The Tools at Hand
“Our most important tools are the ones that involve and empower players to make their own decisions.”
(Church, 2006, p. 379)
This research is grounded in a constructivist epistemology. This chapter discusses the postmodern theoretical perspective of the researchers and elaborates on the user centered design methodology that framed the research process. It covers each chosen method and why they were selected.
2.1 A Theoretical Perspective
Games deal in providing fun experiences, but what is defined as “fun” differs greatly from person to person. Fun has not been operationalised (Baranowski, 2014) and arguably cannot be. Too much is dependent on what the audience wants or expects from a game. This research was conducted under a constructivist epistemology, where meaning was drawn from the views of all parties (Crotty, 1998, p. 58); the clinicians, the participants and the researchers’ own interpretation of acquired knowledge. A postmodern theoretical perspective was adopted in acknowledgment of the multiplicity, ambiguity and ambivalence of the subject matter (Gray, 2014, p. 28). Many researchers of digital games have acknowledged the subjectivity of their findings (Alankus et al., 2010, p. 2121, Chen et al., 2014, p. 8, Flores et al., 2008, p. 383, Holden & Dyar, 2002, p. 70, Nap, Kort, & Ijsselsteijn, 2009, p. 260). Therefore, we consolidated the contributing perspectives to minimize personal bias and improve the integrity of the research.
This research was informed by a user centred design (UCD) methodology (Gregor, Newell, & Zajicek, 2002, Shirzad et al., 2015). Several studies recommend such an approach as the involvement of users and stakeholders in the design process allows for usability issues to be uncovered and resolved (Gerling et al., 2011). A user centred designer is constantly aware of the contextual needs of their users and the functional needs of the product (Shirzad et al., 2015, p. 363), combining these elements to improve the final design’s accessibility. Gregor et al.’s “user sensitive inclusive design (USID)” methodology (2002, p. 152) was found to be particularly relevant for a target audience of older adults and influenced the design process as a result.
To develop a digital game system for rehabilitation, the involvement of professional clinicians and medical researchers became integral. We consulted regularly with specialists from the Auckland University of Technology (AUT). The research was developed in collaboration with fellow Master’s researcher William Duncan. The goal of our collaboration was to generate a game system that enabled play through physiotherapy exercises. This system required a custom physical controller, designed by Duncan, that interfaced the user’s exercises with digital media. The process was conducted under approval from the Health and Disabilities Ethics Committee (for the letter of approval, see Appendix document A, p. 174).
The methods chosen were literature reviews, case study reviews, design precedent analysis, questionnaires, interviews, matrix evaluation, observation and iterative design.
2.3 Methods by Aim
2.3.1 First Aim
The first research aim, to define criteria for designing a digital lower-limb rehabilitation game that engages older adults recovering from stroke, allowed for the findings of our background research to be embedded in the design process.
The first objective of this aim prompted research into literature to gain a scope of existing knowledge. This was followed by the revision of case-studies and design precedents to systematically analyse how digital games have been applied in a physiotherapeutic context, where these projects had failed and what elements fellow researchers deemed worthy of further investigation.
The second objective targeted the acquisition of a professional perspective. Interviews with specialists from AUT (Dr. Nada Signal, Dr. Nicola Kayes and Prof. Denise Taylor) were conducted in the form of research workshops. These workshops provided specialised information (Kuniavsky, Goodman, & Moed, 2012, p. 135) on the treatment of stroke, lower-limb rehabilitation and the expected difficulties patients might experience. These contacts also allowed us to conduct a clinical observation session, which granted insight into the nuances typical for patients undertaking rehabilitation (Zeisel, 1993, p. 113).
The knowledge obtained from this background research was embedded into criteria that became the driving force behind design decisions. These criteria functioned as context-specific heuristics (Nielsen & Molich, 1990) and were used to assess the effectiveness of the game for supporting engagement. Several studies have effectively used criteria-based development processes as a means of prioritising important factors when the context of the project is complex and time is limited (Flores et al., 2008, Martin, Götz, Müller, & Bauer, 2014, Moreira et al., 2010). The criteria were refined as developments were made to ensure the most authentic and applicable version possible (Hartson, Andre, & Williges, 2001, p. 153).
2.3.2 Second Aim
The second research aim, to create and test a prototype game that facilitates lower-limb recovery from stroke, focused on the application of our learnings within the context of a game for stroke rehabilitation.
The first objective of this aim was to explore game mechanics and genres that incorporated rehabilitation into gameplay. Clinicians were key stakeholders, who evaluated the concepts produced in this phase. We employed matrix evaluation (also known as Pugh’s Method) to rapidly process several concepts simultaneously. The variant we acquired from Milton & Rodgers (2013, p. 146) suggests using the strongest concept as a “datum” and comparing other concepts in a matrix using plus (+), minus (-) or equals (=). This process generates a value for each concept that ranks its effectiveness at fulfilling the needs of the stakeholders (Ulrich & Eppinger, 2004, p. 131).
The second objective addresses the iteration of the final output based on the prior evaluation. The rapid turnaround for matrix evaluation allows it to be integrated seamlessly into an iterative design process. Such an approach has been seen in several similar papers (Alankus et al., 2010, Gerling, Schulte, & Masuch, 2011, Shirzad et al., 2015, Trefry, 2010) and has been claimed by Vijay & William as particularly effective when the theory surrounding the designed system is minimal (2015, p. 224). Game design research is too new to have any definitive best practices, meaning such an experimental approach seemed appropriate.
Each iteration was subject to formative evaluation from multiple sources (Barnum, 2011, p. 12, Hartson et al., 2001, p. 149). Designs were critically assessed by the established criteria, as well as by the specialists from AUT. Critique and advice from these doctors supported the integrity of the system as a physiotherapeutic aid. Prior to formal user testing, preliminary user tests were conducted on an audience unaffected by stroke to gain an alternative view of the game’s playability and aesthetics. These tests involved a brief playthrough and questionnaire (for more information on the questions, see Appendix document C1, p. 208) and helped establish whether or not the game was usable by such an audience i.e. caregivers, friends and family. The physiotherapeutic functionality and the generic usability of the system needed to be balanced because without usability the physiotherapeutic aspects would be inaccessible.
The final objective involved formal user tests with the complete system, i.e. the digital game plus custom game controller (smart shoe) designed by William Duncan. Recruiting for user testing began with presenting our research project, as a combined system of hardware and software, to local stroke clubs. Those who expressed interest were contacted over the phone for an interview to test whether or not they met our inclusion criteria (for more information on the inclusion criteria, see Appendix document C2, p. 210). Three of those who met the criteria were asked to participate in user test sessions. For the comfort of our users, the option to conduct the test in their home environment was provided (and taken by all three). Tests involved users putting on Duncan’s device, completing the tutorial phase of the application and playing a standard version of the game. During this process users were encouraged to use “think aloud protocol” (Martin & Hanington, 2012, p. 180). The test sessions concluded with an interview to discuss any points the participants missed with think aloud protocol and to explore their thoughts on the system and how it might be changed (for more information on the interview questions, see Appendix document C3, p. 211). The results of these tests informed subsequent iterations for both aspects of the system. The iterations were tested again with the same users, with a shortened interview afterwards. The findings from these sessions contributed to the final evaluation of the game system, helping articulate strengths and limitations of its design and inform future investigations.
Information obtained from these tests provided a small qualitative sample and cannot be expected to represent a holistic view on the opinions of older adults. It does, however, provide usability engineering measures that may benefit future research (Hartson et al., 2001, p. 150). Given adequate time and resources, it may be possible to measure if engagement with the game system promotes greater adherence to physiotherapy, therefore better recovery. Measurement systems such as the Berg Balance Scale (Chen et al., 2014, p. 3), Fugl-Meyer Test and Wolf Motor Test (Holden & Dyar, 2002, p. 68) could be utilized to generate quantitative data.