Preference for increased uncertainty in a computer-based task: testing an educational hypothesis derived from neuroscience.
Skevi Demetriou
(University of Bristol)
Paper presented at the British Educational Research Association Annual Conference, Institute of Education, University of London, 5-8 September 2007
Abstract
This research aimed at exploring the role of uncertainty in learning as a factor of motivation and engagement. In particular, the investigation of the motivational impact of uncertainty was framed in a specific educational computer game designed for the purposes of this study that allowed a thorough examination of uncertainty. Specifically, the investigation focused on the identification of motivation arising by uncertainty juxtaposing students’ preference together with their experiential assertions and their expressional reactions during their interaction with the computer-based task. The research was conducted in school settings, and the student participants were asked to perform a computer-based gaming task. It should be highlighted that the researcher team was observing students interacting with the game and recording their reactions. In addition, a proportion of the sample was interviewed right after the completion of the task. The teacher’s point of view was also recorded in the research procedure, through an interview, expressing his opinion, from his perspective, on what was going on (video analysis). The statistical analysis of the collected quantitative data, derived from students’ interaction with the task, demonstrated a statistically significant difference between students’ preference in the certain and the uncertain choice, favouring uncertainty, indicating motivation emerging from it. Additionally, the qualitative part of the data linked motivation for engagement deriving from reward uncertainty with affectional correlates of reward relevant to pleasure and enjoyment. However, emergent themes were also taken into consideration throughout the analysis of the data illuminating the research question, with individual differences, gender and contextual issues being some of the representative ones.
Introduction
Computer games are a rapidly growing part of our culture (Kirriemuir & McFarlane, 2004). The impact of games on people of all ages and especially on youth has lead to the development of a vast global community of users, affinity groups and entrepreneurs investing millions of dollars on game related projects (ibid). Within this progression, the potentially beneficial contribution of technological means and especially of computer games, on education and learning has been highly emphasised (Gee, 2004).
Even though some scholars are rather sceptical regarding the link between computer games and learning, others don’t hesitate to overtly acknowledge their beneficial influence on children’s education and learning in general, usually grounding it on the argument that fun learning is effective learning (Lepper and Cordova, 1992). As Prensky (2003) asserts, “video games[1] are not the enemy, but the best opportunity we have to engage our kids in real learning”. From a broader view, the game element seems to carry added value in pedagogy, enhancing students’ motivation and making learning derived from such instruction more efficient compared to that achieved in more conventional settings excluding game features (Whitehall and McDonald, 1993; Ricci et al., 1996; Randel et al., 1992). However, research concerning the implementation of computer games in mainstream education is relatively novel, but growing (Kirriemuir & McFarlane, 2004).
The prevalent reason for the interest in this research area is that computer games motivate people and especially children in a way that formal education doesn't (Facer, 2003) and this is also my personal motivation in pursuing this study; A better understanding of why both children and adults are motivated by computer games may help devise more effective learning experiences in and out of formal school settings. Literature points out that satisfaction, and hence motivation, arises within a gaming environment due, partially, to encountering challenges with unknown and uncertain results (BECTA, 2001). Therefore, uncertainty seems to be one of the engaging and motivating elements of computer games.
Educational literature acknowledges a wider beneficial contribution of uncertainty in academic achievement (Sorrentino & Roney, 2000) and in students’ motivation to engage in leaning activities (Zaslavsky, 2005). Also, psychological evidence indicate that we are particularly excited by different levels of uncertainty when seeking rewards, with motivation peaking near 50% probability (Weiner, 1985; Csikszentmihalyi & Csikszentmihalyi, 1988), exactly where moderate risk taking is required. Additionally, studies on the primate brain show that uncertain reward has rewarding and reinforcing properties (Fiorillo et al., 2003) and may increase motivation, thus suggesting a brain-basis for the pleasures of gambling (Shizgal and Arvanitogiannis, 2003). In a complimentary argument, Fischbein (1987) regards the need for certitude that follows our intuitive notions as a strong driving force for learning, suggesting that people attend to uncertainty in order to reduce it. The previously mentioned arguments indicate the diverse ways in which uncertainty has been implicated with motivation and learning and determined my interest in this research field and in launching the specific research.
Methods
The current study is based on providing students with an educational computer game, especially designed for the purposes of this research. In particular, the game allows the user to choose to incorporate elements of uncertainty, and the researcher’s role is to examine whether, when and how students choose to introduce uncertainty in a specific learning context. Particularly, in pursuing the general aim of understanding how uncertainty impacts on students’ motivation for engagement in a computer-mediated, game-based, learning task, this project seeks to answer the following interrelated research question: “How does uncertainty influence 11-12 year old students’ motivation and therefore their appetite to engage in a learning task within a computer-based, game?”
This study will address the research question by examining children’s responses to uncertainty in a computer-game-based learning environment. The research has been designed as an experimental task-based study of chiefly quantitative nature combined with some qualitative aspects. The task is incorporated in a computer-based game (figure 1) designed for the purposes of this research. This experimental task is included in a broader research methodology that also includes the collection of qualitative data regarding the context of the learner and the learning situation. As Johnson and colleagues (1988) mentioned, traditionally, task-centred analysis investigates what people do when they carry out one task and involves collecting multi-perspective information about how and why they perform each segment of the task. This will contribute to the triangulation (Denzin, 1970) of the evidence (mixing of data and methods) which is one of the most pertinent aspects of my methodological approach.
(a) Sample
Fifty 11-12 year old students from one Cypriot primary school, comprising two “Sixth Grade” classes of twenty-five pupils each and one of the two teachers of these classrooms comprised the sample of this study. The participants were drawn from an inner-city primary school of Lemesos[2] (Cyprus) with substantial ICT infrastructure. I’ve decided to do so in my attempt to minimize the intervention of unexpected variables, like students’ lack of ICT skills, affecting the procedure of the research.
Formal permission for access to schools was given to the research team from the Cyprus Ministry of Education and Culture. In addition, a signed declaration from parents was requested from the students wishing to participate in the study so that parents’ consent would be explicitly ensured prior to the launch of the research. Moreover, it was made clear to the participants that they could not take part or withdraw from the research whenever they wished to.
(b) Research settings
The experimental part of the research was carried out by two researchers in a quiet room of the school premises using two laptops. Each researcher was using one of the two laptops that were placed at significant distance. Two students were entering the room at a time and each approached one researcher. After that, every researcher was explaining the task in detail to the participant that approached him/her and was also trying to acclimatise him/her with the situation. Later on, the participants would complete the task while the researchers would observe the procedure. In particular, each researcher was sitting next to the participant taking field-notes of his/her reactions while undertaking the task. It should be highlighted that each group (1 researcher-1 student-1 laptop) was carrying out the research independently with no interference by the other.
Interviews were conducted selectively from a proportion of the student-participants right after the student had completed the task, in a quiet room of the school which’s access was restricted to anyone else apart from the interviewer (researcher) and the interviewee (student). The same settings were used for the teacher’s interview.
(c) Methods of data collection
(i) Game
The methodology is experimental task-based. The task was framed within a computer-based game designed for the purposes of this study (figure 1) that allowed the role of uncertainty to be closely explored. It was designed in order to measure students’ preference between the certain and the uncertain pathway while answering a set of mathematical questions (figure 1). A majority of choices in the uncertain pathway was interpreted as reflecting greater motivation towards uncertainty when engaging in the task. This makes sense, since free choice can be considered as part of an explicit approach to a task, expressing the “wanting” or desire for something and corresponding to the psychological concept of “motivation” (Berridge & Robinson, 2003). Similar interpretations operate in the case where certainty dominates students’ choices, i.e. they are motivated by the certain option.
The game included a set of 30 mathematical questions that the participants were asked to answer. Before each question the students had to choose between the certain and the uncertain pathway appearing on the screen in the form of boxes containing questions. The game explains to the player in advance that the certain choice gives zero points for a wrong answer and one point for a correct answer whereas the uncertain gives again null points for the wrong answers but could give either zero or two points, with a 50%-50% probability, for the right answers, depending on the result of a coin being tossed in such cases. The same procedure was followed for each one of the thirty questions. It should be underlined that it was explained to the participants beforehand that the two pathways provided the same probability to earn points. The overall aim of the game was to collect as many points as possible.
The software was programmed to keep track of what was going on during students’ interaction with it. Additionally, it estimated the overall percentages of choice between the two options (certain and uncertain).
(ii) One-to-one overt partly pre-structured observations
Schiefele (1991) supported that motivation and engagement involve both feeling-related characteristics (enjoyment, excitement, involvement) and value-related ones (attributing significance to the activity). However, the software didn’t afford recording students’ non-verbal reactions and body language during the task (Argyle, 1987), like frustration, excitement, facial expressions, hesitation, speed of response, enthusiasm, boredom, etc that could make implications concerning motivation and its link with affectional correlates of reward. In addition, based on assertions made by Ryan and Deci (2000) motivation concerns “energy, direction, persistence and equifinality”. These are aligned to the assertions made by Berridge & Robinson (2003) according to which motivation is linked to the affectional pleasures of reward. They also supported that all these are aspects of activation and intention. Drawing from these definitions, which seem to prevail in literature, I decided to also include observations in my methodology in order to absorb these kinds of behavioural evidence that could carry added value upon the findings.
Overall, literature, analysed above, seems to link motivation with energy related non-verbal active expressions that resemble responses of excitement. Due to this it was decided to observe and report students’ expressions of excitement while observing them performing the task. As suggested by the previously analysed literature (Schiefele, 1991; Argyle, 1987; Ryan and Deci, 2000; Berridge & Robinson, 2003) these expressions are related with motivation and this relation is attempted in data analysis. Moreover, interpreting the previously mentioned assertions by researchers, it decided to also report students’ expressions indicating boredom as signs of the absence of motivation. In particular, literature seemed to associate motivation with energy related expressions of excitement that are obviously in opposition to expressions of boredom. It should be noted that the predetermined list of issues / category system[3] (figure 2) that was chosen to focus on, also included “frustration”. It was decided to also record students’ expressions of frustration after piloting the software with one child before launching the actual research. The participant seemed to get rather frustrated when choosing the uncertain pathway, answering correct and winning no points. This was a quite interesting emergent theme which was considered important to include in the list of predetermined issues under observation. This was based on the aspiration that frustration could enrich our understanding regarding students’ behavioural responses in the task.
Even though we had a pre-defined list of categories, the observations were not strictly systematic but maintained the element of flexibility. Therefore, the nature of the observations allowed spontaneous re-structuring of the categories and creation of new categories in order to record emergent themes associated to the research topic, but without reaching ethnography.
(iii) Semi-structured interviews
Even though one-to-one observations try to understand real situations, they are still outsider perspectives (Papanastasiou, 1996). In our attempt to triangulate and collect data that incorporated the experiential knowledge of the participants it was decided to conduct semi-structured interviews to a small proportion[4] of the participants (Bogdan & Biklen, 1992) that were recorded using an audio recorder. The aim was to gain insights to students’ justifications of their choices (certain/uncertain) that enhanced the findings and helped improve our understanding concerning uncertainty and motivation. Thus, we were exploring their reasons for their preferences as well as their feelings from the insiders’ experiential perspective.
In addition, we interviewed one of the two teachers of the two classrooms, Mr Yiannis, in an attempt to take his point of view into account as well when interpreting the findings. The teacher was co-observing with the researchers a small proportion of students undertaking the task. Therefore, while interviewing him we were seeking for his interpretations on how students interacted with the game.
We had also video taped a proportion of students performing the task and analysed the videos a posteriori with the teacher. Consequently both interview and video analysis with the teacher were seen as a way for joined analysis of students’ behaviour between me and the teacher. This was extremely helpful in terms of providing an outsider, but not completely outsider, perspective of what was going on.
The fact that the teacher had close relationship with his students, that formed the main body of participants of this study, and that he was in pedagogy interacting with the specific students everyday, grasping their way of thinking, was making his point of view respond to a mixture of outsider-insider perspective. It was like he was bridging the gap between my perspective, which resembled a rather outsider form, together with students’ insider perspective.
The CDs with the interviews were transcribed using a table. Afterwards, the important issues from the data collected form students’ interviews were included in a column incorporated in the transcript of the field notes and thus the analysis of the data was conducted on the final expanded transcript containing data from both observations and interviews. This was an attempt to merge and integrate the data collected from the observations with those recorded from the interviews so that I could analyse them in a more effective way.
(iv) General remarks – Route sheet
All the findings were interpreted qualitatively combining evidence collected from all the perspectives together with pertinent literature. It is important to mention that I haven’t examined correlations between the choices of the participants and any other variables, like their general tendency in risk taking. What I have examined is which one of the two options gathered the majority of choices and why. The most frequent choice was interpreted as a motivational preference for how to engage with the learning task and data from the observations as well as those derived from the interviews helped to illuminate with greater depth how participants were making their decisions with regard to the research question. Particularly, the analysis focused on the following:
- Choosing (quantitative data from students’ scores in the game) was translated in preference, which indicates motivation derived from the specific choice (Berridge & Robinson, 2003). Also verbs like prefer and want, that link to the meaning of “wanting”, as proposed by Berridge & Robinson (2003), found in qualitative data from students’ interviews were also translated into motivation (approaching a task).
- Phrasal expressions (qualitative data from students’ interviews) and bodily expressions[5] indicating students’ excitement (consuming a task), derived from qualitative data from students’ observations while performing the task. I’ve chosen to focus particularly on those expressions that appeared related to hedonic pleasure/liking/affect and motivation/wanting/desire. “Liking” seems to be associated with hedonic pleasure and emotional correlates of reward consumption, possibly linking to motivation (Berridge & Robinson, 2003). Therefore, the link between verbs like “enjoy” and “like” found in students’ interviews, which refer to the consumption of the task, with motivation/wanting/preferring was closely examined in order to identify association or dissociation. This was attempted having in mind Fiorillo’s (2003) findings that seem to link motivation with pleasure in reward. According to the results derived from this study, dopamine (and thus motivation) is peaking at uncertain conditions of 50%-50% probability for reward but when the reward is something the participant finds personal interest and pleasure in (difference in monkeys’ response between the syrup-reward and the visual-stimuli-reward analysed in previous chapters).