Goal-Oriented Design Methodology Applied to User     Interfaces in AmI Systems for Functional                    Assessmen...
The second method used, the AttrakDiff questionnaire ([9],        deviation indicates that the data points are spread out ...
application: “My name is Laura, I am 29 years old and I live        same, real-time registering of various parameters such...
screen mock-up, exactly the User Settings mock-up, located tothe right.                                                   ...
bar at the top with a text informing the user the screen where       During the evaluation, the users were working with th...
innovative, very intuitive and highly usable. It is expected that               [6]    W. Weber; J.M. Rabaey and E. Aarts....
Upcoming SlideShare
Loading in...5

Paper Gloria Cea - Goal-Oriented Design Methodology Applied to User Interfaces in AmI Systems for Functional Assessment


Published on

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Paper Gloria Cea - Goal-Oriented Design Methodology Applied to User Interfaces in AmI Systems for Functional Assessment

  1. 1. Goal-Oriented Design Methodology Applied to User Interfaces in AmI Systems for Functional Assessment G. Cea#1, E. Gaeta#2, M.T. Arredondo#3 # Life Supporting Technologies, Universidad Politécnica de Madrid. Ciudad Universitaria s/n, 28040, Madrid, Spain. 1 gcea@lst.tfo.upm.es 2 eugenio.gaeta@lst.tfo.upm.es 3 mta@lst.tfo.upm.es The Goal-Oriented Design methodology [1] follows theAbstract - Currently, the functional assessment of aerobic fitness general tendency of giving more importance to the user in theis very expensive, time consuming and needs to be done in process of designing and developing interactive products. Thespecialized laboratories. There are still no solutions based on method revolves around three fundamental aspects: themobile applications that allow people to perform this assessmenton their own. This paper describes the User-Interface design of characters, the goals and the scenarios. Goals are representedan Android application as a key element of the Functional by characters and these characters appear on scenarios, whichAssessment System (FAS) [4], an innovative, intuitive and usable in turn are used to achieve design solutions iteratively [3].AmI System for functional assessment of aerobic endurance. The The characters are fictitious and imaginary models createdFAS has been developed using the Goal-Oriented Design [1] by Alan Cooper‟s methodology [2], which have a name, anmethodology and is intended to be used by people who exercise appearance, a specific age, education, motivations,very often. The obtained results after evaluating the user preferences, life histories, and so on. There must not be someinteraction with the system have been satisfactory. arbitrary attributes, but should be based on studies of real I. INTRODUCTION users and real-world observations [2]. Cooper‟s reasoning is that the „user‟ concept is too broad, and while a user canNowadays, there is neither the knowledge nor the valid assume all kinds of knowledge, attitudes and behaviours, amethods to carry out the functional assessment of aerobic real and concrete person cannot [1]. Thus, he suggestedendurance without medical support in a simple, friendly and designing for a few people, with specific needs, goals,intuitive way. In general, this assessment analyses the knowledge and context. That is, the method replace thefunctional adaptation of the organism when it undergoes to a definition of average user, for an accurate representation: acertain physical effort. The aerobic capacity of the organism model of user called „Persona‟.to maintain a given exercise intensity for a fixed time is a The character-based scenarios are concise narrativedirect reflection of cardiorespiratory functional capacity. The descriptions of one or more characters using a product tobest quantitative measure of this capacity is the maximum achieve specific goals [1]. Thus, designers play withoxygen consumption (VO2max). Normally, this evaluation is characters‟ role and the characters in these scenarios, similarcarried out by means of stress testing [5], which is expensive to how the actors play improvisations on theatre. The use ofand requires specialized resources. These tests measure contextual scenarios help to understand how they affect usersparameters such as heart rate, electrocardiogram or maximal motivations, how users prioritize the tasks and accordinglyoxygen consumption (VO2max). will help them to answer the following questions: What Following the Ambient Intelligent (AmI) [6] approach, a should this product do? Or, how should this product come upprototype of a medical application that allows users to conduct and behave?the functional assessment of aerobic endurance by themselves Finally, the goals are the driving forces behind thehas been designed and validated, the Functional Assessment behaviour obtained from the user; therefore, a person withoutSystem (FAS). This application had two main objectives: the goals is not an effective communication tool. The charactersfirst one was to propose an alternative solution to the current allow the designer to understand and differentiate task goals.stress tests avoiding the devices invasiveness, while the A goal is a final condition, while a task is an intermediatesecond was to promote personal self-care in people who often stage, a way to achieve a goal. The goals are directed by thepractice sports. motivation and are perennial, while the tasks are almost II. MATERIALS AND METHODS completely determined by existing technology at that time. Consequently, Cooper argues that we must focus primarily on The application design has involved the use of two main goals, and later design tasks (the functionality of the product)methods: the Goal-Oriented Design methodology [1] and the that allow the user to achieve their goals [2].usability questionnaire called AttrakDiff ([9], [10]).
  2. 2. The second method used, the AttrakDiff questionnaire ([9], deviation indicates that the data points are spread out over a[10]) is based on the Hassenzahl model [12] and allowed a large range of values.relevant approach to this study, since it takes into account The standard deviation of a statistical population, data set,both the pragmatic aspects (quality and usability) and hedonic or probability distribution is the square root of its variance. Itaspects, namely, those aspects relating to what the application is algebraically simpler though practically less robust than theprovides the user (autonomy, security, competition or average absolute deviation [14]. A useful property of standardrelationships with others). deviation is that, unlike variance, it is expressed in the same The AttrakDiff questionnaire consists of 28 lines with two units as the data. The formula applied to calculate the standardopposing concepts in each of them, such as simple / complex. deviation is:Each line is composed by 7 holes and the user must indicateone of them based on their product opinion. When a usermarks the fourth hole, means that the product is balancedbetween the two concepts (in the example above, neither toosimple nor too complex). If the user indicates one of the firstthree holes, it would be because his opinion about the product Where n is the number of elements, xi is the value of eachis closer to the left concept (e.g. simple). Conversely, if the element, is the average and σ is the standard deviation.user marks one of the last three holes, his opinion would beclose to the right concept (e.g. complex). The results are III. RESULTSdivided into 4 dimensions: Based on the Goal-Oriented Design methodology, the  Pragmatic Q gives us insight into what users have felt Functional Assessment System (FAS) has been designed and in relation to learning and using the application, if they developed, an AmI system developed for performing a have found it easy or difficult. functional assessment of aerobic endurance. It is composed by  Identity refers to the users feeling with the application. three main elements (Fig. 1): an intelligent textile that records It analyses if having it makes them feel special or just the user‟s data (GOW Running [7] from Weartech Company the opposite. [8]), and an electronic module that collects data and send them,  Stimulation is related to the motivation that brings the via Bluetooth®, to the Android application, that displays data user to use the application frequently or the opposite. to the user.  Attractivity refers to the visual appeal of the application. The AttrakDiff dimensions are measured from 0 to 6,where 0 is the lowest score and 6 is the highest. To provide quantitative results of the AttrakDiffquestionnaire, two statistical parameters have been studied:the average and standard deviation. An average is a singlevalue that is meant to typify a list of values. In mathematics,an average of a data set is a measure of the middle value of thedata set. If all the numbers in the list are the same, then thisnumber should be used. If the numbers are not the same, theaverage is calculated by combining the values from the set ina specific way and computing a single number as being the Fig. 1 Functional Assessment System (FAS)average of the set [13]. The formula applied to calculate the average is: A. User description The FAS has been designed to be used by men and women of any age who practice sports regularly. These users are normally interested in knowing how their hearts react to the demands of exercise and if they are training correctly or not. Where n is the number of elements, xi is the value of each In general, users are familiar with technologies such as mobileelement and is the average. phones, smartphones, personal computers and PDAs. The other statistical measure is the standard deviation and To learn how to use the FAS, users do not need a very highis a widely used measure of variability or diversity used in number of attempts for handle it almost perfectly since thestatistics and probability theory. It shows how much variation application automatically assume many functions and it isor "dispersion" there is from the average (mean, or expected more comfortable for the user. The terminology used in thevalue). A low standard deviation indicates that the data points application is known by the users, because some are verytend to be very close to the mean, whereas high standard general concepts and others are specific to their sport. Following the Goal Oriented Design methodology, the following description of “Persona” was used to design the
  3. 3. application: “My name is Laura, I am 29 years old and I live same, real-time registering of various parameters such as hearttogether my partner in Barcelona. I work at a day-care during rate, electrocardiogram from a lead, duration and maximum8 hours per day from Monday to Friday. I practice sports oxygen consumption (VO2max). Finally the user would reviewregularly, about 3 times a week. I like running and gym the recorded data. Moreover, the FAS would allow the user toactivities such as spinning, lift-training and body-tonic. For navigate by graphics of the vital signs (ECG and HR) to keepseveral years, I cared for my fitness and my diet, so I would track of sports training.love to have some information about my cardiovascular health Taking into consideration the previous requirements, it wasat any time, in a simple, convenient and reliable way. I don´t decided to the Smartphone to be used for developing thelike to go to a specialized place for a basic control of application would be Android based equipped with Bluetooth,physiological parameters during exercise such as maximum touch screen and high process capacity.heart rate or oxygen consumption, because it is expensive andvery time consuming.” D. Description of the interaction frameworkB. Scenario Description The application running on the Smartphone is divided into Once the user description was completed, the scenarios of 3 functional levels: main screen, secondary screens and pop-use where depicted. Users would use the FAS whenever they ups. The main screen contains the main menu through whichcarried out a stress test or practiced exercise. They would users access all the application tasks. It includes the menuwear a smart shirt with the electronic module complemented buttons: Perform Test, Data Review, User Settings andwith a Smartphone. Connection. To start using the system, user would have to enter their Each of the features of the main screen takes users within adata into the system and then, connect the smart shirt and the workflow though different secondary screens. The lowerSmartphone via Bluetooth. Additionally, before performing a levels in the hierarchy of screens are the pop-up which are atest for functional assessment, they should select the protocol text in the top of the screen with a question, and two buttonsto be implemented depending on the place, the parameters to to answer it at the bottom. Pop-ups can have different formatsevaluate and the duration of the test. such as a progress bar, a list of elements for selection and a Once the test or exercise training begins, the system would confirmation dialog. Fig. 2 shows the previously describedrecord different parameters in real time: heart rate, hierarchy of screens where different levels are represented byelectrocardiogram, duration, and VO2max. Users would be able colours.to visualize real-time instantaneous heart rate and its relationto the theoretical maximum frequency using a colour code. When the test is finished, users could review all data thatwere stored in the memory of the Smartphone. They couldalso transfer the generated file to their computer to bringgreater scrutiny of its exercise. Finally, the Smartphone andthe smart shirt would be disconnected.C. System Requirements The next step in the design of the FAS was the definition ofthe system requirement for the smartphone application, thatwas made based on the previous user and scenariodescriptions. The application should be friendly, consistent and intuitiveto use. The user should have full control of the interaction andthe FAS should provide assistance with unfamiliar terms for Fig. 2 Hierarchy of screensthem. It also needed to meet the eight usability guidelines [12]:easy to learn, flexibility, error recovery, consistency, The first step in the design process of the interactionrobustness, response time, adaptation to work and reduced framework was the creation of a simple mock-up of the user‟scognitive load. interface which served to document the main layout of the The application was designed to help the user perform four different screens. The mock-up consisted of several boxesmain tasks. The first one was to fill in the form of the representing each functional group and/or containers withconfiguration screen. The data to be introduced here are names and descriptions of the relationships between differentrelevant for a correct functional assessment of aerobic areas.endurance, and include age, weight, sex, height, the The bottom of all screens, except those graphs that includedtheoretical maximum heart rate and heart rate at rest. After the vital signs (ECG and Heart Rate), were reserved for thecompleting the form, the user should connect the smartphone interaction buttons between the user and the application (e.g.and the smart shirt via the wireless technology Bluetooth. The save data, connect, disconnect, start stress test, etc.). Fig. 3third task would be to perform the stress test, or what is the shows the main screen mock-up at the left and secondary
  4. 4. screen mock-up, exactly the User Settings mock-up, located tothe right. Fig. 6 Buttons, from top to bottom: normal, selected and pressed Additionally, Fig. 7 shows final design of the tabs after applying styles, colors and icons. Fig. 3 Mock-ups: main screen (left) and secondary screen (right) The next step according to the Goal-Oriented Designmethodology was the description of the key path, that is, theactions to be done in order to complete a task of theapplication in a case where everything goes fine. Fig. 4 showsan example of one defined key path. In this case, itcorresponds to the revision of the recorded data, where thesquares represent the application elements and the circles Fig. 7 Tabs, from top to bottom: normal and selectedrepresent the buttons that perform a specific action. The dialog windows are used in various formats. They all include a title at the top of the window whose content can be a question or a simple title. Fig. 8 shows the final design of the dialog window after applying styles, colors and icons. At top, from left to right, the elements are: a progress bar and a list of selectable elements. A questionnaire is place at the bottom. Fig. 4 Key Path: Data ReviewTo properly follow the Goal-Oriented Design methodology,the next step was the „Refinement‟ of the graphical elements.Here the Main Menu was designed with four buttons, whoseappearance would change depending on the state (normal,selected, or pressed). Fig. 6 shows the final design of themenu buttons after applying styles, colors and icons. It shows Fig. 8 Pop-upthe change in color of the fonts, the icon and background. The final step was the definition of the look & feel, which shows the final aesthetics of all screens. A background composed of three different grades of blue color was chosen. To provide a good contrast, white was the color selected for the fonts. In all screens, except the main one, there is a gray
  5. 5. bar at the top with a text informing the user the screen where During the evaluation, the users were working with thethey are. FAS to complete all possible tasks. First, they received a little In Fig. 9 two screenshots are presented, each of them explanation about how to use the application along with thebelongs to a hierarchical level. These images represent the smart shirt and the electronic module. Then they were allowedfinal outcome of the application after the whole design to utilize it freely and, afterwards, they completed theprocess applying the GOD methodology. The main screen is AttrakDiff questionnaire.located at left and a secondary screen is located at right. The results obtained with this questionnaire indicate theAdditionally, Fig. 10 shows two different screenshots with the user sensations when using the application. This is thefinal appearance, on the left is the Connection Screen and the difference between the feeling while users were testing theright is the Data Review Screen. FAS and their expectations before using it. As the dimensions of the AttrakDiff questionnaire are valued from 0 to 6, the average value is 3. Table 1 shows the results of each parameter with its average and standard deviation. All dimensions average are upper than 3, so users have valued very well the application. TABLE I ATTRAKDIFF RESULTS Pragmatic Identity Stimulation Attractivity Q Average 4.3 4.3 4.9 5.3 Standard 0.4 0.5 0.4 0.4 Deviation Fig. 11 shows that all results are above average and therefore they can be considered acceptable. Given that the scores of 98% of users surveyed are between the mean value and double of the standard deviation is even above average, we can conclude that most users have good scored the Fig. 9 Look & Feel: Main Menu and User Settings application evaluation, mainly in Stimulation and Attractivity parameters. Fig. 11 AttrakDiff results V. CONCLUSIONS Fig. 10 Look & Feel: Connection and Data Review ECG This paper has presented the application of the Goal- Oriented Design methodology to the design and development IV. VALIDATION of a medical application that allows people who practice In an application development is crucial know how users sports frequently.to perform a functional assessment of theirreact when use it. The AttrakDiff ([9], [10]) questionnaire was aerobic endurance. The final system has been validated usingemployed to evaluate the user experience when using the the AttrakDiff questionnaire and the outcomes have been verydeveloped system FAS. positive. Attractivity and Stimulation parameters have been the The study was carried out with 11 users, 8 men and 3 best evaluated.women, whose ages were between 24 and 33 years. Based on the outcome of the performed test it can be derived that the FAS system has resulted to be quite
  6. 6. innovative, very intuitive and highly usable. It is expected that [6] W. Weber; J.M. Rabaey and E. Aarts. Ambient Intelligence. New York: Springer, 2005.this application will benefit people who practice sports [7] (2011) The GOW Trainer website. [Online]. Available:regularly by providing them with a tool for completing http://www.gowtrainer.com/ (Last access: 18th November 2011).functional assessments of the aerobic endurance by their own [8] (2011) The WearTech website. [Online]. Available:and, therefore, helping them to exercise in a healthier and http://www.weartech.es/ (Last access: 18th November 2011). [9] M. Hassenzahl; M. Burmester and F. Koller. AttrakDiff: Einsafer way Fragebogen zur Messung wahrgenommener hedonischer und pragmatischer Qualität. In Mensch & Computer 2003: Stuttgart, REFERENCES Leipzig, 187-196.[1] A. Cooper (Ed.). The Inmates Are Running the Asylum: Why High Tech [10] M. Hassenzahl; A. Platz; M. Burmester and K. Lehner. Hedonic and Products Drive Us Crazy and How to Restore the Sanity. Indiana, Ergonomic Quality Aspects Determine a Software’s Appeal. Proc. of EEUU: Sam-Pearson Education, 1999. the CHI 2000 Conference on Human Factors in Computing. ACM,[2] A. Cooper; R. Reimann and D. Cronin. About Face 3: The Essentials New York, NY, 201-208. of Interaction Design. Indiana, EEUU: Wiley Publishing, 2007. [11] M. Hassenzahl (Ed.). The thing and I: Understanding the relationship[3] Peinado I, Arredondo MT, Villalba E, Salvi D, Ottaviano M., "Patient between users and product. In Funology, 2003: From usability to interaction in homecare systems to treat cardiovascular diseases in the enjoyment, M.A. Blythe, K. Overbeeke, A.F. Monk, P.C. Wright, Eds. long term", Conf Proc IEEE Eng Med Biol Soc. 2009;2009:308-11. Kluwer, The Netherlands, 31-42.[4] G. Cea. “Desarrollo de una aplicación médica para la valoración [12] J. Nielsen and R.L. Mack. (Eds.) Usability Inspection Methods. New funcional durante pruebas de esfuerzo en cardiología y medicina York: John Wiley & Sons, 1994. deportiva” M. Eng. thesis, Universidad Politécnica de Madrid, Spain, [13] J. Bibby. Axiomatisations of the average and a further generalization Jul. 2011. of monotonic sequences, Glasgow Mathematical Journal, 1974. vol. 15.[5] G. Rodas Font; E. Garrido Marín. Valoración funcional y cardiología [14] S. Ghahramani. Fundamentals of Probability. New Jersey: Prentice previa al entrenamiento físico, ser. Prescripción del ejercicio físico Hall, 2000. para la salud. Barcelona, Spain: Centro de Estudios de Alto Rendimiento Deportivo (CEARE), 2005. Cap. 2.