1
Brain-training games and metacognitive
development:
The effect of a modelling agent
Matteo Rinaldi 11580461 Master’s Thesis
Graduate School of Communication
Research Master’s Programme Communication Science Supervisor: Dr. Jessica Taylor Piotrowski
2 Abstract
Metacognition is one of the most powerful predictors of learning. Having high levels of metacognition positively impact academic success, as well as logical and emotional skills. Brain-training games have been shown to be an easy, fun, and accessible way to improve metacognitive knowledge and self-regulation of cognition, especially among young people. However, research has also pointed out that a modelling agent, modelling metacognitive behaviour, may be an even stronger way to support metacognition than brain-training games. With the new movement towards live-streaming of game play in which streamers voice their game play decisions, we have the opportunity to assess whether the combination of brain-training games with a live modelling agent (in the form of a Twitch streamer) may be extremely effective at enhancing metacognitive skills. With this in mind, a 5-week
experiment was conducted with 172 Italian elementary and middle school children (Mage =
9.30) to investigate whether brain-training games, augmented by a streamer, may lead to differential effects on metacognition when compared to gameplay without a streamer. Moreover, gender was formally investigated as a potential moderator of effects given a host of literature suggesting differential benefits of gameplay for boys and girls. Results were counter to expectations. Specifically, metacognitive skills were unaffected by study
condition. Reflections as to why this study was unable to replicate previous work, and next steps for the field are presented.
Keywords: brain-training games, digital games, metacognition, modelling, streamer,
3 Introduction
Digital games have been undergoing the (sometimes) hypercritical eye of society, academics, and even politics an innumerable amount of times (Kirsh, Mounts & Olczak, 2006). This huge deal of attention comes from the widely implicit accepted notion that children do learn from media messages (Valkenburg & Piotrowski, 2017), as well as the fact that researchers have repeatedly proven negative effects of media on young people ranging from problems such as eating disorders and body shame (Littleton & Ollendick, 2003) to hostile emotional states and aggression (Anderson & Bushman, 2001).
While it is true that media are a dominant force in children and adolescent lives – it is the leading activity for children and teenagers other than sleeping (Rideout, 2010),
apocalyptical visions of media are far from being the only side of the coin.
In fact, media – and especially digital play - have been repeatedly proven to be a valuable tool to better young people self-regulation and arousal (Goldstein, 1995), and to positively impact social (Ferguson & Garza, 2011), cognitive (Bavelier, Green, Pouget, & Schrater, 2012), and learning skills (Blumberg, 2007; Blumberg, & Fisch, 2013; Zimmerman & Christakis, 2005; Hogan, 2012), as well as their attention and vigilance functions (Trick et al, 2005).
Especially on the educational front, educational games have been defined as the perfect way to merge learning with the classic enjoyment deriving from play (Squire, 2011) and have been praised both by researchers and practitioners (Bavelier, Green, Pouget, & Schrater, 2012; Green & Bevalier, 2003). When learning is placed in a game-based environment, children learn more and retain content better compared to more traditional approaches to teaching (Cordova & Lepper, 1996), since learning occurs more effectively when it is active, experience and problem based, and when it offers immediate feedback – all features that digital games can offer (Valkenburg & Piotrowski, 2017). Digital games can in
4
fact act as “brain-trainers”, when presenting challenges and requiring problem solving skills (like solving an enigma in the game). Far from being “just” playing, videogames have the potential to better students’ skills and improve the way they learn.
Brain-training videogames, through their cognitive demands, can also positively enhance children’ metacognition – more specifically metacognitive knowledge and self-regulation skills (Morris, Croker, Zimmerman, Gill, & Romig, 2013; Van Deventer & White, 2002; Aliya, 2002), helping kids to understand what, when, and why a strategy is best to be implemented, especially in learning contexts (Flavell, 1979).
Yet, with the exception of a few studies (Cristoph, 2006; Morris, Croker, Zimmerman, Gill, & Romig, 2013), digital games and metacognitive development are still a widely
unexplored area of research when it comes to young children, despite calls for more on how metacognitive processes may evolve in computer supported contexts and how other agents can scaffold/model these processes (Salovaara 2005; Azevedo & Hadwin 2005; Arvaja, Salovaara, Hakkinen, & Jarvela, 2007), since it has been proven that seeing how external agents solve problems improves metacognitive skills (Pifarre & Cobos, 2010, Wouters, Paas, & van Merrienboer, 2008).
With this in mind, the aim of this study is to understand the potential influence of brain-training videogames on children’s metacognition – with particular attention to how modelling agents (in the form of videogame streamers) may scaffold potential processes. This thesis also tries to investigate the conditional effect of gender over these processes, given the recognized importance non-media variables have on media effects (Valkenburg and Peter, 2013b, Dresel & Haugwitz, 2006; Valkenburg & Peter, 2013; Blumberg & Sokol, 2010; Mohd Suki, 2013; Kwak, Zinkhan, & Dominick, 2002; Lopez, Corona, Halfond, 2013), and since gender has been proven to be an important factor in how metacognition develops and
5
can be enhanced in children (Richard, 2013; (Kolić-Vehovec, Bajšanski, & Zubković, 2010; Tarhini, Hone, & Liu, 2014). This study seeks thus to answer the following question:
What are the effects of brain-training videogames on children’s metacognition when a
modelling agent is included, and how does gender moderate these effects?
Theoretical Framework
Metacognition and Self-regulation processes
Metacognition is the experience of and ability to reflect on one’s cognitive processes and to control it (Flavell, 1979; Perfect & Schwartz, 2002). More simply, it is the act of “thinking about thinking” (Livingstone, 2003). Even though the concept seems at first glance puzzling, the truth is that people engage in metacognition every day. For example, a student evaluating his own performance is a sign of metacognition. Though metacognition as a concept has been going through a challenging theoretical refining (Veenman, Van Hout-Wolters, & Afflerbach, 2006), it has also been deemed as one of the most important goals for children and adolescents when considering their cognitive development (Kuhn 2000), since it is a powerful predictor for learning (Wang, Haertel, & Walberg, 1990; Zulkiply, 2009). The concept first appearance was in the work of Flavell (1979) who, in later writings, divided metacognition into metacognitive knowledge and metacognitive (self-)regulation (1993).
Metacognitive knowledge (MK) is the awareness acquired by an individual about his/her own cognitive processes, such as knowing “about” things (“I am better at biology than at statistic”), “how”, “when”, and “why” to do things (“I should use a different strategy than highlighting here”).
Self-regulation (SR), by comparison, is the act of exercising control over cognitive processes, such as planning, managing information, monitoring and debugging cognitive
6
processes, and evaluating them (Schraw & Dennison, 1994). Examples can be deciding to read a chapter per day when studying for an exam and monitoring how fast one is learning with this plan.
Research has shown that metacognitive knowledge (MK) starts developing between the ages of 3 and 5 years but continues to do so during the entire life course (Veenman, Van Hout-Wolters, & Afflerbach, 2006). Self-regulation of cognition (SR), on the other hand, starts developing only from 8 to 10 years old (Whitebread et al., 2005). Metacognition, overall, experiences dramatic growth between the ages of 8 and 14 years (Weil et al., 2013) making early adolescence the “the golden age of metacognition”, and as such, a highly relevant period for empirical investigation and the focus of this work.
Videogame Play Meets Metacognition
Research has shown that digital play can be beneficial in many ways for young people, from social (Ferguson & Garza, 2011), to learning and cognitive skills (Bavelier, Green, Pouget, & Schrater, 2012). One of the aspects digital play is able to enhance is metacognition (Morris, Croker, Zimmerman, Gill, & Romig, 2013¸ Van Deventer & White, 2002; Aliya, 2002), particularly as the result of immediate feedback that games provide, a factor known to elicit metacognitive processes (Scardamalia, 1984; Fisher, 1998). Such feedback – also in the form of a modelling agent – is said to activate a process whereby the player is forced to review his own strengths and weaknesses (Sadler, 2010), as it stems from the theory of self-regulated learning by Zimmermann (1989, 2000) and Bandura’s Social Cognitive Theory (1977), that acknowledge both the role of external modelling in enhancing skill acquisition.
When players analyze the games of themselves or others, they engage in
7
recognize their own strengths or flaws from other players’ actions. Research showed that especially strategy games are able to encourage players to engage in metacognitive behaviors (Antonietti & Mellone, 2003; Doolittle, 1995; Henderson, 2005; Hong et al., 2012; Horak, 1990; Ke, 2007; Pillay, Brownlee, & Wilss, 1999; Pillay, 2002) but that even the most simple riddle or puzzle can elicit creative thinking about strategies and evaluation of the efficacy of those strategies (Doolittle, 1995). This, especially if the game supports development of strategies for problem solving (Moncarz, 2012) as where “the player takes on the role of a fantasy character moving through an elaborate world, solving various problems” (Gee, 2003).
From Playing to Streaming
While videogames, particularly the so-called “brain-training games”, seem to have some merit when it comes to metacognition, the field makes the powerful point that the “in-game” modelling of metacognitive skills may be a particularly effective way to support metacognition. Specifically, studies have shown that modelling (e.g., verbalizing internal thoughts) can help individuals (particularly children) become accustomed to metacognition. Modelling agents convey information and feedback to learners in a way that positively shapes the thinking needed to make decision about the task at hand (Corden, 2001) and these
“thinking out loud” practices can actively show how thought processes are made, rather thank just why and which ones (Fisher, 2002). Fisher praises the use of “think aloud” and “meta-teaching” methods as an efficacious way to elicit metacognition in children “by bringing to conscious awareness […] thoughts and feelings” (p.11, 2002).
In recent years, some scholars have questioned whether this modelling could be replicated in gaming. For example, Noor Cristoph (2006) tried to understand the role of a “task model” represented by statements appearing in a game that would encourage
meta-8
cognitive thought1. These models helped the players, but the effect seemed limited, perhaps because of their artificiality.
But what if there was a way to test modelling effects on children’s metacognition using more articulate models, such as a human presence modelling metacognitive behaviours, thus with more human-like characteristics combined with a more natural experience instead of static, digital sentences?
With today’s digital environment, there might be. Specifically, platforms such as Twitch.tv2 allow users to livestream themselves while playing (almost) any kind of games. At the time of this writing, Twitch is the 31st most popular website online (14th in the United States), with more than 9 billion hours watched during 2018 only and more than 1 million concurrent everyday viewers (Twitch Statistics and Charts, 2019).
Videogame streaming presents itself as a natural setting already well established in young people lives. What is most interesting, however, is the fact that it merges digital games and metacognitive strategies together. The streamer (i.e., the player who is recording his/her gameplay) acts as a narrator, verbalizing his/her thought processes during the gameplay, analyzing decisions and strategies, evaluating results, backtracking, and trying new solutions. These behaviors are, indeed, pure verbalization of metacognitive strategies as expressed by Schraw and Dennison (1994), but they happen in a more natural setting and are much richer than the one used by Cristoph (2006).
Inasmuch, it seems reasonable that streamers may be good agents for modelling metacognition for several reasons. First, because of how a streaming session is composed, a streamer playing a digital game usually engages in verbalization of many (if not all)
component of self-regulation of cognition. Second, streamers usually display behaviors of
1 Examples of these models are “But what do we have to do now?” (comprehension monitoring) or “That is strange, it does not influence anything and still you want to focus on it” (evaluating).
2
9
explicit declarative, procedural, and conditional metacognitive knowledge, and third, streamers might elicit metacognitive experiences in the audience whenever a viewer recognizes a limit in his/her own problem-solving capacities (Pifare & Cobos, 2010).
Taken together, streamers modelling metacognition might enhance metacognitive skills (operationalized as both MK and SR) via “involuntary scaffolding” even more than directly engaging with the game. As the study from Pifare and Cobos (2010) found, students extremely benefit from knowing about how others (like a streamer) solve problems, which lead to an improvement of their metacognitive skills overtime. At least one study, moreover, considered analyzing the gameplays of others to activate metacognitive processes (Foster, Esper, and Griswold, 2013). Noticing decisions that the other player makes serves the
purpose of learning new strategies and enhance control over the game (Hamilton et al., 2014). It is then logical to consider that the verbalization of the streamer’s thought process may be able to recreate specific moments of self-reflection in the viewers through the continuous natural conversation typical of the streaming practice.
In his Social Learning Theory (SLT), Bandura theorized that most human behavior is learned through observation and modelling (1977). SLT suggests that individuals pay
attention to the behaviors of other people around them – including media – especially when they need to solve new or difficult tasks reproducing those modeled behaviors (a motivation that could stem from a goal to achieve, like advancing in a videogame). In this optic,
individuals trying to learn how to solve a problem by watching those who engage in a modelling behavior might have a “glimpse” of other people’s cognitive and behavioral patterns who show them “how things are done”.
With this lens in mind, it is possible to see streamers as modelling agents who
promote guided instructions which ultimately provide users conceptual tools needed to learn metacognitive strategies (Bandura, 1989). Inasmuch, reasonable hypotheses are:
10
HP1a: Users who watch a streamer modelling metacognitive behavior while playing a
brain-training game will have higher MK than those users playing the game.
HP1b: Users who watch a streamer modelling metacognitive behavior while playing a
brain-training gam will have higher SR than those users playing the game.
Conditional Effects
As mentioned before, levels of MK and SR differ among different people, depending on a variety of factors – most notably gender. Contemporary media-effects theories such as the Differential Susceptibility to Media Effects Model (DSMM) argue that media effects are conditional – that is, they are contingent on many different non-media variables (Valkenburg and Peter, 2013b). In recent years, the importance of recognizing these conditional effects has been stressed by a rising number of communication science scholars, who acknowledge that “ignoring conditional media effects may easily lead to invalid conclusions about the
magnitude of media effects on certain subgroups of individuals” (p. 203, Valkenburg & Peter, 2013a). In particular, existing literature would suggest that gender may have a meaningful influence on the relationship between brain-training games, modelling, and metacognition.
Gender and Metacognition. Gender has been widely recognized to have a
moderating role on media effects (Dresel & Haugwitz, 2006; Valkenburg & Peter, 2013a; Blumberg & Sokol, 2010; Bushman & Green, 1990; Mohd Suki, 2013; Kwak, Zinkhan, & Dominick, 2002; Lopez, Corona, Halfond, 2013). In this paper, gender effects are considered in two ways. First, social stereotypes (such as the fact that videogames are often considered to be a “a boy-thing”) are one reason why gender could moderate the relationship between
11
play and metacognitive outcomes (Richard, 2013). Stereotypes may lead to different level of motivation for use, moderating responses to the content and, in turn, influence associated outcomes (Valkenburg and Peter, 2013b).3
Second, as far as MK and SR are concerned, girls outperform boys at both during early adolescence (Kolić-Vehovec, Bajšanski, & Zubković, 2010; Kolić-Vehovec and Bajšanski, 2006a; Kolić-Vehovec and Bajšanski, 2006b). Moreover, girls seem to be more likely to take into consideration the advice of experts (such as a modelling figure) in the learning process (Wang et al. 2003; Tarhini, Hone, & Liu, 2014).
At the same time, digital play motivate boys more often than girls (Tarhini, Hone, & Liu, 2014), especially in informal settings (Blumberg & Sokol, 2004), and their better general expertise, enjoyment of competition, and frequency of use (Terlecki, Brown, Terlecki,
Brown, Harner-Steciw, Irvin-Hannum, Marchetto-Ryan, Ruhl, & Wiggins 2011) could balance their lower levels of metacognition.
Inasmuch, one can expect that gender will play a moderating role in two ways: first, we might expect boys to acquire more metacognitive skills through a playing condition that would allow them to satisfy their desire for mastery and competition (Kafai, 1996; i.e., a brain-training game). Second, we might expect that girls would be better able to maximise their metacognition from a streaming condition where instructions and verbalizations of metacognitive strategies are prominent (i.e. a condition where both the streamer and the brain-training game are present).
H2: Gender will moderate the relationship between condition and metacognition (MK and
SR) such that boys in the playing condition will have higher MK and SR skills than girls
3 For this reason, video game use and game enjoyment will be considered as control variables, given the importance they have towards MK and SR. See Method section for a more detailed description of all control variables included in this study
12
(H2a) while girls in the modelling condition will have higher MK and SR skills than boys
(H2b).
Method Design and Participants
A 2 (condition: playing, streaming) x 2 (gender: male, female) between-subjects experiment was conducted, in schools, to assess study hypotheses. A review of the literature on teaching metacognitive strategies to students revealed no clear pattern as to the minimal length of time for game exposure. Some studies suggested a minimum of 3 class sessions (2 hours) (Kistner et al., 2010) while others suggested nearly 5-times that amount (i.e., 14 class sections; Zimmerman, 1998). With such broad guidelines in hand, an effort was made to find a middle point between these extremes while at the same time being sensitive to the external confines of participating schools (e.g., student time away from classroom; available
infrastructure for conducting the study at the schools). In practice, this meant that the entire study was conducted in five weeks whereby each week participating students played their assigned game or watched the assigned video for a period of 15 minutes. During the first and final session, pre- and post-test assessments were also completed.
Participating children were randomly assigned to one of two conditions: playing a brain-training videogame (playing condition) or watching a live-streaming (streaming condition) where a professional streamer was playing the same game4. The participants themselves were tweens (N = 172, 69 girls; MAge = 9.27 years, SD = 1.24; age range: 8 – 12
4 A control condition was deemed not necessary for the following reasons. First, the playing condition already acts as a sort of control since playing a videogame can be considered as the most common media use and thus the baseline when considering digital games. Second, the presence of a modelling agent in the streaming condition would be held constant in the playing condition. Third, the aim is to check whether children in the streaming condition will have higher MK and SR skills than those in the playing condition at the end of the study. While a certain level of metacognitive learning would be considered in the playing condition, we are interested in any significative increase between these two conditions - thus making a control condition not necessary once again. Finally, due to time and resource limitations, adding a control condition was judged not a feasible choice for this research.
13
years). After receiving ethical approval from the University of Amsterdam’s ethical review board, all participants were recruited from an Italian elementary and middle school in the North-Western area of the country (Piemonte Region). The choice of the school was convenient in nature, with a purposive element (only children between 8 and 14 years old were involved in the study). The study took place in the informatic lab of the school. As part of the recruitment process, the lead author of the study contacted the dean of the school system and associated teachers and presented the research proposal during two meetings. After school approval was received, parents were informed of their child’s potential participation in the project and could opt-out of the study if they so desired (via passive consent protocol). The total number of participants from the 6 elementary classes and 2 middle school classes was 184 children. The families of 2 children revoked consent to
participate, 9 children were absent during the last meeting where they had to take the post-test questionnaire, and one was absent from school for health reasons. The final dataset is
comprised of data from 172 children.
As Table 1shows5, at pre-test, the average age of the sample was of 9.36 years (SD = 1.31) in the playing condition and 9.23 years (SD = 1.24) in the streaming condition (t (170) = .36, p =.48). Girls accounted for 40.9% of the participants in the playing condition and for 39.3% in the streaming condition (χ2(1, N = 172) = -.02, p = .83). Participants were highly unfamiliar with the game, with 97.7% of those who played the game and 92.9% of those who watch the streamer playing who had never heard about or played the game before (χ2(1, N = 172) = -.12, p = .13). Moreover, the streamer was mostly unknown to the children: 86.4% of those in the playing condition and 83.3% of those who watched the videos had never heard about the streamer (χ2(1, N = 172) = -.04, p = .58).
5
14
Participants in the playing condition were not heavy videogame players at home (M = 2.09, SD = .74), although they had a higher (but not significant) videogame use of those in the streaming condition (M = 2.22, SD = .77; t (170) = -1.16, p = .25). On the other hand, children in the playing condition had significantly higher epistemic curiosity (M = 3.06, SD = .58) than those in the streaming condition (M = 2.84, SD = .51; t (170) = 2.61, p < .05).
Stimuli
Brain-training game. One of the most successful brain-training adventure games is
the Professor Layton franchise, where the player tries to solve a mysterious plot (Professor Layton, 2019)6. The game is a puzzle adventure game distributed by Nintendo™ worldwide for the Nintendo™ DS platform. For this study, the first episode - Professor Layton and the
Curious Village – was used.
This game was selected as an appropriate game because of its previous use in classroom setting to develop children information literacy and writing skills (Learning and Teaching Scotland, 2010) and because of features (National Research Council, 2011, 2012) that have been shown to support metacognitive skills among early adolescents, namely:
- Puzzle solving as the primary gameplay activity. Professor Layton and the Curious Village makes puzzle solving the core of its gameplay. Puzzle solving skills include logic, pattern recognition, sequence solving, the ability to adapt to novel situations, all characteristics that are very likely to make participants engage in metacognition. - Increasing level of difficulty. In Professor Layton players face numerous puzzles of
increasing difficulty. This has been deemed as extremely important when it comes to learning cognitive skills (Federation of American Scientists, 2006).
- Immediate feedback. The game continuously monitors progress and gives immediate feedback, enabling players to engage in immediate metacognitive evaluation and to start over with a new plan, keeping in mind their previous mistakes.
- Possibility of repetition. The game virtually allows players to retry each puzzle an infinite number of times through a “start again” button. This allows players to re-start
6
15
their metacognitive processes after evaluating their mistakes. Moreover, many
enigmas and puzzles give the player the possibility to draw, write, and sketch directly on the screen, allowing to devise strategies and modify the course of action.
- Cues, hints, and partial solutions. Professor Layton and the Curious Village includes the possibility to ask for cues for a maximum of 3 hints per puzzle. In this sense, cues and hints may act as “augmented information” the player can focus on in order to acquire new information and engage in self-regulative processes.
- Goal as Challenge. The goal of Professor Layton and the Curious Village is to unfold a murder and to find a treasure in the town of St. Mystere. Each step towards the solution of the game is presented as a cognitive challenge to the player. This design element makes Professor Layton and the Curious Village an interesting choice when interested in stimulating cognitive and metacognitive abilities of the audience.
In addition to these characteristics that are deemed important for supporting metacognitive skills, the game was also selected due to its popularity (Darren, 2008), wide availability, and appealing design (Valkenburg & Piotrowski, 2017) – which together provided added
confidence that the game would be appealing to participants and provide important external validity for the study.
The Streamer. With a total of over 57.000 live channels on Twitch7, each of them led by a streamer with his own style, games preferences, and so on, it was necessary to establish conditions for selecting the streamer used in this study. In addition to streaming the selected game content, the streamer should be appropriate for a young audience and be a good modelling figure for metacognitive behaviors8. In practice, this meant that the following characteristics were considered.
- Language and vocabulary: the selected streamer’s language and vocabulary should be appropriate for the age group, not too complex but clear and entertaining. Each part of the video deemed inappropriate should be edited out.
- Attitude: a positive, welcoming, and cheerful attitude is to be preferred, since it is logical to think this would increase participants’ enjoyment in the video and
16
motivation to follow her reasoning, as well as engaging in metacognitive behaviours (Schraw, Crippen, & Hartley, 2006).
- Easiness of retrieval: the video of the streamer should be easy to retrieve online, either on YouTube, Twitch or other platforms. This would increase external validity. - Distribution: the streamer should be active on well-known streaming platforms in
order to have a good reach on audience, especially the youngest one. This would increase external validity as well.
After reviewing different Italian streamers9, the choice fell on the streamer Roberta Sorge (Twitch name: CKibe). Her language is well-articulated and greatly understandable also by a younger audience, rarely including vulgarity and profanities10. Her welcoming and out-going attitude leads her towards a regular and abundant engagement about her thought processes, making the verbalization of metacognitive strategies consistent enough for this research. Second, her Professor Layton and the Curious Village gameplay videos are easily retrievable onlineand thus would be easy to incorporate in the study. Third, and last, Twitch is the most famous live streaming platform in the world. Her presence on this platform make her part of a group creating media content in the area of streaming that is among the most accessed online.
Measures
Metacognitive Knowledge and Self-Regulation Skills – Child Report. To assess
the dependent variables in this study – namely metacognitive knowledge (MK) and self-regulation of cognition (SR) – the Junior version of the Metacognitive Knowledge and Self-Regulation scale (J-MAI) developed by Sperling, Howard, Miller, & Murphy (2002) was used. The J-MAI measures metacognition broadly and across subject areas. The J-MAI is composed by 18 items, where items 1-5, 12-14, and 16 represent Metacognition Knowledge (for example, “I know when I understand something”), whereas items 6-11, 15, 17, and 18
9 The necessity to select an Italian streamer was since the entire data collection will be conducted in an Italian elementary and middle school
10
17
represent Self-Regulation processes (for example, “When I am done with my schoolwork, I
ask myself if I learned what I wanted to learn”). The present research used a scale from 1 =
Never to 4 = Often, instead a 5-point scale using Always as a highest item as in the original
version11. Moreover, the third item for metacognitive knowledge (“I try to use ways of
studying that have worked for me before”) was dropped given the low correlation it had with
the total score both when all 18 items (r = .19) and only MK ones were analysed for
reliability (r = .16), and since it was not loading on the same component both in the scale as a whole and when only MK items were analysed (r < .30). When removed, the Cronbach alpha coefficient was .82 for the whole scale, .62 for items measuring MK and .76 for questions measuring SR. Two scores were created: one for MK (M = 3.15, SD = .45) and one for SR (M = 2.97, SD = .56), with higher scores indicating greater meta-cognition.
Metacognitive Knowledge and Self-Regulation Skills – Teacher Report. In
addition to the child report, a questionnaire, also developed by Sperling, Howards, and Murphy (2002) was distributed to the classroom teacher during the last meeting, where they had to report students’ metacognitive skills on a scale from 1 = Very Low Metacognition to 6 = Very High Metacognition. This resulted in one score (M= 3.60, SD = 1.23), with higher scores indicating greater meta-cognition. As shown in Table 212, the teacher report positively correlates with child report of MK (r = .15, p = .05) and SR (r = .12, p = .11).
Epistemic Curiosity. Epistemic curiosity (EC) – the desire that motivates individuals
to obtain new knowledge (Piotrowski, Litman, & Valkenburg, 2014) - has been proven to be associated with better and deeper learning outcomes. Developing in early childhood, EC motivates individuals to persist in their search for answers when facing a problem (Litman & Mussel, 2013). EC was included as a potential covariate for this study.
11 This measure was taken since early piloting of the questionnaire gave indications that children were taking the option Always too literally and consequently never selecting it.
12
18
Epidemic curiosity was measured using the Epistemic Curiosity (I/D-YC) Questionnaire, a 10 items questionnaire (from 1 = Almost never to 4 = Almost always) developed by Piotrowski, Litman, and Valkenburg (2014). In this study, the (I) scale has an alpha of .61 and the (D) scale an alpha of .52, but an overall alpha of .72. Changes in the alpha of the two separate scales might be attributed to the translation and changes made to the questions in order to make them more kid-friendly. One score was created for epistemic curiosity (M = 2.95, SD = .56) with higher scores indicating greater EC.
Stimuli familiarity. Knowing the game and the puzzles in advance might affect the
enjoyment of the experience while playing through a lower perceived difficulty of the game (Klimmt, Blake, Hefner, Vorderer, & Roth, 2009). Due to the game’s heavy presence in the streaming environment we opted to assess familiarity of the stimulus for a potential control variable. The variable was measured with a dichotomous Yes/No question (“Have you ever
played the game Professor Layton and the Curious Village?”). Results indicated that the
95.3% of the sample did not know the game at pre-test.
Streamer familiarity. Since metacognition has been proven to be affected in its
processes by affective states (Cross & Paris, 1988), knowing the streamer might play a role in how metacognition develops if the streamer is deemed more “authentic” or familiar.
However, since it is not the aim of the current research to understand how the relationship between audience and streamers affects the enjoyment of the content, we kept knowledge of the streamer as a control variable. The variable was measured with a dichotomous Yes/No question (“Have you ever watched video or live streams of the Youtuber and Twitch streamer
CKibe?”). Results showed that 84.9% of the sample did not know the streamer at pre-test.
Game Enjoyment. Game enjoyment serves as a potential control variable in this
study. Enjoying a game means to be more in touch with the characters, more immersed in the flow of the story, and it has positive effects on learning outcomes (Gomez, Wu, and
19
Passerini, 2010). Five items from the Game Experience by Verhoeven et al. (2015) was used for this study, ranging on a scale from 1 to 5 (for example, “How did you think Professor
Layton and the Curious Village was fun?”). Items were averaged to create one scale (α =
.76., M = 4.12, SD = .65).
Video game use. Video game use serves as a potential control variable for this study.
Items were adapted from Piotrowski, Jordan, Bleakley, & Henessy (2015) and comprised 7 items on a scale from 1 = Never to 4 = Always. The questions enquired about video game use during a typical school day and during a typical week end day (i.e. “Think about a day of the
week when you go to school, say a Tuesday. How often during one of those days do you play
videogames between when you are back from school and dinner time?”). Items were averaged
to create one scale (α = .85, M = 2.15, SD = .76), with higher values indicating higher video game use.
Demographics and Children Characteristics. Participants were asked about their
gender (Boy or Girl) and age (Day, Month, and Year of birth).
Procedure
Participants were picked up from class in small groups (from a minimum of 7 to a maximum of 12, depending on the class size). In class, the researcher greeted them and then gave a brief summary of what they had to do that day. He then proceeded to call them one by one and handed out a piece of paper with a personal unique code to each of them. Children were then led to the school informatic lab. A copy of each code was sitting next to each computer and children were instructed to search for their code and sit, so that they would always use the same computer in each session. Each participant completed a computer
20
end of the meeting). Each of the children took part in one of the following two randomly assigned experimental conditions:
- Playing condition. An opensource Nintendo™ DS simulator (DeSmuME©13) was installed in the computers of the school lab. A copy of Professor Layton and the Curious Village was installed on each of them and run through the simulator. Participants had a fixed location in the lab and they always used the same computer on which their own game was saved at the end of each session. At the beginning of every session, the researcher reminded participants of the existence of the in-game helps, intervening only in case participants were experiencing issues or were seriously stuck in the game. Participants were told to play as if they were at home and they were left free to play at the pace they preferred for 15 minutes for each session. - Streaming condition. Children watched videos of a walkthrough of the game run by
the streamer using the computers in the school lab. The videos were accessed using the online platform YouTube, via the researcher private channel. Specifically, the videos entailed the first episode of the game from the streamer’s Twitch channel14. Each video was edited to cut away unnecessary parts in order to keep the flow of the experience close to that of the playing condition, apart from the streamer presence, so that they could be comparable. The total amount of videos was five of the duration of circa 15 minutes each, with the last one being of 9 minutes. Absent kids continued the video from where they left it the time before. During the last meeting, participants all watched the final videos of 9 minutes so that they were able to conclude the narrative arch of the game together.
For the first and final session, identical pre-test and post-test assessments were completed online. At the end of each session, kids were brought back to class and the new group of kids was picked up until the whole class had participated. Each of the participants received a gift in form of a small toy, regardless of their participation in the research.
Results Analytic Approach 13http://desmume.org/download/ 14 https://www.youtube.com/watch?v=GG5gpTUqfOo
21
Preliminary analyses indicated that Game Habits, Game Enjoyment, and Epistemic Curiosity were appropriate control variables. Analysis was carried out with the using SPSS Statistics™ software (version 23.0). Hypotheses were examined with a 2 x 2 (Condition [Playing, Streaming] x Gender [Male, Female]) Multivariate Analysis of Variance
(MANOVA). MANOVA was chosen in order to adjust for the possibility to incur in Type 1 errors against using many ANOVAs tests. Univariate normality was investigated for the variables included in the analysis across the two main conditions. Only MK in the streaming condition had a normal distribution. However, given a reasonable big sample (N = 172), since each cell reached the minimum threshold of 20 cases between the main and interaction effects (min.: 33) (Tabachnick and Fidell, 2001), given the values of skewness and kurtosis (ranging for all variables between -1 and +1), and after reviewing the histogram plots for each
variable, the distributions were “reasonably normal”. The multivariate normality assumption was respected (Mahalanobis max distance < 13.82 for two dependent variables). No major issues with outliers nor multivariate outliers were found (Mahalanobis max distance < 13.82). The assumptions of linearity and multicollinearity (r = .64) of dependent variables were respected, as well as the assumption of homogeneity of variance-covariance matrices (p = .086).
Hypothesis Testing
Results from the MANOVA15 revealed no statistically significant differences between the streaming and playing condition (H1a and H1b) (p = .90), nor did gender serve as a moderator on MK and SR. As such, both the main effect (H1) and interaction effect (H2) was rejected. Even more, patterns suggested that the effect of condition was in the opposite direction of what expected, with children who played the game having higher MK and SR on
15
22
average than those in the streaming condition (MStreamingMK = 3.12, MPlayingMK = 3.19;
MStreamingSR = 2.94, MPlayingSR = 2.99)16. When looking at the teacher reported metacognition as
the only dependent variable, an ordinary least squares (OLS) regression using PROCESS (Hayes, 2017) was performed. Results showed that teachers evaluated children almost the same between the two conditions (MPlaying = 3.60, SD = 1.23; MStreaming 3.61, SD = 1.23) with
no moderation by gender, even though the effect of gender was in the expected direction17, with girls outperforming boys in any group18. Thus, all the hypotheses were rejected.
Discussion
The development of metacognitive skills is one of the crucial roles of early
adolescence, and continue to influence all aspects of one’s life throughout the life course – influencing, for example, academic achievement (Zulkiply, 1994), especially for people with learning disorders (Zemira & Bracha, 2014; Flavell, 1979), but also emotions control (Keith & Frese, 2005), and can overall help children and adults to “to make wise and thoughtful life decisions” (Flavell, 1979, p 910).
At the same time, our digitalized culture has transformed early adolescence into a space that is increasingly augmented by digital media - including digital games. Research pointed out that these games (particularly of the “brain-training” variety) have been shown to support metacognitive skills – although empirical investigations into this space are relatively limited. In other researches on metacognitive support, work suggested that verbal modelling of metacognitive strategies can make significant inroads in supporting such skills. These two areas of work come together in the contemporary space of brain-training game streamers – whereby gamers verbally review their gameplay (thus modelling metacognitive strategies) of
16 See table 3, in Appendix B 17 See table 5, in Appendix B 18
23
brain-training games. It would seem, based on the literature, that such behaviours would be effective – perhaps even more so for girls than boys. The results, however, provided no such evidence. Instead, results showed that there were no significant differences in metacognition by form of play (standard gameplay versus viewing a streamer) nor was there any moderation by gender.
It is hard to identify a clear reason for null effects, particularly given the rich literature behind this question. That said, it may be that using a streamer who was verbalizing
metacognitive processes in a natural setting, while boosting ecological validity, was not enough to encourage children to focus on those tasks and instead their attention deviated to other elements of the game (the characters, the funny drawings, etc.). Since the streamer was not a trained professional conveying examples of metacognition, it also may be that the metacognitive elements in her behaviours were not strong enough to elicit a response from the participants. Also, of note, Social Learning Theory (Bandura, 1977) states that children learn from media and modelling especially when they need to solve new or difficult tasks. In the modelling condition, participant might have not perceived they had to achieve anything: the streamer was doing everything for them, without the opportunity to “use” the
metacognitive tools learned. Inasmuch, while they undoubtedly enjoyed the game, the modelling condition might not have been “task oriented” enough to activate learning
processes through motivation that are the fundamentals of SLT. In the future, an extension of this study might decide to use a modelling agent using specific scaffolding strategies devised by the researcher(s). Moreover, children should be allowed to directly apply their
metacognitive skills in the game (or in a game of the same domain), so that the activation of metacognitive processes could be adequately measured.
Equally surprising, gender differences were unapparent in this project. That said, trends (in both child and teacher report) showed that females demonstrated higher levels of
24
metacognition – regardless of condition. This is consistent with prior studies indicating that girls in this age of reference overperform boys at both MK and SR (Kolić-Vehovec,
Bajšanski, & Zubković, 2010), and provided added confidence to the measures that were used in this study. Yet, the fact that both genders responded the same way to the manipulation was indeed a surprise. While it is unclear why this is the case, it may be that it was the design itself that muddled the waters. Specifically, it may be that gender influences selection more than processing – whereby girls would opt for something different than boys, related to in-game strategies and actions – but when faced with modelled content they experience it in similar ways. The DSMM, indeed, specifically acknowledges that individual differences may influence selection, processing, or both. Previous research pointed out that variation in
responses to media is due to differences between participants more than the content itself, but that more research is needed to understand what individual differences are influencing what response (Fikkers & Piotrowski, 2019). An extension of the present study with an appropriate design looking at selection processes, like adding a part where participants can play and apply metacognition, in addition to how content is processed would be a valuable addition.
Limitations and Recommendations
This research carries with it a certain number of limitations. First, due to lack of financial resources, the informatic lab of the school had to be used instead of a more
standardized one. The lab was underfunded, and the available technology reflected this. This meant that participants used different kind of computers to take the questionnaire and
participate in the experimental conditions, (e.g. some used laptops and others used available desk computers). The difference in screen resolution and size might have induced different enjoyment of the game or made more difficult to play it, thus creating some “noise” in the model. Second, some of the kids, for a total of 28, were absent for one or more days: this
25
might have affected how conditions influenced metacognition. However, no outliers were found to be the kids who were absent for at least one day or those who were noticed being distracted during the sessions. Third, during two separate sessions there were some technical issues, namely two power outages, but no major disruptions happened and all participants in the sessions finished their questionnaire when the power returned.
The major key challenge with this research is how metacognition was measured. The J-MAI measures metacognition across subject areas while evaluating behaviours and habits of the respondents. Since habits and behaviours are difficult to change, the expectations of change on this assessment in only 5-weeks may have been too great. However, the J-MAI (and its teacher component) was chosen for its ability to measure both MK and SR, because it was built for the target sample, and because of the lack of resources for employing another method, such as qualitative interviews. For future research, it would be reasonable to consider an additional and more sensitive measure that is more task specific and closer to the
experimental condition (like the one devised by Veenman et al., 2014).
In line with the points raised above, it seems pertinent to reiterate that extensions of this work should consider a different way to manipulate the streamer condition. In this study, anecdotally, most of the young people in the streamer condition verbally expressed
disappointment about only watching the video – they wanted to play the game as well. A variation to this design whereby young people can enact their metacognitive strategies may be closer to the theoretical mechanism of modelling, be enjoyable for the participants, and even more closely approximate a ‘normal’ situation (such as a class setting), thus adding even greater external validity. Such work is necessary given the clear implications for both every day life and school settings.
In conclusion, this research ultimately makes a plea for more studies on how modelling metacognitive behaviour can be beneficial for metacognition. Adding a more
26
sensitive measurement alongside a task-based use of a modelling agent may be better able to detect potential change, allowing for greater precision, potentially unveiling novel ways for supporting youth’s metacognition through gaming.
27
Aliya, S. K. (2002). The role of computer games in the development of theoretical analysis, flexibility and reflective thinking in children: A longitudinal study. International
Journal of Psychophysiology, 45, 149.
Anderson, C. A., & Bushman, B. J. (2001). Effects of violent video games on aggressive behavior, aggressive cognition, aggressive affect, physiological arousal, and prosocial behavior: A meta-analytic review of the scientific literature. Psychological science, 12(5), 353-359.
Antonietti, A., & Mellone, R. (2003). The difference between playing games with and without the computer: A preliminary view. The Journal of Psychology, 137(2), 133-144.
Arvaja, M., Salovaara, H., Häkkinen, P., & Järvelä, S. (2007). Combining individual and group- level perspectives for studying collaborative knowledge construction in context. Learning and Instruction, 17(4), 448-459.
Azevedo, R., & Hadwin, A. F. (2005). Scaffolding self-regulated learning and
metacognition. Implications for the design of computer-based scaffolds. Instructional
science, 33(5/6), 367- 379.
Bandura, A. (1977). Social Learning Theory. New York: General Learning Press.
Bandura, A. (1989). Social cognitive theory. In: Vasta, R. (1995) Annals of child
development. Psyccritiques, 40(11).
Bavelier, D., Green, C. S., Pouget, A., & Schrater, P. (2012). Brain plasticity through the life span: learning to learn and action video games. Annual review of neuroscience, 35, 391-416.
Blumberg, F. (2007). Digital games as a context for cog devel, learning, & devel research. New Directions for Child and Adolescent Development, (115), 83–97.
28
Blumberg, F. C., & Fisch, S. M. (2013). Introduction: Digital games as a context for cognitive development, learning, and developmental research. New directions for child and adolescent development, 2013(139), 1-9.
Blumberg, F. C., & Sokol, L. M. (2004). Boys' and girls' use of cognitive strategy when learning to play video games. The Journal of General Psychology, 131(2), 151-158, doi: 10.3200/GENP.131.2.151-158
Bushman, B. J., & Geen, R. G. (1990). Role of cognitive-emotional mediators and individual differences in the effects of media violence on aggression. Journal of
personality and social psychology, 58(1), 156.
Christoph, L. H. (2006). The role of metacognitive skills in learning to solve problems. SIKS.
Cordova, D. I., & Lepper, M. R. (1996). Intrinsic motivation and the process of learning: Beneficial effects of contextualization, personalization, and choice. Journal of
educational psychology, 88(4), 715.
Coutinho, S. (2008). Self-efficacy, metacognition, and performance. North American Journal
of Psychology, 10(1).
Cowan, D. (February 15, 2008). Sailing The World: Professor Layton Tops North American
DS Charts. Gamasutra. Retrieved May 30, 2008 from URL:
https://www.gamasutra.com/php-bin/news_index.php?story=17417
Cunningam et al., (2018). Retrieved 3rd June 2019 from URL:
https://skillful- learning.org/wp-content/uploads/2018/06/2018-ASEE-Metacognition-Workshop-Workbook.pdf
Doolittle, J. H. (1995). Using riddles and interactive computer games to teach problem-solving skills. Teaching of Psychology, 22(1), 33-36.
29
Dresel, M., & Haugwitz, M. (2006). The relationship between cognitive abilities and self‐ regulated learning: evidence for interactions with academic self‐concept and gender.
High ability studies, 16(02), 201-218.
Federation of American Scientists (FES). (2006). Summit on educational games: Harnessing
the power of video games for learning.
Ferguson, C. J., & Garza, A. (2011). Call of (civic) duty: Action games and civic behavior in a large sample of youth. Computers in Human Behavior, 27(2), 770-775.
Fikkers, K.M., Piotrowski, J.T., & Valkenburg, P.M. (2019). Child’s play? Assessing the Bidirectional longitudinal relationship between gaming and intelligence in early childhood. Journal of Communication. doi:10.1093/joc/jqz003
Fisher, R (1998). Thinking about thinking: Developing metacognition in children. Early
Child Development and Care, 141, 1–15.
Fisher, R. (2002). Shared thinking: metacognitive modelling in the literacy hour. Reading, 36(2), 63-67.
Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive developmental inquiry. American Psychologist, 34(10), 906–911.
Flavell, J. H., Miller, P. H., & Miller, S. A. (1993). Cognitive development (3rd ed.). Englewood Cliffs, NJ: Prentice Hall.
Foster, S. R., Esper, S., & Griswold, W. G. (2013). From competition to metacognition: designing diverse, sustainable educational games. In Proceedings of the SIGCHI
Conference on Human Factors in Computing Systems (pp. 99-108). ACM.
Gee, J. P. (2003). What video games have to teach us about learning and literacy. Computers
30
Goldstein, H. (1995). Hierarchical data modelling in the social sciences. Journal of
Educational and Behavioral Statistics, 20(2), 201-204.
Gomez, E. A., Wu, D., & Passerini, K. (2010). Computer-supported team-based learning: The impact of motivation, enjoyment and team contributions on learning outcomes.
Computers & Education, 55(1), 378-390.
Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective attention. Nature, 423(6939), 534.
Hamilton, W. A., Garretson, O., & Kerne, A. (2014). Streaming on twitch: Fostering participatory communities of play within live mixed media. In CHI '14 proceedings
of the SIGCHI conference on human factors in computing systems (pp. 1315 - 1324).
New York: ACM
Hayes, A. F. (2017). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach. Guilford Publications.
Hogan, M. J. (2012). Prosocial effects of media. Pediatric Clinics, 59(3), 635-645.
Hong, J. C., Hwang, M. Y., Tam, K. P., Lai, Y. H., & Liu, L. C. (2012). Effects of cognitive style on digital jigsaw puzzle performance: A GridWare analysis. Computers in Human Behavior, 28(3), 920-928.
Horak, V. M. (1990). Students' Cognitive Styles and Their Use of Problem-Solving
Heuristics and Metacognitive Processes.
https://files.eric.ed.gov/fulltext/ED347069.pdf
Kafai, Y. B. (1996). Gender differences in children’s constructions of video games. In NJ: Ablex P. M. Greenfield & R. R. Cocking (Eds.), Interacting with video (pp. 39–66). Norwood
Ke, F. (2008). Alternative goal structures for computer game-based learning. International
31
Keith, N., & Frese, M. (2005). Self-regulation in error management training: emotion control and metacognition as mediators of performance effects. Journal of Applied
Psychology, 90(4), 677.
Kirsh, S. J., Mounts, J. R., & Olczak, P. V. (2006). Violent media consumption and the recognition of dynamic facial expressions. Journal of interpersonal violence, 21(5), 571-584.
Kistner, S., Rakoczy, K., Otto, B., Dignath-van Ewijk, C., Büttner, G., & Klieme, E. (2010). Promotion of self-regulated learning in classrooms: Investigating frequency, quality, and consequences for student performance. Metacognition and Learning, 5(2), 157-171.
Klimmt, C., Blake, C., Hefner, D., Vorderer, P., & Roth, C. (2009, September). Player performance, satisfaction, and video game enjoyment. In International Conference on
Entertainment Computing (pp. 1-12). Springer, Berlin, Heidelberg.
Kolić-Vehovec, S., & Bajšanski, I. (2006a). Dobne i spolne razlike u nekim vidovima
metakognicije i razumijevanja pri čitanju [Age and gender differences in some aspects of metacognition and reading comprehension]. Društvena Istraživanja, 15, 1005-1027.
Kolić-Vehovec, S., & Bajšanski, I. (2006b). Metacognitive strategies and reading
comprehension in elementary school students. European Journal of Psychology of
Education, 21, 439-451.
Kolić-Vehovec, S., Bajšanski, I., & Zubković, B. R. (2010). Metacognition and reading comprehension: Age and gender differences. Trends and prospects in metacognition
research (pp. 327-344). Springer, Boston, MA.
Kuhn, D. (2000). Metacognitive development. Current directions in psychological science, 9(5), 178-181.
32
Kwak, H., Zinkhan, G. M., & Dominick, J. R. (2002). The moderating role of gender and compulsive buying tendencies in the cultivation effects of TV show and TV
advertising: A cross cultural study between the United States and South Korea. Media
Psychology, 4(1), 77-111.
Litman, J. A., & Mussel, P. (2013). Development and validation of German translations of interest- and deprivation-type epistemic curiosity scales. Journal of Individual
Differences.
Littleton, H. L., & Ollendick, T. (2003). Negative body image and disordered eating behavior in children and adolescents: what places youth at risk and how can these problems be prevented? Clinical child and family psychology review, 6(1), 51-66.
Liu, D. R., & Ke, C. K. (2007). Knowledge support for problem-solving in a production process: A hybrid of knowledge discovery and case-based reasoning. Expert Systems
with Applications, 33(1), 147-161.
Livingstone, J. A. (2003). Metacognition: An Overview.
Lopez, V., Corona, R., & Halfond, R. (2013). Effects of gender, media influences, and traditional gender role orientation on disordered eating and appearance concerns among Latino adolescents. Journal of adolescence, 36(4), 727-736.
McClure, R. F., & Mears, F. G. (1984). Video game players: Personality characteristics and demographic variables. Psychological Reports, 55(1), 271-276.
Mohd Suki, N. (2013). Young consumer ecological behaviour: The effects of environmental knowledge, healthy food, and healthy way of life with the moderation of gender and age. Management of Environmental Quality. International Journal, 24(6), 726-737.
Moncarz, H. T. (2012). The relationship between playing games and metacognitive awareness (Doctoral dissertation).
33
Morris, B. J., Croker, S., Zimmerman, C., Gill, D., & Romig, C. (2013). Gaming science: the "Gamification" of scientific thinking. Frontiers in psychology, 4, 607.
doi:10.3389/fpsyg.2013.00607
Much Needed (2019). Retrieved on 12/12/2019 from URL: https://muchneeded.com/twitch-statistics/
National Research Council (2011). Learning Science Through Computer Games and
Simulations. Washington, DC: National AcademiesPress.
National Research Council (2012). A Framework for K-12 Science Education: Practices,
Crosscutting Concepts, and Core Ideas. Washington, DC: National Academies Press.
Perfect, T. J., & Schwartz, B. L. (2002) Introduction: toward an applied metacognition. In Perfect, T. J., & Schwartz, B. L. (2002). Applied metacognition. Cambridge University Press.
Phan, M. H., Jardina, J. R., Hoyle, S., & Chaparro, B. S. (2012, September). Examining the role of gender in video game usage, preference, and behavior. Proceedings of the
human factors and ergonomics society annual meeting (Vol. 56, No. 1, pp.
1496-1500). Sage CA: Los Angeles, CA: Sage Publications.
Pifarre, M., & Cobos, R. (2010). Promoting metacognitive skills through peer scaffolding in a CSCL environment. International Journal of Computer-Supported Collaborative
Learning, 5(2), 237-253.
Pillay, H., Brownlee, J., & Wilss, L. (1999). Cognition and recreational computer games: Implications for educational technology. Journal of research on computing in
education, 32(1), 203-216.
Piotrowski, J.T., Jordan, A.B., Bleakley, A. & Hennessy, M. (2015) Identifying Family Television Practices to Reduce Children’s Television Time, Journal of Family
34
Piotrowski, J.T., Litman, J.A., & Valkenburg, P.M. (2014). Measuring epistemic curiosity in young children. Infant and Child Development, 23(5), 542-533. doi:
10.1002/icd.1847
Professor Layton (2019). The Ultimate Puzzle Awaits. URL:
http://professorlayton.nintendo.com/azran-legacy/
Richard, G. T. (2013). Gender and gameplay: Research and future directions. Playing with virtuality. Theories and methods of computer game studies, 269-284.
Rideout, V. J., Foehr, U. G., & Roberts, D. F. (2010). Generation M 2: Media in the Lives of 8-to 18-Year-Olds. Henry J. Kaiser Family Foundation.
Salovaara, H. (2005). An exploration of students’ strategy use in inquiry-based computer-supported collaborative learning. Journal of Computer Assisted Learning, 21(1), 39– 52.
Scardamalia, M., Bereiter, C., & Steinbach, R (1984) Teachability of reflective processes in written composition. Cognitive Science 8, 2, 173–190.
Schraw, G., & Dennison, R. S. (1994). Assessing metacognitive awareness. Contemporary
educational psychology, 19(4), 460-475.
Schraw, G., Crippen, K. J., & Hartley, K. (2006). Promoting self-regulation in science education: Metacognition as part of a broader perspective on learning. Research in Science
Sjöblom, M., & Hamari, J. (2017). Why do people watch others play video games? An empirical study on the motivations of Twitch users. Computers in Human Behavior, 75, 985-996
Sperling, R. A., Howard, B. C., Miller, L. A., & Murphy, C. (2002). Measures of children's knowledge and regulation of cognition. Contemporary educational psychology, 27(1), 51-79.
35
Squire, K. (2011). Video games and learning. Teaching and participatory culture in the
digital age. New York, NY: Teachers College Print.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Boston: Allyn & Bacon
Tarhini, A., Hone, K., & Liu, X. (2014). Measuring the moderating effect of gender and age on e-learning acceptance in England: A structural equation modelling approach for an extended technology acceptance model. Journal of Educational Computing Research, 51(2), 163-184.
Terlecki, M., Brown, J., Harner-Steciw, L., Irvin-Hannum, J., Marchetto-Ryan, N., Ruhl, L., & Wiggins, J. (2011). Sex differences and similarities in video game experience, preferences, and self-efficacy: Implications for the gaming industry. Current
Psychology, 30(1), 22-33.
Trick, L. M., Jaspers-Fayer, F., & Sethi, N. (2005). Multiple-object tracking in children: the “Catch the Spies” task. Cognitive Development, 20, 373–387.
Twitch Statistics and Charts (2019, April the 30th). Retrieved from URL:
https://twitchtracker.com/statistics
Valkenburg, P. M., & Peter, J. (2013a). The differential susceptibility to media effects model.
Journal of Communication, 63(2), 221-243.
Valkenburg, P. M., & Peter, J. (2013b). Comm Research—Views from Europe Five Challenges for the Future of Media-Effects Research. International Journal of
Communication, 7, 19.
Valkenburg, P. M., & Piotrowski, J. T. (2017). Plugged in: How media attract and affect
36
Valkenburg, P. M., Peter, J., & Walther, J. B. (2016). Media effects: Theory and research.
Annual review of psychology, 67, 315-338.
van der Stel, M., Veenman, M.V. (2010). Development of metacognitive skillfulness: A longitudinal study, Learning and Individual Differences, 20, 3, pp. 220-224, URL:
https://doi.org/10.1016/j.lindif.2009.11.005
Van Deventer, S. S., & White, J. A. (2002). Expert behavior in children’s video game play.
Simulation & Gaming, 33(1), 28-48.
Veenman, M. V., Bavelaar, L., De Wolf, L., & Van Haaren, M. G. (2014). The on-line assessment of metacognitive skills in a computerized learning environment. Learning
and Individual Differences, 29, 123-130.
Veenman, M. V., Van Hout-Wolters, B. H., & Afflerbach, P. (2006). Metacognition and learning: Conceptual and methodological considerations. Metacognition and learning, 1(1), 3-14.
Verhoeven, K., Abeele, V. V., Gers, B., & Seghers, J. (2015). Energy expenditure during Xbox Kinect play in early adolescents: The relationship with player mode and game enjoyment. Games for Health Journal, 4(6), 444-451.
Wang, M. C., Haertel, G. D., & Walberg, H. J. (1990). What influences learning? A content analysis of review literature. The Journal of Educational Research, 84(1), 30-43.
Weil, L. G., Fleming, S. M., Dumontheil, I., Kilford, E. J., Weil, R. S., Rees, G., ... & Blakemore, S. J. (2013). The development of metacognitive ability in adolescence. Consciousness and cognition, 22(1), 264-271.
Wells, D. (2008, June). Professor Layton and the Curious Village. Hyper. Next Media (176): 60. ISSN 1320-7458
37
Whitebread, D., Anderson, H., Coltman, P., Page C., Pino Pasternak D., & Mehta S. (2005). Developing independent learning in the early years. Education, 33, 40–50. URL:
http://dx.doi.org/10.1080/03004270585200081
Wouters, P., Paas, F., & van Merrienboer, J. J. G. (2008). How to optimize learning from animated models: A review of guidelines based on cognitive load. Review of
Educational Research, 78(3), 645–675.
Zemira, M., & Bracha, K. (2014). Educational Research and Innovation Critical Maths for
Innovative Societies. The Role of Metacognitive Pedagogies: The Role of Metacognitive Pedagogies. OECD publishing.
Zimmerman, B. J. (1998). Developing self-fulfilling cycles of academic regulation: An analysis of exemplary instructional models. In D. Schunk & B. Zimmerman (Eds.),
Self-regulated learning: From teaching to self-reflective practice (pp. 1 - 19). New
York, NY: Guilford.
Zimmerman, B. J. 1989. A social cognitive view of self-regulated academic learning. Journal of Educational Psychology, 81: 329–339.
Zimmerman, B. J. 2000. “Attainment of self-regulation: A social cognitive perspective”. In Handbook of self-regulation, Edited by: Boekaerts, M., Pintrich, P. R. and Zeidner, M. 13–39. San Diego, CA: Academic Press.
Zimmerman, B. J., & Christakis, D. A. (2005). Children’s television viewing and cognitive outcomes: a longitudinal analysis of national data. Archives of Pediatrics &
Adolescent Medicine, 159(7), 619-625.
Zulkiply, N. (2009). Metacognition and its relationship with students' academic performance. The International Journal of Learning, 15(11), 97-106.
38 Appendix A
Image 1
The Differential Suceptibility to media Effects Model (DSMM)
Image 2 and 3
39
Image 4
Screenshot of a part of the game in Professor Layton and the Curious Village
Image 5
The streamer CKibe solving a puzzle in Professor Layton and the Curious Village (Italian version)
40 Appendix B Table 1
Sample Descriptives by playing and streaming condition Conditions Playing Streaming M SD % M SD % N Age 9.36 1.31 9.23 1.31 172 Videogame use 2.09 .74 2.22 .77 172 Epistemic Curiosity 3.06 .58 2.84 .51 172 Game Enjoyment 4.13 .63 4.12 .67 172 Self-esteem 3.54 1.14 2.11 .78 172 Gender (females) 40.9 39.3 Game Familiarity (unfamiliar) 97.7 92.9 Streamer Familiarity (unfamiliar) 86.4 83.3
41
Table 2
Correlation Table for MK, SR, and Teachers Assessment
Notes: *** p < .000
Variable 1 2 3
1. MK
2. SR .64***
42
Table 3
Descriptives for MK, SR, and Teachers Assessment by playing and streaming condition Conditions Playing Streaming M SD Min. Max. M SD M SD N MK 3.15 .45 1.75 4.00 3.19 .47 3.12 .44 172 SR 2,97 .56 1.22 4.00 2.99 .58 2.94 .53 172 Teachers Assessment 3.60 1.23 1.00 6.00 3.60 1.23 3.61 1.23 172
43
Table 4
Multivariate Analysis of Variance with Wilk’s Lambda Tests
Wilks’ Lambda Tests Between-Subjects Effects
DV Value F (df) Mean Square F (df) Adj. R2
Intercept .70** 34.60 (164) MK 12.69*** 68.52 (1) SR 4.51*** 16.81 (1) Epistemic Curiosity .90** 9.58 (164) MK 2.23** 12.05 (1) SR 4.80*** 17.90 (1) Videogame Use .97 2.17 (164) MK .20 1.09 (1) SR .18 .69 (1) Game Enjoyment .98 1.93 (164) MK .22 1.17 SR 1.04 3.87 Gender .99 1.04 (164) MK .25 1.35 SR .52 1.93 Exp. Condition 1.00 .11 (164) MK .01 .06 SR .01 .02 Condition*Gender 1.00 .13 (164) MK .02 .12 SR .00 .01 Corrected Model MK .77** 4.15 (6) .10 SR 1.43*** 5.33 (6) .13
R squared = .16, Adj. R Squared .13.
Notes: **p < .001, ***p < .000. Independent variable: Exp. Condition. Dependent Variables: MK, SR. Conditional Effect: Gender. Covariates: Epistemic Curiosity, Videogame Use, Game Enjoyment.
