the elevator because it was or appeared easier. Healthier lifestyle” means they take the stairs because it’s better for their health. “No reason” means they didn’t make clear decision and didn’t know why they did take the stairs. A) Represents the number of Elevator users, grouped by the reason they chose to take the elevator. B) Rep-resents the number of Stairs users, grouped by the reason they chose to take the stairs.)

When just examining the users’ thought pattern when making their choice, after doing a thematic analysis of their answers (Figure. 6.10), we concluded that elevator users make their choice dependant on how simple they find using the elevator and their perceived lack of ability to take the stairs because of barriers (Figure. 6.11). These answers coincide di-rectly with Fogg’s behavioural model regarding ability (Fogg, A Behavior Model for Persuasive Design, 2009). 0 1 2 3 4 5 6 7

B

Automated action Barriers Healthier lifestyle No reason Simplicity 0 1 2 3 4 5 6 7 8 9 10 11

A

(22)

18

(Figure. 6.10 The users were asked what they thought about before choosing the elevator or the stairs. These reasons were sorted into themes.)

(Figure. 6.11 The Elevator users were asked what they were thinking about before they made the choice to take the elevator. “Barriers” means they thought about reasons to not take the stairs. “No” means they weren’t thinking about anything in specific or can’t re-member. “Simplicity” means they were thinking about how it would be easier to take the elevator. “Health” means they were thinking about the health benefits of taking the stairs. A) Presents the number of elevator users, grouped by theme. “Social” means they were considering the social consequences of their choice. B) Presents the number of stairs users, grouped by theme.)

Elevator users see the elevator the easier choice and as easy to use, regardless if it has a que or not and we have determined that they see other barriers as more difficult in order

0 1 2 3 4 5 6 7 8 9

A

Barriers No Simplicity 0 1 2 3 4 5 6 7 8 9 10 11 12 Summa

B

(23)

19

to justify their choice, despite of the elevator being occupied (Figure. 6.12). This is further emphasised by the fact that elevator users claimed different automatic action reasons for their choice while the elevator was occupied. An explanation for this is described by Cialdini’s thoughts on power of inconsistency or reason (Cialdini, 2001), if we assume that the users simply see themselves as people who take the elevator and keep finding reasons to do so. It is further supported by what Fogg says about people’s viewed ability (Fogg, A Behavior Model for Persuasive Design, 2009) is affected by time and physical effort, which were mentioned as barriers by the users. When examining the stairs users, also when the elevator is occupied, we can determine that the users are motivated by automatic actions and view the elevator as hindering, not a means to simplify the action. They are also more likely to avoid the elevator if they are motivated by health reasons, supporting Fogg’s theory on hope/fear being a good motivator (Fogg, A Behavior Model for Persuasive Design, 2009).

(Figure. 6.12 The users were asked what motivated their choices and it was recorded if the elevator was or wasn’t occupied at the time. “No” means the elevator was available and “yes” means elevator was occupied. “Automated action” means they didn’t know their motivation but could see reasons to why they made their choice. “Barriers” means they were trying to avoid what they perceived as barriers. “Simplicity” means they wanted to take the simplest option. “Healthier lifestyle” means they want live healthier. “No reason” means they don’t know why they made that choice. A) Presents what motivated the ele-vator users’ choice, grouped into if the eleele-vator was occupied or not. B) Presents what motivated the stairs users’ choice, grouped into if the elevator was occupied or not.)

0 1 2 3 4 5 6 7 8 No Yes

A

Automated action Barriers Simplicity 0 1 2 3 4 5 No Yes

B

Automated action Barriers Healthier lifestyle No reason Simplicity

(24)

20

When comparing the users’ reasons depending on if others were using the stairs, we can tell that elevator users again refer to how simple it is to use the elevator, even giving it as a clearer reason if the stairs are being used (Figure. 6.13). Since elevator user keeps using this as a motivator for their choice we make the assumption that it has to do with what Cialdini refers to as power of reason, no matter if the reason makes sense or not, we as people will follow it (Cialdini, 2001). We can also see a slight increase of automatic behav-iour in the stairs users when joined by other on the stairs, hinting at what Cialdini (2001) says about the power of social proof and Fogg’s motivator of social acceptance/rejection (Fogg, A Behavior Model for Persuasive Design, 2009).

(Figure. 6.13. The users were asked what motivated their choice and it was recorded if the stairs were being used by others. “No” means the stairs was at least almost clear and “Yes” means it was at least somewhat crowded. “Automated action” means the user didn’t think about their motivation but just acted. “Barriers” means the user was motivated to avoid the most difficult choice. “Healthier lifestyle means the user wants to live healthier and that influenced their choice. “No reason” means the user had no motivation. “simplicity” means the user was motivated by how easy their choice seemed. A) Presents how the elevator users motivated their choice, grouped by if the stairs were crowded by others at the time. B) Presents how the stairs users motivated their choice, grouped by if the stairs were crowded by others at the time.)

There wasn’t a clear connection between how much exercise a user gets and how often they take the stairs/elevator (Figure. 6.14). But the graph shows a slight preference to take the elevator if they exercise less, but not all fit that description. Just as active stairs users are a little more likely to take the stairs than those who are less active. From the people we surveyed, every stair user that claimed to exercise very much almost always took the stairs. This could be partly because of what Cialdini mentions as the power of inconsistency and power of commitment, since less active people tend to make less active choices and vice versa (Cialdini, 2001). It could also have to do with what Fogg says about the user’s ability

0 1 2 3 4 5 6 7 No Yes

A

Automated action Barriers Simplicity 0 1 2 3 4 5 No Yes

B

Automated action Barriers Healthier lifestyle No reason Simplicity

(25)

21

being hindered by the physical effort which would be viewed as more prominent if not exercising regularly (Fogg, A Behavior Model for Persuasive Design, 2009).

(Figure. 6.14 A) Presents how the amount of exercise influences how often elevator users take the elevator B) Presents how the amount of exercise influences how often stairs users take the stairs)

We could not find any useful data when analysing their energy level (Figure. 6.15) or when purely analysing how often they take chosen path (Figure. 6.16). Gathered data regarding question 2 didn’t say anything on its own but can be studied in Figure. 6.17. The same goes for question 3 (see Figure. 6.18) and question A (see figure 6.19). These questions were instead used to analyse other answers as seen in previous text.

(Figure. 6.15. A) Presents how much energy the elevator users claimed to have had. B) Presents how much energy the stairs users claimed to have had.)

0 1 2 3 4 5 6 A little Pretty

much Very little Very much

N um be r o f u se rs

How often they use the elevator

A) Amount of exercise

Almost always Not often Often Very seldomly

0 1 2 3 4

A little Prettymuch Very little Very much

N um be r o f u se rs

How often they use the stairs

B) Amount of exercise

Almost Always Not often Often Very seldomly

0 1 2 3 4 5 6 7

B

A bit tired Pretty energized Very energized Very tired 0 1 2 3 4 5 6 7 8 9 10 11 12

A

(26)

22

(Figure. 6.16. The users were asked how often they use the same option the chose today. A)Presents how often a certain number of the elevator users, uses the elevator. B) Presents how often a certain number of the stairs users, use the stairs.)

(Figure. 6.17. It was recorded if the elevator was occupied when the users made their choice. “No” means it wasn’t occupied and “Yes” that it was. A) Presents the number of elevator users grouped by if the elevator was occupied or not. B) Presents the number of stairs users grouped by if the elevator was occupied or not.)

0 1 2 3 4 5 6 7 8

A

0 1 2 3 4 5 6 7 8 9 10 11 12

B

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

A

No Yes 0 1 2 3 4 5 6 7 8 9 10 11

B

No Yes

(27)

23

(Figure. 6.18. It was recorded if the elevator was occupied when the users made their choice. “No” means it wasn’t occupied and “Yes” that it was. A) Presents if the stairs were being used by others when the elevator users made their choice. B) Presents if the stairs were being used by others when the stairs users made their choice)

(Figure. 6.19. A) Presents how much the elevator users exercises. B) Presents how much the stairs users exercises)

0 1 2 3 4 5 6 7 8 9 10 11 12

B

No Yes 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

A

No Yes 0 1 2 3 4 5 6 7 8 9

A

A little Pretty much Very little Very much

0 1 2 3 4 5 6 7 8

B

A little Pretty much Very little Very much

(28)

24

Analysis

During the semi structured interviews analysis, we concluded that we need methods to handle 8 different noticed situations. These are divided below, into if they are referring to elevator or stairs usage.

Stairs users needed methods

 Since many users responded that they would change to the elevator if their friends wanted them to, we need a method which motivate the user to keep using the stairs.  Several stairs users pointed out the health benefits being a reason for choosing the

stairs. We therefore need methods to keep the user feeling active and accentuate these health benefits.

 Users often choose the stairs if they see their friends doing so. A method is needed to highlight those the user knows, who are currently using the stairs.

 The stairs can seem like they take more effort than the elevator. This means that the user needs to feel active and reminded to keep using the stairs.

Elevator users needed methods

 Since many users responded that they would change to the stairs if their friends wanted them to, we need a method to highlight the friends that are already using the stairs. A method for motivating user and friends to all use the stairs would also be helpful.  Mobile devices, sounds or other distractions limit the user’s available brain cycles and

makes them more likely to do automated actions. We therefore need to disrupt their distractions and pull their focus toward the stairs, in an effort for the user to make an active choice.

 Many of the elevator users said that using the stairs had many barriers and difficulties. A method is needed to limit the perceptions of these barriers, making the stairs seem more appealing.

 Elevator users tend to either not think before their choice or get hung up on the simplicity of using the elevator or the barriers to use the stairs. We want them to stop and focus on the reasons for choosing the stairs, meaning the goals they have set.  Elevator users will not change their choice to taking the stairs unless the elevator is

unavailable. A method to achieve this is needed along with making the stairs more appealing.

 Since elevator users tend to see the barriers for using the stairs as greater than the barriers for using the elevator, we need to give them further reasons in support of taking the stairs.

 Again, barriers towards taking the stairs seem very important to elevator users, mean-ing that the stairs need to be made to look simpler and more appealmean-ing. In accordance

(29)

25

with using people’s automatic behaviour it would help to make the users just follow the movement of people surrounding them, up the stairs and avoiding the elevator.

7. Scenarios/Real world system

In this section we describe todays real-world scenario that highlights factors that affects users’ choices in everyday situations. This scenario involves users making two choices be-tween taking the stairs or elevators to get to their destination in a building. We observe the user’s interaction and interview the users to get in-dept information about why did their choices. The information from the interviews we gathered are then tied and analysed with our literature study. With the data and information, we created a vision scenario where we describe how the future AR-system could persuade the users into different action and what kind of factors the system must utilized to persuading the user. We will also show our readers how the AR-glasses interface could look like in this scenario.

The main factors that affects users’ choices in this specific scenario are:  Availability (available elevator)

 Social (friends have a great impact on our decision making)  Social barriers (many people, long waiting lines, awkwardness)  Personal barriers (soreness, laziness, tiredness, motivation)

Our scenario begins with a user of an AR system walking into a building, wanting to get to one of the top floors of said building. Both the elevators and stairs are easily accessible in the building used in this scenario. Depending on the amount of people, movement, conditions and distractions on the ground level, the user decides on which path to choose. According to both Fogg (2009) and Cialdini (2001) these decisions are usually pre-decided and automatic. This means the system needs to take in the environment immediately and steer the user’s attention away from the elevator and towards the stairs.

Since the system can be worn always, it is important that the system does not feel disrup-tive or in the way. We want the user to feel like they are not actually wearing any glasses and not really consider the manipulations and visual ques.

While walking towards the stairs and elevators, the system should register any information and object which could help persuade the user in steering their steps toward the stairs. Depending on social situation, accessibility of the elevator and barriers, the user will either walk towards the elevator or stairs. The system should highlight the stairs and make them look accessible and easy to use, and make the elevator seem less accessible and problematic. If the user is about to or walking towards the elevator for any reason, they should be redirected to the stairs.

(30)

26

8. Concept model

At the beginning of this paper we proposed our conceptual model that could be used for future development towards persuasive AR-systems. The foundation of our conceptual model is based on three areas which are psychology, color theory and persuasive technology (see figure 8.1).

(Figure 8.1. Conceptual model)

In this chapter we describe how psychology, color theory and persuasive technology can work together to influence user behaviour, as a result of our study. We chose to create a table that show how things are tied together in a simple way (Table 8.1). The column “Psychology” contains theories on human behaviourisms. The column “Scenario” contains real-world situations, we observed people being in, while using the stairs or elevator. When combining the psychology and specific scenario, we could see what type of model could be helpful in that situation and wrote it down in the column called “Models and Theories”. The last column, called “Figure”, you can find a number that correlates with a picture in the section “Vision scenario” where the related models are applied.

(Table 8.1 Connections between methods)

Psychology Scenario Models and theo-ries

Figure Fogg motivator

hope/fear

Before task exe-cution

Gamification: mo-tivation self-ac-complishment.

3.1 Color influence:

As-sociation with achievement

Before task exe-cution

Gamification: mo-tivation self-ac-complishment

(31)

27 Fogg brain cycles User is

dis-tracted by other tasks such as screen viewing

Subtle gaze guid-ance and gamifi-cation: Task sup-port

3.2

Fogg: Ability phys-ical effort

Before choosing stairs or eleva-tor (without so-cial or environ-mental distrac-tions) Color theory, gaze-guidance, Gamification: At-tractiveness 3.3 Influence: power of social proof, Fogg: Ability, social devi-ance Walking/talking with friends Gamification: Task Support-self accomplishments, subtle gaze-guid-ance, Color theory 3.4 Fogg: Ability, physical effort, In-fluence: Power of commitment, Fogg: non-routine Elevator is available, about to choose eleva-tor over stairs

Color theory, gaze-guidance 3.5 Influence: Power of commitment, in-consistency

Used stairs Gamification ( Social accom-plishments & Task support & self-accomplish-ments)

3.6

Fogg: Ability, time constraints

(32)

28 Influence: power of

social proof, Fogg: Ability, social devi-ance

User sees friends using the stairs

Subtle gaze guid-ance

3.8

Vision scenario

Keep in mind when reading the scenarios that the theories and model in our conceptual model and the scenarios was produced based on the human visual sense only and that in our future vision. The users are controlling the AR-glasses through a smartphone but does not interact with smartphone and AR-glasses at the same time. The user set-up specific goals or settings before using the AR-glasses and the AR-glasses works on its own inde-pendently. During the vision scenario we will frequently use the method of warm colors (5.7.4 Warm colors and red in particular), so we will not reference it, for ease of read. Before the user has the opportunity to choose between stairs or elevator, we need to moti-vate them into choosing the stairs. As earlier mentioned in the section persuasive de-sign/fogs behaviour model, Fogg says that hope is a powerful motivator (5.5.1 Motivation). We have also learned that self-accomplishment in gamification is a great motivator, such as rewards and collection (5.6 Gamification). By combining these two facts, in the shape of badges and progress bar, we hope to prime the user into choosing the stairs (figure 8.2). Since we want to be able to use colors to further strengthen our persuasions, a correct association is also needed (5.7.1 Association). This means that the badges and progress bars should trigger an association with achievement, in order for us to use red as a deterrent later on. We do not simply want to use red because it means danger in connection with achievement, it has as mentioned also been proven to facilitate greater motor output.

(33)

29

(Figure 8.2 Before the user starts to walk towards the stairs/elevator they are primed using a gamification: progress bar. To start the association with achievement)

If the system notices that the user is pre-occupied with their phone, they need to be per-suaded to put the phone away out of sight or direct their attention towards the stairs. Since taking the stairs requires more cognitive load, many seems to take the elevator so they can focus on their devices (6.2 Result). According to the information we gathered about subtle gaze-guidance, we think we can make the user look away from the device and direct attention towards the stairs (5.3 Subtle gaze-guidance). In figure 8.3 you can see a man’s face is illuminated to attract attention, which is removed as soon as the user look that way. The illumination would be moving a little, since it is more attention grabbing. This gives them the ability to use the stairs, which according to Fogg’s discussion about ability is vital (5.5.2 Ability). When the user is directing their eyes towards the stairs, a trigger should be presented, reminding the user of the goal. Since we have already primed the user before they are close to the stairs or elevator, we think a simple task support method (5.6 Gamification) from gamification should be enough. One possible design sug-gestion could be a to show checkbox getting checked.

(34)

30

(Figure 8.3 If the user is distracted by their phone or other screen, their eyes are redirected to the stairs using subtle gaze-guidance: illumination, which can be seen as a light bubble in the left picture. A checkbox tells the user, in the right image, that they accomplished the wanted task and a gaze-guidance: path appears for the user to follow.)

We rely heavily on colors when guiding users next path, so the way the stairs and elevator are colored will further persuade users into walking up the stairs (figure 8.4). The infor-mation about color theory tells us that warm colors excites people, making them feel more energetic (5.7.4 Warm colors and red in particular). According to gamification theory, attractiveness is important in gaining user attention (5.6 Gamification). We propose col-oring the stairs in warm colors except for the color “red” which has been chosen for the elevators as a sign to stop. The warm colors have an increase of triggering high arousal, leading to quicker and spontaneous decisions.

(Figure 8.4 The stairs are colored warmly to show where we want the user to go while leaving them feel energetic and up to the task)

(35)

31

If a user is walking/talking with friends and moving towards the elevator, the system needs to persuade the user into redirecting the group’s route towards the stairs. The stairs are colored to remind the user of their task and make them feel energetic enough to have the ability to choose the stairs (figure 8.5). However, this still means the user is has to work against what Cialdini tells us about social proof and what Fogg says about social deviance (5.4.5 Power of social proof, 5.5.2 Ability). We previously mentioned that a lot of distrac-tions calls for a stronger stimulus, we therefore chose to include a colored path directing the user’s steps up the stairs (5.2 Gaze-guidance). This scenario also holds a possibility of using gamification self and social accomplishment in the form of a badge for influencing friends into choosing the stairs, if they managed to redirect the group.

(Figure 8.5. When a user is walking with others, a gaze-guiding path reminds them of their commitment to the task of taking the stairs. A warmly colored staircase makes them feel up for the task. Coloring the elevator red makes the user feel like there is a barrier to using it.)

If the user normally takes the elevator, the power of commitment and routine is making the elevator appear less challenging. The user will follow their automatic behaviour up to the elevator. In that case, we need the user to stop and redirect their attention. Since we primed achievement before the user came this far, red means stop or danger (5.7.4 Warm colors and red in particular). We therefore color the elevator red, making it seem unavail-able and reduce ability (figure 8.6). The warm colored path tells the user to act fast and help guide their gaze and steps towards the stairs.

(36)

32

(Figure 8.6 Coloring the elevator red makes it appear unavailable and a warm colored gaze-guidance: path leads the user away back to the stairs.)

When the user has used the stairs, we want them to commit to this choice because Cialdini tells us this forms new automatic behaviour and Fogg says it makes it harder to make a different choice in the future (5.4.4 Commitment, 5.5.2 Ability). One possible solution would be to use gamification methods and share the achievement on social media (8.7 Gamification). Another option is to show a progress bar that show how many more times the user has to take the stairs before they get a specific badge or achieve their set goal. However, the user should always be rewarded with a badge when having used the stairs as a reward for self-accomplishment to add motivation (figure 3.6). Since we don’t want the system to feel intrusive or distracting, the system should choose to show the most rewarding and commitment-inducing badge or progress bar received, not several.

(Figure 8.7 A gamification: badge after having completed the task confirms the action and motivates the user into taking the same action again)

(37)

33

A user in a hurry will feel like he/she has less ability to make” the right choice” because of time constraints (5.5.2 Ability). Using warm colors guiding up the stairs should make them feel like it is the quickest path (figure 8.8). If the user is not aware of the system being able to do this, it would be unethical (9.4 Resolving the ethical issue of autonomy). That is why the user is informed, that they will be manipulated in this way, when activat-ing the system for the first time. The elevator should be colored red, makactivat-ing it appear unavailable and the user think that route consumes more time.

(Figure 8.8 Using warm color we make the stairs look like it’s the fastest route and the red elevator make them appear unavailable for hurried users. The warmly colored gaze-direc-tion: path further the feeling of excitement and influences the user into taking a less thought through decision.)

Seeing as the social environment has such great impact on our choices, according to Fogg, Cialdini and our study we have concluded that the system should take advantage of “friends” already using the stairs. If a person, known to the user, is using the stairs the system should attract the user’s eyes to that fact (as seen in figure 8.9 where the girl on the stairs is illuminated until the user is looking directly at her). This could be done using the illumination method from subtle gaze-guidance, where something is illuminated in the periphery of the user’s gaze (5.3 Subtle gaze-guidance). Acknowledging a friend using the stairs, should be trigger (5.5.3 Trigger) enough to remind the user of their choice to use the system and taking the stairs. If not, the social influence could be enough to push the user into the wanted action.

(38)

34

(Figure 8.9 A friend using the stairs could serve as a motivation for choosing the stairs. The user’s friend is using the stairs and her face is subtle gaze-guidance: illuminated to draw attention)

Summary

In summary, the conceptual model says that when the user enters a building and is moving towards stairs and elevators, they should be primed with the association of achievement. This is done using gamification methods and results in us being able to use red as a warn-ing/stop signal and motivate motor output with warm colors.

If at any time the user is preoccupied or not looking towards the stairs when they should, subtle gaze-guidance can be used to redirect their attention. Illuminating a moving point in the stairs (such as a friend’s face in the periphery of the user’s gaze) and then removing it as the user looks that way, will achieve this.

This method is also helpful when the system notices and want to show a friend of the user taking the stairs, because it could influence the user into doing the same thing.

When the user does what the system wants, such as viewing the stairs instead of their phone, a gamification task-support method such as a check box is showed.

To increase the user’s ability and motivation, the stairs are colored a warm color such as yellow or orange because it makes people feel energetic. It also makes the user come to faster decisions with less afterthought, hopefully automatically following the color.

A conceptual model for future wear-able augmented reality system: Influence user behaviour into choos-ing stairs over elevator

Examination Thesis, Computer science