Business Information Systems
MASTER THESIS
The integration of natural interaction within
Data Analytics visualization by using a Kinect
Student:
Dhr. ing. Joël Pattynama 10430423
University of Amsterdam
Supervisors:
Dhr. dr. G.P.A.J. (Guus) Delen
Dhr. drs. R. (Ronald) Kleijn
Carried out at:
Dr. Guus Delen, Signature:
Drs. Ronald Kleijn, Signature:
Prof. Dr. Tom van Engers, Signature:
Abstract
Version, Date 1.0, Amsterdam June 27, 2014 Author Joël Pattynama
University University of Amsterdam Faculty Faculty of Science
Master Business Information Systems Student number 10430423
Company Avanade, Netherlands Thesis title The integration of natural
interaction within Data Analytics visualization by using a Kinect
Three different groups participated in a study on the additional benefits of natural interaction using a Kinect when presenting the equivalent Data & Analytics information. The purpose of this study is to investigate the question of whether a Kinect-controlled presentation provides improved engagement over that of a traditionally controlled presentation. Therefore three groups of participants took part in both presentations. To assess the quality of engagement, an own engagement model was proposed. This consisted of three engagement phases in order to determine which particular phase of engagement the participants found the most advanced. The results showed that the participants experienced a better quality of engagement during a Kinect presentation than a more traditional one. These results indicate that a Kinect presentation provides better control, and this is of particular interest due to its novelty value. The conclusion drawn from this is that a Kinect presentation based on a one-time experience provides a better quality of engagement.
ACKNOWLEDGEMENT
This thesis completes my study in Business Information Systems. It has been a considerable challenge, particularly as I once started on the preparatory secondary vocational education (VMBO). I could not have achieved this without help. I would like therefore to take the opportunity to thank some of the people who helped me through this difficult period:
I am grateful to my teachers’ dr. Guus Delen and drs. Ronald Kleijn for their continuous support, critical opinions, and much more helpful advice during my thesis study.
Nick Holthuijsen, classmate and colleague, has also been a great support for my thesis. I thank him especially for all the insightful discussions and opinions.
A special thanks goes to my colleague Jan Staal who helped me in completing my thesis, and to Gijs Rademaker who guided me through all the additional issues during the writing of it. In addition, I would like to thank all the other Avanade and Schuman employees who have helped me during this thesis experiment.
I would like to thank my parents for their undivided support and for their appreciation of my work. They always believed in me and supported my decisions in every way they could, and their unconditional love has been the backbone from which I was enabled to complete this long journey.
At last I want to thank my girlfriend who inspired me and encouraged me to go my own way. Without her, and her motivation, I would have been unable to complete this thesis.
Amsterdam, June 2014
TABLE OF CONTENTS
ACKNOWLEDGEMENT ... 3 1. INTRODUCTION ... 6 2. THEORETICAL BACKGROUND ... 8 2.1 CONCEPTUAL FRAMEWORK ... 10 2.2 HYPOTHESES ... 15 2.2.1 POINT OF ENGAGEMENT ... 15 2.2.2 PERIOD OF ENGAGEMENT ... 16 2.2.3 DISENGAGEMENT ... 17 3. METHOD ... 17 3.1 DESIGN ... 17 3.2 PARTICIPANTS ... 17 3.3 MATERIALS ... 18 3.3.1 MICROSOFT KINECT ... 183.3.2 MICROSOFT KINECT FOR WINDOWS SDK ... 18
3.3.3 PRESENTATIONAID ... 18 3.3.4 MICROSOFT POWERPOINT ... 18 3.3.5 TEST PROCEDURE ... 19 3.3.6 QUESTIONNAIRE ... 19 3.4 DATA ANALYSIS ... 19 4. RESULTS ... 20 4.1 ATTRIBUTES OF ENGAGEMENT ... 20 4.1.1 ATTRIBUTES RANKING ... 21 4.2 PHASES OF ENGAGEMENT ... 22
4.2.1 INDEPENDENT SAMPLES T-TEST ... 22
4.3 QUESTIONNAIRE ... 24
5. DISCUSSION AND CONCLUSION ... 25
5.1 HYPOTHESES ... 25
5.2 STUDY LIMITATIONS... 26
5.2.1 DURATION OF THE STUDY ... 26
5.2.2 STUDY VALIDITY ... 26
5.2.3 PRESENTATIONAID KINECT APPLICATION ... 26
5.3 CONCLUSION ... 27 5.3.1 RECOMMENDATIONS ... 27 5.3.2 FURTHER RESEARCH ... 27 6. REFERENCES ... 28 7. APPENDICES ... 31 7.1 APPENDIX A: QUESTIONNAIRE ... 31
7.2 APPENDIX B: QUESTIONS CHI-SQUARE ... 36
7.3 APPENDIX C: ATTRIBUTES STATISTICS... 37
1. INTRODUCTION
Today’s world deals with a great deal of data; data created by humans and computers. Data used to be a technical problem but has become these days a business opportunity. Ackoff (1989) developed the Data-Information-Knowledge-Wisdom (DIKW) hierarchy. “The DIKW hierarchy is widely accepted in
knowledge management circles as a way to represent the different levels of what we see and what we know”
(Schumaker, 2011). The DIKW hierarchy provides the process to translate data into wisdom. The first step, in the DIKW hierarchy, focuses on the translation from data to information. Ackoff (1989) defines information as: “information is data that has been given meaning by way of relational connection. This
“meaning” can be useful, but does not have to be. In computer parlance, a relational database makes information from the data stored within it”. Thus, according to Ackoff data can only be seen as information
when it is useful. Because of the enormous and increasing amount of data, organizations face problems in controlling, managing and especially understanding their data. Understanding this data can obviously be useful for organizations, and they automatically process it into information. But in order to understand the data, organizations also need to analyze it. In order to explain the process of data analysis some relevant terms will be explained.
Business Intelligence became a popular term in the business and IT communities only in the 1990s, and in
the late 2000s Business Analytics was introduced to represent the key analytical component in business intelligence (Chen, et al., 2012). More recent terms include Big Data and Big Data Analytics. According to Chen et al., those terms have been used to describe the data sets and analytical techniques in applications that are extremely large, from terabytes to exabytes; and also increasingly complex, from sensor to social media data, that they require advanced and unique data storage, management, analysis, and visualization technologies.
Business Intelligence (BI) refers to tools and technologies for collecting, storing and analyzing available data about an organization’s operations (Senan, et al., 2008). This helps with decision-making within organizations. The role of BI is becoming increasingly important. In a survey Gartner found that BI projects were the number one technology priority for 2007 (Watson & Wixom, 2007). They mentioned that “BI has
become a strategic initiative and is now recognized by CIOs and business leaders as instrumental in driving business effectiveness and innovation”. They mentioned that “the future of business intelligence is bright”
and “major vendors such as Microsoft, Oracle, and SAP are increasing their commitment and investments
in business intelligence”. They claim that there will be other developments soon on the horizon that promise
to keep BI fresh and exciting by the integration of structured and unstructured data, rules engines, guided analytics, and enhanced data visualization. Organizations tend to focus more on predictive analytics, which is what most people refer to as Data Analytics. This gives the probability of various outcomes and is future-oriented. Therefore organizations talk about Big Data which refers to the tools, processes, and procedures that allow on organization to create, manipulate, and manage very large data sets and storage facilities (Knapp, 2013).
Business analytics is also still in the “emerging stage” according to research conducted by Bloomberg Businessweek (2011). They define “business analytics” as a technology that uses data analysis to understand business issues in a way that can guide decision-making. Bloomberg Businessweek (2011) launched a research program in order to determine the current state of business analytics. Their program focused on the entire business of data analytics; spreadsheets, business reporting, data and text mining, general statistics, model management, optimization, web analytics, social media analytics, interactive data visualization, and text, audio and video analytics. In their survey they asked the question Which of the
following would you consider to be business analytics capabilities/tools? From the 14 suggested tools,
participants believed that interactive data visualization can be seen as a useful tool. Both answers comprise a relatively low percentage of the total amount. This means that despite the phenomenon being known, it is probably not yet sufficiently enticing.
According to Theus (2003) data visualization has been acknowledged as an important tool in decision support. To enhance data visualization, attention ought to be paid to Human Computer Interaction and Ergonomics (Human Factors). Human Computer Interaction (HCI) refers to the interaction between human beings (users) and computers. HCI can be defined as “a discipline concerned with the design, evaluation,
and implementation of computing systems for human use and with the study of major phenomena surrounding them” (Fetaji & Loskoska, 2007). The term Human Factors is more widely used in the United
States while the term Ergonomics is more prevalent in Europe and the rest of the world (Sanders & McCormick, 1987). Currently the term Ergonomics is often combined with human factors, known as Human Factors Ergonomics (HFE). HFE focuses on systems in which humans interact with their environment. The environment is highly complex and consists of the physical environment, the organizational environment, and the social environment (Dul et al., 2012). Thus HFE focuses on human beings and their interactions with systems and other peripherals.
Although various kinds of definitions are discussed on dealing with information useful for organizations, this study consider why organizations should use HCI in order to present their information within Data Analytics.
Every human-being is born with limbs and senses; senses to smell, taste, feel, hear, and see. Using their limbs and senses a human being is able to perform all kinds of activities such as, for instance driving a car, typing a document or making toast and eating it. When considering HCI, it is clear that the limbs and senses are needed to interact with computers. Interaction with computers is provided by using a keyboard or mouse. But are the keyboard and mouse outdated? Should we not go back to basics and make it easier for us to
control the technology simply by using voice and gestures to create human interaction?
Besides making life easier, this study has already shown that gathering business information can be useful for organizations and CIO’s in helping with decision-making. But presenting business information while maintaining the audience’s attention still present challenges, especially when presenting something to a difficult group. Jelemenska et al. (2011) undertook some research on students’ engagement in interactive presentation. They mentioned that the quality of a presentation determines the ability of students to understand a given topic and that most of the presentation applications available for large auditoriums do not support active entries into the presentation which are necessary to improve student-teacher interaction. Thus to improve the interaction between the student and the teacher or between the presenter and the audience, active entries are needed. Therefore HCI has an important role in improving presenter-audience interaction.
Although Webster & Ho (1997) have done a similar study on measuring audience engagement, they only used two different software programs to measure audience engagement. This study focuses on the integration of natural interaction both to improve engagement and to keep the audience involved for longer during the presentation. Therefore a Kinect will be used as a tool to provide natural interaction. To measure audience’s engagement the proposed model of engagement of O’Brien & Toms (2008) will be used. Although this provides a tool for defining user engagement with technology, other attributes will be adapted to their model to ensure suitability for this study. This will create an own proposed model to measure engagement.
This paper focuses then on the integration of natural interaction, or to integrate our human skills, within data analytics visualization, or today’s technology environment, by using a Kinect. Natural interaction will be used to improve the interaction between human and computer and to create a higher quality audience engagement.
This creates the following research question: Does natural interaction provide an improved engagement
when presenting “Data Analytics” information with a Kinect?
The remainder of the paper is organized as follows. The following section describes the theoretical background of the integration of natural interaction by using a Kinect. Therefore, relevant research on the use of a Kinect will be addressed. In the same section the foundation of the theoretical framework will be discussed and an own proposed model with engagement attributes will be introduced. The own proposed model of engagement provides the three phases of engagement which will be used to determine the hypotheses. In section three the study method will be discussed, as will the design of the experiment. Section four will show the results of the experiment. Section five will discuss the results and provide the conclusion of the study. Section six will address the references, and section seven comprises the appendices.
2. THEORETICAL BACKGROUND
Integrating our human skills within today’s technology environment is possible by making use of a Kinect. Kinect is a sensor that provides the ears and eyes of an application in order to create interaction with and control of your computer. It enables users to do this through a natural user interface using gestures and voice. Kinect has been developed by Microsoft for Xbox, a gaming console, to create a better gaming experience without using a controller. Since February 2012 Microsoft have had Kinect fully developed specially for Windows, which is able to construct a human-computer interaction solution. Microsoft describes this experience as: “exciting human-computer interaction solutions you once thought were
impossible”.
Lee & Oh (2014) undertook some research on a Kinect Sensor-based Windows Control Interface. In their study they propose an intelligent interface for the Windows OS using Microsoft Kinect instead of using a mouse or keyboard. Their system provides humans with the ability to interact with the computer using voice and hand gesture recognition, and their study showed that the combination of motion and voice not only replaced the mouse but also provided an interface for users to add more operations to by employing a registration method. They concluded that users can control their computer more easily, depending on their needs. They also measured the user satisfaction of the system interface, the recognition feature, the register mode and the input mode. Evaluations from the participants showed that they were satisfied, except with the input mode. The overall level of satisfaction proved its acceptability.
Hansen (2004) undertook some research in hospitals on interaction without keyboard and mouse. He mentioned that the work in hospitals differs from office work in that it is more mobile and undertaken in a specialized environment that limits the possibility for interaction with computers. Currently employees are dealing with different working styles, such as working from home, mobile working, and places with flexible workspaces so that employees have a greater choice on where they will work regardless of time, day or location, whether inside or outside the office. The advent of modern technologies such as smartphones and tablets in the work-place, has meant that a vast majority of professionals now use these devices as their primary working tool thereby integrating corporate data and personal data in the process that is also known as BYOD, Bring Your Own Device (Self & Kestle, 2013). According to Johnson (2012) one of the reasons for BYOD is its affordability, as well as the experience of power that ubiquitous communication and information access is able to confer. Organizations recognize and embrace this more flexible way of working, such as Hansen has mentioned.
In the light of Lee & Oh and Hansen, this study considers why users still employ a keyboard and mouse while they make increasing use of mobile devices that are touch-controlled, or swiped with hands, or voice-controlled. Dealing effectively with all the current technology requires a simplified usability. Formerly humans only used their hands and their feet to get things done. Life could be so much easier now, if we could just make more optimal use of the current technology. Increasing human computer interaction is therefore required, and this entails less reliance on remote controls, keyboard and mouse in return for simply using our voice or gestures to control the technology. As Sanders (1987) hasmentioned in his article about human factors in engineering and design: “our ancestors lived in an essentially “natural” environment in
which their existence virtually depended on what they could do directly with their hands (as in obtaining food) and with their feet (as in changing prey, getting to food sources, and escaping from predators)”. This
study then goes back to basics and integrates our human skills within today’s technology environment in order to control it by using a Kinect.
When integrating natural interaction greater attention should be paid to natural interaction within interactive data visualization. Bloomberg’s research (2011) has already shown that interactive data visualization can be seen as a data analytical tool, but is simply not sufficiently appealing. Therefore the level of interaction requires improvement. To do this, this study focuses on user engagement. According to Attfield & Kazai (2011) user engagement is “a quality of the user experience that emphasizes the positive aspects of
interaction – in particular the fact of being captivated by the technology”. When such user engagement is
involved with a technical resource Attfield & Kazai mention that “it is not just about how a single
interaction unfolds, but about how and why people develop a relationship with technology and integrate it into their lives”.
Webster & Ho (1997) argue that users describe their feelings when interacting with engaging software as curiosity, interest, confidence, and surprise. Thus engagement focuses on the quality of user experience and emphasizes the interaction by describing the user’s feelings. Therefore engagement itself will be measured in order to improve the interaction when presenting Data Analytics information.
Over the last few decades much research has been done on engagement. Particularly in human-computer interaction, research engagement appears frequently. Homer, et al., (2014) made a study on the effects of interactivity in a Kinect-based literacy game for learner readers. The results showed that readers from the Kinect group had a higher interest and engagement during the study’s activities. McCrindle, et al., (2014) discovered that the Microsoft Kinect can engage a patient and encourage them to follow a post-hospital rehabilitation program in the case of brain trauma injuries. Instead of following the formal program for brain injury rehabilitation, a patient with a traumatic head injury followed a rehabilitation program with a Kinect. The results showed that the patient’s engagement had many positive benefits. The patient also demonstrated vocal improvements as well as improvements in his ability to sustain and to switch his attention.
Thus interaction with a Kinect will increase engagement. Although Jacques (1995) mentioned that “too
much engagement can be a disadvantage. Games for example, can be very compelling and distract users from their initial learning goal”. This can also occur with a Kinect. Disengagement will then arise from
excessive engagement or by frequent use of the Kinect. The frequency of use is directly related to the degree of novelty. According to Martínez-Román & Romero (2013) several factors influence the degree of novelty. They claim that it depends on the stage of the product life cycle in which the innovation is introduced. Attfield & Kazai (2011) assert that novelty appeals to our sense of curiosity, encourages inquisitive behavior and promotes repeated engagement. “It can arise through freshness of content or innovation in
Vázquez & Yamanaka (2014) made a study that involved the appraisal of novelty on users’ emotions, and an evaluation of user experience with commercial products. Participants in their study were asked if they had seen and if they had used the products before. The results showed that participants who had not seen a product before experienced an exceedingly high degree of novelty compared to participants who had already seen it. These latter participants experienced the products as “predictable” and “easy to understand”. Yvonne Eve Walus, an education specialist and a senior consultant to Creative Learning Systems in Auckland, wrote an article about this called: Let the games commence1. She made a study of different
gaming consoles in order to assess activity interest and levels. She mentioned that the novelty of the Kinect would eventually wear off unless she kept on supplying new games.
O’Brien & Toms (O’Brien & Toms, 2008) also noticed that novelty played an important role in their study on defining user engagement with technology. They mention that novelty is needed to sustain user engagement. Video gamers, for instance, would look for unexpected auditory or visual stimuli, and web shoppers were interested in seeing new products.
Thus, despite the fact that previous research has shown that a Kinect will increase engagement, disengagement can occur after a long-term use due to lack of novelty. The Kinect then becomes dull, lacking all interest, or even worse it could possibly even distract. In cases such as this the user will need to decide for what purpose he or she will use the Kinect; either to sustain or increase engagement.
In order to discover whether the use of a Kinect will increase engagement a proper foundation for measuring that engagement is required. A great deal of research has therefore been done and numerous methods have been proposed to measure user engagement. Little has been done however to validate and correlate these measurements and so provide a firm basis for assessing the quality of user experience (Attfield & Kazai, 2011). O’Brien (2006) also mentioned that no theoretical frameworks exist for articulating engagement, although several theories may be drawn upon to explore and understand it. Thus this study will address the question of a proper foundation to measure engagement, and this topic will be discussed in the next section.
2.1 CONCEPTUAL FRAMEWORK
O’Brien & Toms (2008) made a study of what constitutes user engagement and they provide a conceptual framework for defining user engagement with technology. O’Brien used a concept from his framework in an earlier study to measure user engagement with information systems. In this study O’Brien (2006) defines engagement as a “dynamic process in that it reflects the way in which the two-way interaction between the
user and the interface changes and responds to the other”. O’Brien & Toms first determined the attributes
of engagement after an extensive literature study. The results are shown in Figure 1. The attributes of engagement are based on four different theories, flow theory, aesthetic theory, play theory and information interaction theory. They used the theories to inform their understanding of user experiences.
According to O’Brien & Toms the flow is the condition in which people are so engrossed in an activity that nothing else seems to matter; the experience itself is so enjoyable that people will do it even at great cost, for the sheer sake of doing it. The flow theory has been used “to explore situational and personality
variables associated with computer-based task, to predict and design for flow experiences, and to understand users ‘reactions to and motivations for using applications” (O’Brien & Toms, 2008). They
suggest that engagement may share some attributes with flow, such as focused attention, feedback, control, activity orientation, and intrinsic motivation.
The second theory focuses on the aesthetic. According to Jennings (2000) aesthetic experiences occurs
“when a person is sagaciously engaged and immersed in an activity. It is perceived as an experience that is sensuous and singularly unique”. Aesthetics is important for engagement, but clearly is only one aspect
of engagement according to O’Brien & Toms (2008).
The third theory is the play theory. Play is difficult to define but it has the following attribute: “it involves
some level of active, often physical, engagement” (Rieber, 1996). O'Brien & Toms mentioned that play has
been associated with increased frequency and satisfaction of system use, and has been attributed to increased motivation, challenge, and affect. “Thus elements of play are intrinsic to engagement”. The fourth theory is information interaction. “Information interaction is specifically the process that people
use in interacting with the content of a system” (O’Brien & Toms, 2008). They argue that it is the interaction
between users and systems operating within a specific context that facilitates an engaging experience.
Figure 1. Attributes of flow, aesthetic, play, and information interaction theories, and proposed relevancy to engagement. (O’Briens & Toms, 2008)
O’Brien & Toms (2008) table maps the characteristics that emerge from their literature of flow, aesthetics, play, and information interaction theories. In the final column they hypothesize which of the attributes may be intrinsic to engagement. Based on their exploration of the four theoretical frameworks and research on engagement, an understanding of engagement emerges. They argued that “the theoretical framework that
informs engagement integrates system attributes (i.e., feedback, challenge, sensory appeal) with the affective propensities of the user”.
Their second step is to explore engagement in four application areas: online shopping, web searching, educational webcasting, and video games. The rationale for choosing these areas is encapsulated in the following: “studies of engagement exist predominantly in education and video games, while there is sparse
research about engagement in online shopping and web searching”. However, O’Brien & Toms provide a
conceptual framework for defining user engagement with technology. They focus on four non-comparable areas of application for this study.
Webster & Ho (1997) produced a study called “Audience Engagement in Multimedia Presentations”. In this they describe engagement and propose that it is conceptually similar to the state of playfulness, a construct increasingly applied to studies of human-computer interactions. It describes how multimedia technologies can influence listeners' engagement during presentations. They developed seven attributes in order to measure engagement: feedback, control, variation, attention focus, curiosity, and intrinsic interest.
They excluded the item control, because it was the presenter and not the participants that controlled the technology. The attributes used to measure engagement were based on earlier research (Webster & Ho, 1997).
Because the attributes of engagement suggested by O’Brien & Toms focus on four areas of applications that are not fully compatible with this study, the attributes of measurement have been integrated based on Webster & Ho (1997) to provide attributes of engagement for presenting Data & Analytics information. The results of the second step, exploring engagement in the Data & Analytics presentation area, are shown in the final column of Figure 2.
Figure 2. Attributes of engagement suggested by O'Brien & Toms (2008) and Webster & Ho (1997)
O’Brien & Toms (2008) final step focuses on other possible other attributes of engagement. Their previous research had identified all the possible attributes of engagement, but they also mentioned being insufficiently convinced to determine all the recognized attributes. Therefore they carried out qualitative research, interviewing 17 participants, in order to examine engagement in general by using the four applications. The result provides a process of engagement with distinguishable attributes inherent at each stage in the process. The steps in the engagement process are as follows: the point of engagement, period of engagement, disengagement and reengagement. Based on their interview results, O’Brien & Thomas indicate both the attributes and manifestations of the experiential threads present during each stage. The outcome of their study is a proposed model of engagement with the identification of four distinct stages of engagement as well as the attributes that characterize each stage.
Figure 3. Proposed model of engagement without its attributes (O’Brien & Toms, 2008)
Although O’Brien & Toms helpfully provide a conceptual framework for defining user engagement with technology, another combination of attributes will be adapted into the framework to make it useful for this study. The attributes will be integrated into the different stages of engagement. Those attributes consist of elements drawn from the theories shown in Figure 1 and from Webster & Ho (1997) shown in Figure 2. The following tables define the attributes of each stage that provide a foundation for the measurement of engagement in this study.
Point of Engagement
Attribute Definition
Aesthetics Visual beauty or the study of natural and pleasing (or aesthetic) computer-based environments (Jennings, 2000).
Attention Focus The concentration of mental activity; concentrating on one stimulus only and ignoring all others (Matlin, 1994)
Curiosity A psychological attribute assigned to thrill seeking and/or information seeking, which may include interpersonal dimensions (Glass & Ogle, 2013a)
Intrinsic Interest
Relating to the essential nature of a thing; inherent (Oxford dictionary) A state of curiosity or concern about or attention to something (Oxford dictionary)
Motivation Elements that bring about focus or a desire to proceed with an activity (O’Brien & Toms, 2010)
Specific or experiential goal
The object of a person’s ambition or effort; an aim or desired result (Oxford dictionary)
Period of Engagement Attributes
Attribute Definition
Aesthetics and
Sensory Appeal
Visual beauty or the study of natural and pleasing (or aesthetic) computer-based environments (Jennings, 2000).
Relating to sensation or the physical senses; transmitted or perceived by the senses (Oxford dictionary)
Challenge A call to engage in a contest, fight, or competition (Oxford dictionary) Feedback Information that is sent back to the user about what action has been done
or what result has been accomplished (O’Brien & Toms, 2010)
Control How “in charge” users feel over their experience with the technology
(O’Brien & Toms, 2010)
Novelty/Variety Variety of sudden and unexpected changes (visual or auditory) that cause excitement and joy or alarm (Aboulafia & Bannon, 2004); Features of the interface that ―users find unexpected, surprising, new, and unfamiliar
Table 2. Period of Engagement attributes
Disengagement Attributes
Attribute Definition
Challenge A call to engage in a contest, fight, or competition (Oxford dictionary) Positive Affect To have a positive influence on or effect a change in (Oxford dictionary) Negative Affect To have a negative influence on or effect a change in (Oxford
dictionary)
Interruptions Something that interrupts, such as a comment, question, or action (Oxford dictionary)
Table 3. Disengagement attributes
Based on the “Proposed Model of Engagement” (O’Brien & Toms, 2008) combined with attributes from Webster & Ho (1997) an own model of engagement is defined to identify the key components that constitute engagement for this study.
This model no longer contains a cycle of engagement; such a cycle is unnecessary for this study because there is no requirement to reengage. In this study we focus on the way the participants become engaged, remain engaged, and why they should disengage. Therefore the crucial difference between a Kinect and a traditional presentation ought to be proven. Whatever presentation method that participants prefer should clearly emerge in the end. Reengagement is therefore superfluous.
The attributes of engagement that are gathered are based on the scope of this study. Therefore attention will be given to audience engagement when presenting information.
Figure 4. Final model of engagement with attributes
In addition to what has already been said, the attributes of varying levels of intensity will be changed. Only the five attributes during the period of engagement will be measured on levels of intensity. Based on these attributes it can be determined which attribute provides the highest level of intensity. The results of this new model of engagement are presented in Figure 4.
2.2 HYPOTHESES
Based on the above literature study and the own proposed model of engagement, the hypotheses will be discussed. Although the model provides a complete process by which to measure engagement, the three phases will be used to measure the hypotheses. Each phase consists of several attributes which will be used to measure the level of engagement. The attributes, according to Webster & Ho (1997), will be identified for each phase that comprises engagement between the participant and the application (presentation).
2.2.1 POINT OF ENGAGEMENT
The point of engagement is the point at which the participant experiences becoming fully engrossed in a task or in interface attributes. “The point of engagement is initiated by the aesthetic appeal or novel
presentation of the interface, the users’ motivations and interests, and users’ ability and desire to be situated in the interaction and to perceive that there is sufficient time to use the application” (O’Brien &
Toms, 2008). Jacques (1995) argued in his study on engagement as a design concept for multimedia that “Learners are “engaged” with educational multimedia when it holds their attention and they are attracted
to it for intrinsic rewards”. As well as involving naturally curiosity, also known as the fundamental
component of optimal human functioning, engagement is considered instrumental in human problem solving and development (Glass & Ogle, 2013). Glass & Ogle mentioned that the key components of human curiosity are engagement, interest and motivation. Malone & Lepper (1987), argued that the sense of curiosity can be increased by providing variations in video or audio effects. Thus to gain the attention of the audience, presentations need to be attractive and have a strong appeal for the curious. Curiosity can be increased by using a Kinect because it provides visual effects by gesture control (video effect) and also provides voice control (audio effect).
(H1) The point of engagement attributes will let the participants experience a higher level of engagement with a Kinect presentation than during a traditional presentation.
2.2.2 PERIOD OF ENGAGEMENT
The period of engagement is the period during which the participant is fully engaged or involved with the interface or task. “Engagement is sustained when users are able to maintain their attention and interest in
the application, and is characterized by positive emotions” (O’Brien & Toms, 2008). According to O’Brien
& Toms feedback was shown to be an essential component for inciting and maintaining engagement. They mentioned that users must be made to feel part of the interaction through an awareness of what the system is doing (feedback). Chapman (1997) found that feedback has influences on engagement. Price et al. (2011) argued that feedback is a two-way process, a dialogue, in which information is exchanged, interpreted and transformed. In addition to feedback, novelty and variety provide engagement by presenting information. In the study made by O’Brien & Toms (2008) novelty played a significant role in sustaining user engagement with the applications. They mentioned that lack of novelty caused participants to disengage. Attfield & Kazai (2011) argued that learners experience higher levels of engagement during multimedia presentations that exhibit higher levels of variety. Berlyne (1960) claim that variety in multimedia will relate directly to flow, because novel, surprising, and complex stimuli preserve sharpness of attention. Although variety provides a higher level of engagement and stimuli preserve attention, O’Brien & Toms (2008) found that participants did not want too much variety, otherwise they became “lost”. Webster & Ho (1997) concluded that feedback, novelty and variety are attributes of engaging presentation software. Thus the period of engagement can be sustained when feedback, novelty and variety are provided. Although the Kinect holds the Guinness World Record as the fastest selling consumer electronic device, with sales surpassing 10 million units as of 9 March 2011 (Boulos et al., 2011), most users no longer need it. According to Peter Lee, owner of ComputerHowtoGuide.com, commercial applications have yet to be released2. He forecasts that when more software is released the Kinect could very easily be built into laptops
in the same way that webcams currently are. This study assumes that most people consider the Kinect a “novelty” device, because it is seldom used for business purposes. Therefore the level of novelty will be higher when giving a presentation with a Kinect. Aside from this, a “traditional” presentation can only controlled by a mouse or keyboard click in contrast to the Kinect, which enables control of a presentation with any kind of a human gesture or any chosen voice command. This will create greater interactive variety when giving a Kinect presentation.
(H2) The attributes of period of engagement provide a higher level of engagement during a Kinect presentation than during a traditional presentation.
2.2.3 DISENGAGEMENT
Disengagement is the period that the participant stops a task due to an event in the external environment or for voluntary or personal reasons. “Users disengage for many reasons such as the usability of the
technology, and distractions in their environments” (O’Brien & Toms, 2008). Disengagement is associated
with positive emotions when users need to become motivated once more to enable them to return with renewed alacrity. It is also associated with negative feelings of frustration, uncertainty, loss of interest or motivation, and lack of novelty (O’Brien & Toms, 2008). Disengagement can occur at any time and is conveniently divided into internal and external factors. Examples of external factors include participants visiting a bathroom or receiving a phone call. Internal factors can arise when the period of engagement is no longer sufficient or when there are some technical problems. When there is an unexpected interruption like a phone call, or a participant needs to go to a meeting, it does not matter whether they are attending a traditional or a Kinect presentation, disengagements of this kind can occur at any time. The challenge is ensure that the participant will not disengage; interruptions therefore should be avoided. To achieve this there should at least be no technical problems. Due to the novelty of the Kinect and the fact it is still little used, problems with it are anticipated. Although these are more likely to be technical problems, the Kinect will certainly provide more positive experiences through motivation and curiosity, and any common technical problems that cause disengagement will be likely to occur during a traditional presentation.
(H3) Distraction provides a higher level of disengagement during a traditional presentation than during a Kinect presentation.
3. METHOD
3.1 DESIGN
This study is based on an independent samples t-test in which each of the participating groups attended both presentations. Both presentations contain information based on text, graphs, tables, pictures, and videos equivalent to Data & Analytics information. The presentations can be defined by two conditions. The first condition represents a PowerPoint presentation controlled by a keyboard and mouse, which can be defined as a “traditional presentation”. The second condition represents a PowerPoint presentation controlled by human gestures and voice, which can be defined as a “Kinect presentation”. Each presentation took approximately 10 minutes. There was a short break between the presentations. After each one was finished the participants were asked to complete a questionnaire, which can be found in Appendix A.
3.2 PARTICIPANTS
There were three groups (A, B, and C) of participants. Two groups (group A and B) consisted of Avanade employees3, varying from interns to full employees, and another group (group C) consisted of Schuman
employees4. No distinction was made between age, gender or function. 41N (N=Participants) took part in
the study. In total there were two presentations. Each group participated in both presentations, the one controlled with keyboard and mouse and the other controlled with human gestures and voice using a Kinect. The subject of the presentation was the opponents of the Dutch team for the forthcoming football World Cup in Brazil. This was chosen in order to attract and excite participants. Group A began with the traditional presentation in which the Chilean opponents were discussed. The second presentation was made using the Kinect, and in which the Australian opponents were discussed. Group B then commenced with the Kinect presentation on the Chilean opponents, and their second one was the traditional presentation on the Australian opponents. Group C then started once again with the traditional presentation.
3Avanade is a joint venture between Accenture and Microsoft that provides IT service focused on the Microsoft platform. http://www.avanade.nl 4Schuman is a collection agency and judicial office. http://www.schuman.nl
3.3 MATERIALS
3.3.1 MICROSOFT KINECT
The Kinect enables users to interact and control with their computer through a natural user interface using gestures and voice. Therefore the Kinect employs complex technical properties (Microsoft Kinect Interface Guidelines, 2013). The Kinect “near mode depth ranges” are 0.4m to 3m versus the “default mode depth ranges” 0.8m to 4m. The Kinect can track up to six people within its view, including two whole skeletons. The Kinect can track 20 joints in a full skeleton mode. In a seated mode it will track only the 10 upper joints. The Kinect detects audio input from + and – 50 degrees in the front of the sensor. The microphone array can be pointed at 10-degree increments within the 100-degree range. The Kinect can cancel 20dB of ambient noise, which is the sound level of a whisper and it can also detect the loudest source within a raucous environment.
Figure 5. Microsoft Kinect for Windows
3.3.2 MICROSOFT KINECT FOR WINDOWS SDK
The Kinect for Windows SDK (Software Development Toolkit) enables the use of C++, C#, or Visual Basic to create applications and experiences that support gesture and voice recognition by using the Kinect for a Windows sensor and a computer or embedded device.5 For this study Visual Studio is used to develop the
Kinect application. Visual Studio is a Microsoft product which is used to develop computer programs for a Microsoft Windows platform, as well as web applications, websites, and web services.
3.3.3 PRESENTATIONAID
An existing Kinect application provided by CodePlex is used. “CodePlex is Microsoft’s free open source
project hosting site. You can create projects to share with the world, collaborate with others on their projects, and download open source software”6
. The application has been developed by Dal Rupnik, PhD Student at the University of Maribor. The application, called PresentationAid, is an application that connects the most common body gestures read by Microsoft Kinect sensor to a small engine that converts them into keystrokes or presentation software commands7. For this study voice control is added to
PresentationAid to enable voice control commands to control a presentation software using voice. The voice commands for PresentationAid are: “Start presentation” to start the presentation, “next slide” to go to the next slide, “previous slide” to go to the previous slide, and “stop presentation” to stop the presentation.
3.3.4 MICROSOFT POWERPOINT
PowerPoint is a slide show presentation program developed by Microsoft that provides numerous features offering flexibility and the ability to create a professional presentation. Although PresentationAid makes it possible to control multiple presentation programs, PowerPoint is used only because it is still the most widely used program.
5http://www.microsoft.com/en-us/kinectforwindowsdev/start.aspx 6http://www.codeplex.com/
3.3.5 TEST PROCEDURE
The presentations for both groups were presented by one presenter, who, during the presentations for Group A, wore the same clothes throughout in order to avoid personal distractions of any sort. The same was done for Group B and Group C. The Kinect reacted only to the presenter and not the audience. Before the Test was started the participants were asked if they had experienced a Kinect-controlled presentation before. Only one girl from Avanade had experience with a Kinect-controlled presentation. Group A and C started with the traditional presentation. When this was finished the participants were asked to complete a questionnaire, after which there was a five minute break. The Kinect presentation then commenced. In group B the presentations were reversed. Group B started with the Kinect presentation and finished with the traditional one. When the Kinect presentation was finished the participants completed the same questionnaire. Note that the presentations were given in Dutch but the questionnaire was in English. Figure 6 shows the setup of both test procedures.
Traditional Kinect
Figure 6. Setup test procedure
3.3.6 QUESTIONNAIRE
One questionnaire was used during this study (Appendix A). It consists of twenty-six questions on engagement. These questions were answered based on a five point Likert-type scale. The questions were divided into three phases of engagement. They addressed the engagement of the participants during the presentation and were the same for both conditions. The age and gender of the participants was also asked, and participants were permitted make any comments they wished to when they reached the last question. Fresh questionnaires, which were of paper, were completed after each presentation. After the results had been gathered in, the results were manually entered into Excel sheets in order to import them in SPSS8.
SPSS was used to analyze the data. To analyze the data, the outcomes of question 24, 25 and 26 were reversed because those questions are negative in contrast to the other questions.
3.4 DATA ANALYSIS
The engagement data were analyzed with crosstabs (chi-square) and with independent samples t-test. The crosstabs were used in order to determine whether the two conditions differ from each other with each engagement attribute. Then the independent samples t-test was used in order to determine the different phases of engagement between the two conditions. The phases of engagement were assessed based on Cronbach’s α to determine whether multiple attributes may form a phase of engagement together.
8The data was analyzed using IBM’s SPSS Statistics
Table Table
Laptop Laptop Kinect
Screen Screen
Presenter Presenter
4. RESULTS
First the results of the attributes of engagement will be presented. Next, the phases of engagement results will be analyzed and then the conditions will be compared in order to determine any significant difference.
4.1
ATTRIBUTES OF ENGAGEMENTThe proposed model of engagement consists of 15 attributes. Each attribute provides a component that measures engagement, and multiple attributes consist of phases of engagement. First, the attributes of engagement will be analyzed based on the questions that were asked in the questionnaire.
The questionnaire was analyzed using a chi-square, based on cross tables, to determine whether the conditions differ from each other from question to question. The results show that three questions provide a significant difference. The first is question 18: “This presentation allows the presenter to maintain in
control over the direction of the presentation”. The result shows that none of the participants strongly
disagreed during either presentations. Additionally, during the Kinect presentation not one of the participants disagreed. Exactly 92.7 % of the participants either agreed or strongly agreed. Although more participants agreed during the traditional presentation, fewer participants strongly agreed. The result of question 18 provides a significant difference of .020. This shows that the presenter retains greater control over the direction with a Kinect presentation than with a traditional presentation. The results are presented in Table 4.
Strongly agree Agree Neither Disagree
Traditional Count 7 24 5 5
% within Group 17.1% 58.5% 12.2% 12.2%
Kinect Count 17 21 3 0
% within Group 41.5% 51.2% 7.3% 0%
Table 4. Cross table question 18
The second question with a significant difference is question 19: “This presentation incorporates changing
presentation styles”. The results show that 10 participants disagreed during the traditional presentation as
against only one during the Kinect presentation. None of the participants strongly disagreed during either presentations. Just as many participants agreed during both presentations. However 18 participants strongly
agreed as against only one participant during the traditional presentation. The results are shown in table 5.
Strongly agree Agree Neither Disagree
Traditional Count 1 19 11 10
% within Group 2.4% 46.3% 26.8% 24.4%
Kinect Count 18 19 3 1
% within Group 43.9% 46.3% 7.3% 2.4%
The result of question 19 provides a significant difference of .000. Thus a Kinect presentation incorporates more changing presentation styles than a traditional presentation.
The final question to provide a significant difference is question 20: “This presentation uses a variety of
presentation styles”. The result of this question shows that 14 participants disagreed or strongly disagreed
during the traditional presentation as against only two during the Kinect presentation. In addition, more participants agreed or strongly agreed. There was a total of 80.5% for the Kinect presentation as against 36.5% for the traditional presentation. The participants experienced a greater variety of presentational styles. The results are presented in Table 6.
Strongly
agree Agree Neither Disagree
Strongly disagree Traditional Count 1 14 12 13 1 % within Group 2.4% 34.1% 29.3% 31.7% 2.4% Kinect Count 9 24 6 2 0 % within Group 22.0% 58.5% 14.6% 4.9% 0%
Table 6. Cross table question 20
This results in a significant difference of .000 on question 20. Thus a Kinect presentation uses a greater variety of presentational styles than a traditional presentation. The results of the three significantly difference questions can be found in Appendix B. Now that the results of the questions have been given, the attributes of engagement will be analyzed to determine which attribute provides the highest level of engagement. The attributes are analyzed and an overview of the results are presented in Appendix C.
4.1.1 ATTRIBUTES RANKING
Based on the results the attributes can be ranked from strongly engaged to strongly disengaged. Therefore the mean of each attribute is calculated in order to determine the attributes ranking. The scale of mean is between one and five. This is derived from the questionnaire based on a Likert-scale. One is strongly agree and five is strongly disagree. The results are presented in Figure 7.
The results in Figure 7 show what the top five represents for the traditional presentation: from one to five,
Aesthetics, Intrinsic Interest, Interruptions, Control, and Aesthetics Sensory. The attribute of Aesthetics
appears twice in this ranking. This means that the participants were pleased, the presentation won their attention and that they were engaged by the “Nature of Art” of the presentation. Interruption, which is part of disengagement and occurs often during a traditional presentation, is also ranked in the top five. Interruption easily leads to distraction.
For the Kinect presentation thee top five represents the following: Control, Intrinsic Interest,
Novelty/Variety, Aesthetics, and Attention Focus. The attribute of control provides the highest level of
engagement. The attributes of intrinsic interest and aesthetics also occur during a Kinect presentation. In contrast to the traditional presentation, the Kinect provides considerable variety and can be considered as novel.
4.2 PHASES OF ENGAGEMENT
The model of engagement defines three phases of that engagement. To determine the reliability of each phase, Cronbach’s α is applied. Cronbach’s α is a method used to measure the internal consistency of multiple items. For this study Cronbach’s α determines if the multiple questions may be framed together as a phase of engagement. This will measure the reliability of the variables; phase of engagement. Therefore each phase needs a minimum Cronbach’s α value of .700. The first phase point of engagement consists of questions 1 to 11 and it has a Cronbach’s α value of .913. The second phase is defined as the period of engagement, and consists of questions 11 to 21. It has a Cronbach’s α value of .884. The final phase is defined as disengagement. Disengagement comprises questions 21 to 26 and it has a Cronbach’s α value of .732. The results are presented in table 7.
Table 7. Reliability based on Cronbach’s α
Thus the three phases of engagement provide sufficient reliability in order to be divided into different phases. Now that the phases of engagement have been determined, the independent samples t-test is applied to measure the engagement levels between the traditional presentation and the Kinect presentation.
4.2.1 INDEPENDENT SAMPLES T-TEST
The independent samples t-test measures any significance difference on the three phases of engagement between the conditions. Both conditions, traditional presentation and Kinect presentation, had a total of 41 participants (N=41). The total mean during each phase and each condition is between 2 and 2.59, which means that the total average of all the answers on engagement is rated as agreed. The standard deviation shows no higher a value than .639 which means that it will not affect the rates. The results are presented in Table 8. Phases of Engagement Cronbach's Alpha N of Items Point of Engagement .913 10 Period of Engagement .884 10 Disengagement .732 6
Condition N Mean Std. Deviation Std. Error Mean
Point of Engagement Traditional 41 2.3537 .62654 .09785
Kinect 41 2.0390 .56297 .08792
Period of Engagement Traditional 41 2.5878 .59590 .09306
Kinect 41 2.0756 .49131 .07673
Disengagement Traditional 41 2.5203 .63922 .09983
Kinect 41 2.2236 .50223 .07844
Table 8. T-test statistics
The hypotheses of this study will now be analyzed. The same independent samples t-test was therefore used. The results are presented in Table 9.
The independent samples t-test provides a Levene’s test outcome and a t-test for e-quality of means outcome. The Levene’s test checks whether or not the variances of the groups are equal. This is to assess the equality between the traditional group and Kinect group. Therefore their will not be assumed that variances of the populations from which different samples are drawn are equal.Based on the Levene’s outcome it is possible to determine whether the equal variances that are assumed or the equal variances that are not assumed should now be used.
The Levene’s test shows that there is no significant equality in variances. The significance is .772 at the point of engagement, .086 at the period of engagement, and .100 at disengagement which means that for all the phases of engagement the significant value is higher than .05. Thus the equal variances assumed row will be used.
The column t-test for equality of means provides the significance results of the experiment. First we look at the mean difference. The mean difference gives the difference between the answers for both conditions. The first phase, the point of engagement, has a mean difference of .314. This means that the given answers between the conditions do not differ that much. The point of engagement provides a significant value of .019 which is lower than .05. Thus the point of engagement provides a significant value between both conditions.
The period of engagement has a mean difference of .512. Here also the given answers between the conditions do not differ very much. The period of engagement provides a significant value of .000 which is lower than .05. Thus the period of engagement also provides a significant value between both conditions. Disengagement has a mean difference of .296 which provides the lowest difference between the response conditions. Disengagement provides a significant value of .022. Based on these results this study shows that all the phases of engagement show a significant difference.
Levene's Test for Equality of
Variances
T-test for Equality of Means
95% Confidence Interval of the Difference F Sig. t df Sig. (2-tailed) Mean Diff. Std. Error
Difference Lower Upper Point of Engagement Equal var. assumed .085 .772 2.392 80 .019 .31463 .13155 .05285 .57642 Equal var. not assu. 2.392 79.102 .019 .31463 .13155 .05280 .57647 Period of Engagement Equal var. assumed 3.017 .086 4.246 80 .000 .51220 .12062 .27216 .75223 Equal var. not assu. 4.246 77.195 .000 .51220 .12062 .27203 .75236 Dis-engagement Equal var. assumed 2.776 .100 2.337 80 .022 .29675 .12696 .04410 .54940 Equal var. not assu. 2.337 75.758 .022 .29675 .12696 .04388 .54962
Table 9. Independent Samples T-test
4.3
QUESTIONNAIREThe last question of the questionnaire provided a comment box to enable participants to add any comments on their experience during the experiment. First, the comments on the traditional presentations will be addressed.
From the 41 participants, four participants left a comment. Two participants stated that the subject of the presentation was just not interesting enough. One participant suggested using a pointer to control the PowerPoint presentation. The last participant said that he noticed no difference between the traditional presentation and the Kinect presentation.
With the Kinect presentation six participants left a comment. One participant commented that the Kinect provides less appeal than expected. Another participant mentioned that the natural gestures were too distracting. The other four participants were positive, mentioning: “Nice natural movements”, and “Nice!”, “Kinect control seems more interesting than traditional control”, and “Smooth presentation”.
5. DISCUSSION AND CONCLUSION
5.1 HYPOTHESES
Based on the study results the hypotheses will now be discussed. The first hypothesis stated that the point
of engagement attributes will let the participants experience a higher level of engagement with a Kinect presentation than during a traditional presentation (H1). For this hypothesis 10 questions were asked to
determine significance. The results show that the point of engagement scored significantly high with a value of .019. This means that the Kinect presentation proved more appealing and won the participants’ attention. Extra attention was also given to the Kinect presentation as it excited curiosity and was generally more fun. Thus this study proves that participants experienced a higher level of engagement with a Kinect presentation. This means that the point at which the average participant experienced engagement in a task or interface attributes is higher than that for a traditional presentation. Hypothesis 1 can therefore be accepted.
The second hypothesis stated that the attributes of period of engagement provide a higher level of
engagement during a Kinect presentation than during a traditional presentation (H2). For this hypothesis
10 questions were asked to determine significance. The results show that the period of engagement scored significantly high with a value of .000. This means that the Kinect presentation was more appealing and won the participants’ attention during the presentation. The Kinect presentation also provided a greater challenged to participants and encouraged them to think. They also experienced the Kinect presentation as novel and varied. Thus this study proves that participants experienced a higher level of period of engagement during a Kinect presentation than during a traditional presentation. Hypothesis 2 can therefore be accepted.
Although Hypothesis 2 can be accepted now that proof has emerged that the Kinect provides a higher level of period of engagement, it still cannot be known how long that period of engagement is. The total duration of the presentation was approximately 10 minutes. The period of engagement could decrease, possibly through distraction, after 15 minutes or later with a Kinect presentation. Therefore it is not just that the presentation time requires measurement; it should also be held with the same participants in order to discover if the Kinect itself distracts. This limitation will be discussed in the next section.
The third hypothesis stated that distraction provides a higher level of disengagement during a traditional
presentation than during a Kinect presentation (H3). For this hypothesis six questions were asked to
determine significance. Disengagement scored significantly high with a value of .022. The third hypothesis allowed for the enjoyment of the participants, alleviated boredom and avoided distractions likely to lead to disengagement. The participants did not encounter any technical issues with the Kinect during the presentation. Although just one participant mentioned distraction due to the Kinect itself, the other participants responded enthusiastically. This was mentioned in section 4.3. Thus this study proves that participants experienced a higher level of disengagement during a traditional presentation than during a Kinect presentation. Hypothesis 3 can therefore be accepted.
The three hypotheses can all be accepted based on the results of the experiment. Thus this study shows that at point of engagement and period of engagement the Kinect provides better engagement than a traditional presentation. Additionally, disengagement provide more engagement during a Kinect presentation. This will answer the central question of this study: Does natural interaction provide an improved engagement
when presenting “Data Analytics” information with a Kinect?
This study shows that natural interaction by use of a Kinect provides better engagement when presenting equivalent Data & Analytics information. Although the results are promising, the study has a number of limitations which will be discussed in the next section.
5.2 STUDY LIMITATIONS
5.2.1 DURATION OF THE STUDY
The total duration of the study was limited. Within the total time occupied by the study the Kinect application had to be developed. During the initial phase it was noticed that there was insufficient time or skills to develop an own Kinect application. Therefore we opted for an existing application. Several applications had been tested and the most useful one was selected. Based on this the experiment was set up. Any limitations arising from these conditions will be explained in section 5.3.2. In addition to the limitations of the Kinect application, one interesting measuring attribute has been omitted due to a lack of time. As already mentioned in the literature study, there is an expectation that after a period of use the Kinect will lose its novelty value and probably distract during a presentation, resulting in disengagement. Over a longer period this will need to be measured with the same group of participants. The results of this study are based on a single Kinect experience. Thus it cannot determine for how many Kinect presentations or for how many minutes participants will disengage due to the expiry of novelty value.
5.2.2 STUDY VALIDITY
Based on the outcomes of the study the validity of the results will be discussed in order to determine if those results are well-founded and correspond accurately to the real world.
The experiment was conducted at Avanade and Schuman. 24 participants from Avanade and 17 from Schuman attended. The youngest from Avanade was 21 and the oldest was 38. Schuman’s youngest participant was also 21 but the oldest was 61. Avanade’s mean age was 25, while Schuman’s was 36. In total 19 men and 22 women took part. Only three men participated during the experiment at Schuman. The above results do not provide an equable distribution of the participating groups from the two companies. The total number of participants was 41, and although they comprised an adequate foundation in order to determine any significant difference, more participants would have increased the experiment’s reliability. Furthermore, this study focused on encouraging Data & Analytics information through the integration of natural interaction. Not all the participants were representative of the total population, although they were all business employees. Therefore more attention should be paid to participants from different companies with a Data & Analytics background.
5.2.3 PRESENTATIONAID KINECT APPLICATION
This study used the PresentationAID application to control the Kinect. The PresentationAID is based on developed keystroke engines. This provides for only a few specific movements that are supported. For example the PresentationAID supports a movement control for waving from right to left or from left to right. It does not support any movements from top to bottom. Programming in more movements could improve the range attributes and thereby increase the engagement.
Although the PresentationAID provides an adequate foundation for the control of a PowerPoint presentation, it lacks several features that might have increased the engagement. For example, the Kinect has a camera recognition system and a voice control function. Both functions could be used to automatically generate results after a vote. Or when, in a meeting, somebody makes a proposal, the Kinect could recognize the raised hands by camera or could generate the results by voice control. Another example is when entering keywords in an application, such as was done by Lee & Oh (2014). When making a presentation, decisions could be made on which keyword should bring the next slide into view, or any number of other task might be executed.
Thus the words “next slide” might be rendered obsolete through the ability to use keywords to control each slide. This voice control function could undertake the entire presentation smoother, which in turn will probably increase the engagement, as there would be fewer distractions during presentation.
