Bringing Stories Back to the Center
of Narrative-Centered Learning Environments
The word narrative alone is complicated, let alone the term narrative-centered learning environment (NLE). Narrative in the simple sense of story-telling has been at the center of learning environments from antiquity onward, but the term NLE in the context of this paper is broadly defined as digital, story-based experiences aimed at promoting thoughtful activity in its users. At issue here is the question of the relationship and possible tension between the pedagogical and narrative objectives of NLEs, and related to this is the question of how to define “off-task behavior” within them.
Part One: Crystal Island, Stories, and Learning
This review discusses the evolving position of the researchers and designers behind the Crystal Island NLE, a science mystery with impressive 3D graphics set on a virtual, volcanic island. I argue that the narrative aspects of NLEs like Crystal Island are being pushed to the margin, despite lip service given to their centrality. While it is valuable in a certain context for the Crystal Island team to work towards the most efficient way to raise short-term, standards-based test scores, other versions of Crystal Island and NLEs should aim towards broader, more interdisciplinary goals and unabashedly embrace their narrative centers. My hypothesis is that an NLE can be radically narrative-centered with clear gains in a variety of learning outcomes, including those presently being assessed by the Crystal Island team, so long as the user’s experience is not timed. When under time constraints, engaging the narrative might come at the expense of engaging with the disciplinary content, lowering performance on post-tests. After reviewing some of the research on Crystal Island regarding the role of narrative, I will suggest the design for a separate Crystal Island iteration and an accompanying set of research questions.
What’s The Task?
In 2009 the Crystal Island team published an article “Off-Task Behavior in Narrative-Centered Learning Environments”. The authors write that “behaviors that are clearly unrelated to the narrative and curriculum are denoted as off-task” (3). Note that this includes behaviors related to the narrative but not to the curriculum. For example, one example cited of common off-task behavior in Crystal Island is exploring the Beach area of the story-world because
…it was found to have a negative correlation with students’ test scores, and it contains only peripheral information relevant to the story of CRYSTAL ISLAND—no curricular goals can be achieved there.
It is what they elsewhere describe as a seductive detail that can distract or divert attention from pedagogical objectives, reducing time-on-task. Seductive detail is defined as “highly interesting and entertaining information that is only tangentially related to the topic”, and the topic for the researchers here is the microbiology curricula within Crystal Island. Another humorous example of “off-topic behavior” observed in Crystal Island is users picking up bananas and place them in the toilets of the virtual world. The conclusion of the study finds a negative correlation between off-topic behavior and post-test scores, leading the authors to question the role of seductive gameplay and “purely” narrative features of NLEs. One problem with the study which the authors acknowledge is that some users may have been exploring the setting in search of relevant information, not knowing that there wouldn’t be any there (6). Relevant information here is narrowly defined as information required for the solving of the microbiology learning task. So why include an interesting, explorable setting, or put thought into the narrative as a narrative and not just as a vehicle for curricular delivery, if user engagement in them reduces learning outcomes?
Shouldn’t NLEs Be Seductive?
The Crystal Island team is keenly aware that these seductive, narrative elements promote motivation and engagement. An article published in 2013 focuses specifically on the motivational aspects of NLEs, among which are narrative elements like dramatic plots, believable characters, and other requirements for high-quality stories (223). Narrative environments receive praise for providing meaningful contexts for learning and problem solving. The guiding principle of the article is that, “explicit consideration of motivational factors can guide analyses of the extent and nature of these elements in narrative-centered learning environments” (224). Despite this praise, again the authors excise narrative elements of the experience that are not explicitly linked to microbiology.
First, the authors distinguish between intrinsic and extrinsic motivation. Simply put, something that is intrinsically motivating is worth doing for its own sake; it is “inherently interesting”. On the other hand, extrinsically motivated behavior is done for the “reward of pleasure or security manifested by something other than the task itself”, (227) such as test scores or money. As a framework for discussion, the authors use Malone and Lepper’s identification of four individual intrinsic motivators: challenge, curiosity, control, and fantasy. It’s in the section about curiosity that the authors touch again on the issue of those seductive, narrative elements. The authors define curiosity as an individual’s desire to explore an unknown object, and they observe that narrative evokes curiosity beyond the core subject matter (230). They offer suggestions lest “story-centric curiosity…detract from learning objectives” (231). For example, they removed a poisoning scenario within the Crystal Island narrative that was intended to enrich character personality and heighten the drama, but which was tangential to the primary microbiology learning goals. In support of such modifications is their observation that they got the best results with a streamlined narrative that had just enough elements to be believable, but with no plot twists or characters that were irrelevant to the microbiology task at hand. In this context of these studies, best results are those that demonstrate the highest learning gains on standards-based post-tests, and the priority subject is microbiology. Importantly, the authors draw attention to preliminary findings suggesting that additions like the poisoning scenario did contribute to engagement in the narrative, but not engagement in the game’s educational objectives in the realm of microbiology. But wait: if it’s the narrative that provides the broader context for the microbiology learning, isn’t any narrative in Crystal Island, to the extent that it creates a more compelling and provocative experience, supportive? Previously in the article, believable characters and dramatic plots were rightfully cited as requisite for the high-quality stories that provoke engagement. Do believable characters only talk about microbiology, and don’t poisoning scenarios make for some good drama?
Bring Back the Poison, Please
This issue stands out to me because when I think of why something like Crystal Island appeals to me and many others, it’s because of the detailed world to explore and the story of which I am the protagonist, both of which could help raise my interest in something that I previously hadn’t found that compelling, microbiology. It might show me aspects of microbiology that are intrinsically engaging and playful, and provide contexts that demonstrate the real-world applicability of microbiology knowledge. The narrative
might be worth engaging with for its own sake, as good books are. These things may not translate to concrete, short-term demonstrations of microbiology knowledge in tests conducted immediately after the experience is over, but they may be formative in my development as a motivated, life-long learner across different disciplines.
I agree that the beach exploring behavior of the first study and the poisoning scenario of the second study sound off-task with respect to the specific microbiology content of the Crystal Island NLE, but it is undoubtedly on-task with respect to the potential provocations of the narrative and setting, which qua narrative affords ample opportunity for significant kinds of learning and activity. In terms of the microbiology content, I suspect that when users are more engaged in the narrative they will end up more engaged with the microbiology content, since it’s the narrative that provides the meaningful and engaging context for microbiology. This hinges on how authentically and creatively the curricular content is woven into the narrative frame of the NLE, such that engagement with the narrative comes to include and depends on engagement with microbiology.
My proposal is that while research continues on the iterations of Crystal Island that prioritize microbiology post-test results, a separate iteration be developed and studied that has broader goals and which takes more seriously the challenge of putting the narrative at the center of the experience. This seductive Crystal Island will include more narrative details, especially ones that don’t involve the user explicitly dealing with microbiology learning. In one study on a computer-mediated math tutoring environment, the presence of text not directly related to the problem solving tasks lessened the likelihood of users trying to simply click their way through the problems without learning anything, and whether or not narrative unrelated to microbiology on Crystal Island causes a similar phenomenon should be investigated (Baker 2009). I will also remove the timed dimension of previous studies, so that exploration of narrative and setting doesn’t come at the expense of engaging with the content that more concretely deals with microbiology. I will have one control group use the seductive Crystal Island and another use a previous iteration, and both groups will be given a post-test whenever the story within the NLE is over. I am also interested in the potential difference between subjects who know that there will be a microbiology post-test after Crystal Island and those that don’t, because I suspect that those who know will be less likely to engage with the narrative, instead focusing too much on the microbiology content. Those who know about the test might achieve higher microbiology post-test scores, but I predict they will be less engaged and not get as much enjoyment out of the experience. I predict they will show less motivation to continue learning microbiology than the group that didn’t see Crystal Island as test-prep before beginning it.
Off-topic behavior in this iteration will be more narrowly defined as deliberate avoidance of exploring any content of the NLE, for example if a user is only walking around parts of the map that have already been thoroughly explored, or deliberately avoiding any interactions with all non-player characters. There will be algorithms that detect this behavior and trigger interactions that force the user, however reluctant, to eventually re-engage with the story. One example of this is making a non-player characters approach the user, rather than requiring the user to approach them. I predict, though, that off topic behavior will be less common in the seductive Crystal Island than in other iterations.
Do We Have Any NLEs?
A challenge for game authors is creating environments so rich that very few activities within them could be seen as unequivocally off-topic. This is made easier when NLEs adopt a more expansive set of learning goals, beyond testable knowledge in a single subject. If the main priority of the game designers is improved learning outcomes as defined by standards-based test scores in as short amount of time as possible in one discipline, then indeed there is little motivation for developers and educators to create nuanced, provocative narrative environments. If the narrative of the story is seen as holding little intrinsic educational value on its own, then narrative elements will be on the chopping block. Instead, we should encourage the development of NLEs that have a different set of goals than might include but are not limited to results on conventional or standards-based assessments, and which don’t see a zero-sum game being played by narrative and pedagogy. Then we could feel justified in our use of the term narrative-centered learning environment. We shouldn’t have to put at risk one of the motivational and pedagogical factors of a compelling NLE: the story.