The Role of Epistemic Curiosity in Game-Based Learning: Distinguishing Skill Acquisition From Adaptation

Abstract

Background

Videogames are emerging as increasingly popular tools for training complex skills, in part due to their potential for improving the transfer of learning to changing demands. One caveat is that training outcomes are primarily influenced by a learner’s capability and willingness to engage in adaptive learning processes. The present study examined the role of epistemic curiosity as a predisposition for seeking information and resolving gaps in knowledge.

Method

Using a game-based learning environment with strong cognitive and perceptual-motor demands, we tested the effects of interest- and deprivation-type epistemic curiosity on performance during skill acquisition and adaptation to unforeseen change. A task-change paradigm and piecewise latent trajectory model disentangled the effects on performance trajectories corresponding to skill acquisition and adaptation.

Results

Interest-type epistemic curiosity, but not deprivation-type, was positively related to baseline performance, even after controlling for cognitive ability and videogame experience. Moreover, the positive effect of interest-type epistemic curiosity on post-change performance (i.e., adaptation) was fully mediated by pre-change performance.

Discussion

These findings indicate that interest-type epistemic curiosity is particularly important in the earlier phases of learning, but explains little variability in adaptive performance beyond its direct influence on performance acquired prior to change. This research speaks to the roles of individual differences in cognition and motivation during game-based learning.

Keywords

adaptation epistemic curiosity game-based learning interest latent trajectory model skill acquisition

Due to the accelerating pace of technological development, workers are increasingly expected to adapt to change (Hoffman et al., 2014; Pulakos et al., 2000). One of the biggest challenges for training the modern workforce is fostering the adaptability of trained skills beyond instructional settings. Fortunately, the same technology responsible for changing work demands has provided many opportunities for improving training. Videogames and synthetic learning environments have become increasingly popular for training complex skills (Bell et al., 2008; Cannon-Bowers & Bowers, 2010). Game-based learning is the use of an interactive, learner-controlled environment with clear rules and objectives, in which learners receive immediate feedback on their actions (Sitzmann, 2011; Vogel et al., 2006). These features of game-based learning enable learners to explore and experiment with different actions, which can motivate continued practice and facilitate deeper processing of training content, thus promoting transfer of skills to a broader range of anticipated and unanticipated demands (Bell & Kozlowski, 2008; Carolan et al., 2014).

Although empirical support is promising, the effectiveness of game-based learning depends on many factors (Clark et al., 2016; Sitzmann, 2011; Vogel et al., 2006; Wilson et al., 2009). Broadly, the complexity of learner-controlled environments can overwhelm the cognitive capacities of learners (Kirschner et al., 2006). Many studies have made progress toward reducing cognitive load and increasing motivation with game design features (Wilson et al., 2009; Wouters & van Oostendorp, 2013). However, less research has examined the role of learners’ characteristics that influence game-based learning outcomes. Individual differences often account for more variability in performance than design features, which makes them essential considerations (Kaber et al., 2002). Learners differ in their attitudes toward games (Landers & Armstrong, 2017), motivation to engage in self-directed learning (Bell & Kozlowski, 2002), and tendency to immerse oneself in gameplay (Sacau et al., 2008). As such, differences in learning outcomes might be attributed in part to individual differences in the tendency to engage in adaptive learning processes.

Epistemic curiosity (EC), defined as the drive to seek information and eliminate gaps in knowledge, is considered an essential predisposition for learning complex skills (Kang et al., 2009; von Stumm et al., 2011). Individuals with higher EC are more likely to be stimulated by complexity, which motivates them to engage in learning toward reducing it (Loewenstein, 1994; Mussel, 2013). Among similar predispositions related to learning (e.g., intrinsic motivation), EC is the most proximal to exploratory behavior, which is a central process in active learning (Hardy et al., 2018; Keith & Wolff, 2014). Despite many reviews advocating for the importance of EC (Grossnickle, 2016; Mussel, 2013; von Stumm et al., 2011), our search of the extant literature found no empirical studies on its role in game-based learning.

In the present study, we examined the relationship between EC and learning a complex videogame with strong cognitive and perceptual-motor demands. Our goal was to develop a better understanding of the extent to which EC plays a role in game-based learning. Drawing on active learning and resource allocation theories (Kanfer & Ackerman, 1989; Keith & Wolff, 2014), we tested hypotheses related to the effects of EC throughout the learning process. Latent growth models were used to distinguish between phases of the learning process. This approach, combined with a task-change paradigm, enabled us to distinguish the processes of adaptive transfer following an unforeseen change in task demands (Lang & Bliese, 2009). Because learners differ in their cognitive capacities and tendencies to allocate resources toward learning, we disentangled the effects of EC from other variables known to be important to learning (i.e., cognitive ability and prior experience). This research contributes to a growing literature by examining the role of individual differences in navigating the complexity of learner-controlled, game-based learning environments.

Epistemic Curiosity

Videogames are designed to be engaging for players, which is often accomplished by stimulating players’ curiosity. Epistemic curiosity (EC) is the “desire for knowledge that motivates individuals to learn new ideas, eliminate information gaps, and solve intellectual problems” (Litman, 2008, p. 1586). EC is distinct from perceptual curiosity, which refers to the “drive to experience and feel” (Berlyne, 1954; Collins et al., 2004). Both epistemic and perceptual curiosity are likely to be elicited during game-based learning, albeit by different game features. EC is stimulated by collative variables (e.g., ambiguity, complexity, and uncertainty), whereas perceptual curiosity is evoked by visual, auditory, and tactile sensations (Collins et al., 2004). Due to the complexity of videogames as active learning environments, stimulation of EC is expected to motivate learning directly.

EC belongs to a network of constructs known as intellectual investment traits and considered important for learning (Mussel, 2013; von Stumm & Ackerman, 2013). Openness to experience, typical intellectual engagement, and need for cognition are just a few examples of constructs that are strongly correlated with EC, raising concerns about its discriminant validity. Based on curiosity theory, EC is distinct for its coupling with domain-specific interest (Grossnickle, 2016). For example, one might be curious about science, but not art (Fayn et al., 2017; Silvia, 2008). Many studies have measured EC as a domain-general personality trait (e.g., Harrison et al., 2011; Kashdan et al., 2009; Koo & Choi, 2010), but research on state curiosity has shown that individuals express different levels of curiosity over time and across situations (e.g., Loewenstein, 1994). Individuals vary in their interests and may subsequently vary in levels of EC for different domains. Indeed, measurement specificity affects the magnitude of correlations with performance (Völkle et al., 2006). The principle of Brunswik symmetry posits that criterion-related validity is maximized when the predictors and criteria are matched in terms of content and breadth (Wittmann & Süß, 1999). As such, EC is best understood in the context of the performance domain in which it is being examined.

Research has distinguished between two dimensions of EC (Litman, 2008; Litman & Jimerson, 2004). Interest-type (I-type) EC refers to the anticipated pleasure of learning, whereas deprivation-type (D-type) EC is the drive to reduce uncertainty and resolve information gaps. Both dimensions are related to openness, intrinsic motivation, and exploration (Litman et al., 2010; Mussel, 2010; Richards et al., 2013), but are rooted in distinct self-regulatory processes (Litman, 2005, 2010). I- and D-type curiosity correspond to broad and narrow appraisals, respectively, which makes them somewhat analogous to diversive and specific curiosity (Litman & Jimerson, 2004). Given that little empirical literature exists regarding their unique relationships with learning outcomes, we examined the following research question.

Research Question 1: What are the relative effects of I- and D-type EC on learning?

Individuals with higher levels of EC experience more intense states of curiosity across a broader range of situations, which promote engagement and exploratory behavior (Kashdan et al., 2009; Naylor, 1981). For example, individuals with higher I-type EC are more likely to play games for longer periods (Kim & Lee, 2017). Studies have identified positive effects of EC on performance (Hardy et al., 2017; Hassan et al., 2015; Kang et al., 2009; Reio & Wiswell, 2000), but little is known about its relationship with learning over time. Likewise, EC has been promoted as a facilitator of adaptation to change (Harrison et al., 2011; Mussel, 2013; Mussel & Spengler, 2015). However, our search of the literature revealed no empirical examinations of EC concerning learning curves or adaptation to change.

Active Learning Theory

Game-based learning is associated with active learning theory, a constructivist perspective that emphasizes the roles of exploration and experimentation for acquiring a deeper understanding of training content (Keith & Wolff, 2014; Wu et al., 2012). Developing complex skills that generalize to changing demands implies that learners are capable of self-regulating their behavior without relying on instructional support. Videogames offer many opportunities for creating immersive environments representative of how tasks are performed outside of instructional settings. Active learning theory highlights the role of learner control for building complex skills, with the caveat that it puts much of the responsibility for learning on the learners themselves.

Many tasks that need to be trained for the modern workforce cannot simply be broken down into components and rehearsed to proficiency. Complex tasks possess many moving parts that interact with each other and change over time (Liu & Li, 2012; Wood, 1986). Although research has shown that active learning environments can be effective for training complex skills (Bell & Kozlowski, 2008; Carolan et al., 2014; Hughes et al., 2013), the empirical evidence is largely mixed (Kirschner et al., 2006). Task complexity strains learners’ cognitive capacities to make sense of the task environment and encode concepts into long-term memory (Kalyuga, 2009). Providing learners with control over their learning can introduce unnecessary cognitive demands that could otherwise be avoided with instructional scaffolding. However, active learning can be more effective for some learners than others (Kalyuga, 2009).

Decades of research have demonstrated individual differences in the acquisition of complex skills (Ackerman, 1988, 1992, 2007). Learners differ in their cognitive resource capacities and their willingness to allocate those resources toward learning. Moreover, the effects of cognitive and motivational variables on learning are known to change throughout the learning process (Kanfer & Ackerman, 1989). These effects are reflected in models of performance change. Theoretically, there are three stages to acquiring skills (Anderson, 1982; Fitts & Posner, 1967; Tenison & Anderson, 2016). In the beginning of skill acquisition, demands on learners’ attention are highest as they initially make sense of the task (Kanfer & Ackerman, 1989). As learners become familiar with the task, they subsequently compile knowledge schemas and performance strategies, which reduce attentional demands. Eventually, learners tend to settle into routines that require minimal attention, which is observed as the rate of performance change decreases toward an asymptote.

In the beginning of skill acquisition, performance is more influenced by one’s cognitive resource capacity. Cognitive ability, defined as a general aptitude for efficiently processing information and solving problems, is well-known to be related to attaining higher levels of performance (Ackerman, 1988). Learners with higher cognitive ability have a greater capacity for processing a greater number of task elements in less time. Aside from receiving direct instruction, there are generally two ways in which a learner can approach the complexity of encountering a new task (Sweller, 2006). First, the learner can “borrow” concepts and procedures from previous experiences that are relevant for the present. Orvis et al. (2009) observed that one of the strongest predictors of performance during game-based learning was previous experience with videogames. Learners with relevant experience can draw on knowledge and skills that are effective at performing the task being learned. In cases where concepts and procedures are not directly applicable, they can sometimes be adapted, which is more efficient than constructing task strategies from scratch.

The second way in which learners approach complexity is by exploring and attempting task strategies via trial-and-error (Sweller, 2006). Hardy and colleagues (Hardy et al., 2014, 2018); have shown that exploration is positively related to learning outcomes in a learner-controlled, game-based environment. After accounting for cognitive ability and prior experience with videogames, exploration is expected to prompt the discovery of effective performance strategies. Both I- and D-type EC are tied to exploratory behavior, albeit with different self-regulatory explanations. I-type EC is associated with optimistic appraisals of uncertainty and the tendency to set learning goals. It correlates with openness to experience (Litman & Mussel, 2013), learning goal orientation (Litman, 2008), and intrinsic motivation (Litman et al., 2010). Like I-type EC, D-type EC is thought to facilitate learning via exploratory behavior. However, exploratory behavior motivated by D-type EC differs from I-type EC in that it is directed toward attaining a state of resolution (Lauriola et al., 2015; Litman, 2008). Rather than exploring for pleasure, individuals experiencing a state of D-type EC are driven to eliminate discrepancies between their knowledge and the environment. As such, learners with higher EC are expected to attain higher levels of performance in the beginning of skill acquisition.

Hypothesis 1: EC is positively related to baseline performance.

As learners continue to practice, the role of aptitude decreases while the role of motivation increases (Kanfer & Ackerman, 1989). Cognitive demands are reduced as learners consolidate many interacting elements into smaller chunks of concepts and procedures (Johnson-Laird, 1980). Learners with higher cognitive ability and videogame experience do so more quickly, as reflected by a faster rate of performance gains (Ackerman, 2007). Eventually, the rate of performance gain tends to plateau as learners settle on automated procedures. These strategies can be effective, but suboptimal, especially in complex tasks without an upper-bound for performance (Dörner, 1980). Achieving higher levels of performance requires continued cognitive engagement, along with further exploration and experimentation (Hardy et al., 2019). Learners with higher levels of EC are expected to maintain higher levels of cognitive engagement (Smillie et al., 2016). Learners with higher EC perceive higher levels of novelty, which motivates them to persist in exploring task strategies that can improve performance (Litman, 2008). Likewise, feelings of deprivation are proposed to be uncomfortable and associated with need for closure (Litman, 2010). Individual differences in the tendency to maintain intense, focused attention have been shown to accelerate learning (Espejo et al., 2005). As such, we expected to see a relationship between EC and the rate of performance gains.

Hypothesis 2: EC is positively related to skill acquisition, such that learners with higher EC gain performance at a faster rate.

Adaptive Transfer

For complex skills, the goal of training is rarely the automation of routine tasks. Rather, training is often needed to promote adaptation to non-routine situations beyond what was initially learned (Pulakos et al., 2000). Advocates of active learning theory suggest that the mixed empirical evidence surrounding learner-controlled training is partly due to the breadth of the criterion being measured (Keith & Wolff, 2014). That is, learner-controlled training might hinder short-term performance while facilitating development of the self-regulatory processes needed to adapt to change. Due to the complexity of learner-controlled environments, minimal guidance can lead to slower learning in the short-term relative to direct instruction (Kirschner et al., 2006). However, learners who construct their understanding of the task via exploration and experimentation are expected to develop more nuanced mental schemas that are robust to changes in task demands (Keith & Wolff, 2014).

Evaluating adaptive transfer is difficult, considering that performance environments vary in the extent to which they resemble initial learning conditions (Barnett & Ceci, 2002). Researchers have distinguished adaptive transfer from initial learning using a task-change paradigm in which performance is measured before and after a manipulated change in task demands (e.g., Bell & Kozlowski, 2008; Hughes et al., 2013). As seen in skill acquisition, individuals perform at different levels and learn at different rates (Ackerman, 1988; Ackerman et al., 1995). Lang and Bliese (2009) extended this reasoning to the task-change paradigm, which distinguishes between two processes of adaptive transfer. Transition adaptation is the level of performance immediately following a change in task demands, whereas reacquisition adaptation is the subsequent rate of performance improvement. Individuals differ in the extent to which they can immediately adapt to change, and the rate at which they can learn to perform in the post-change environment.

Previous research on the role of individual differences in adaptive transfer of learning has produced counterintuitive findings. Lang and Bliese (2009) found that higher levels of cognitive ability were associated with larger decrements in performance following a change in task demands, which contradicted traditional expectations (Beier & Oswald, 2012). Cognitive ability is frequently associated with the ability to adapt to change. However, as suggested by Lang and Bliese (2009), learners with higher cognitive ability might be quicker to achieve higher proficiency with particular task strategies, which are subsequently relied on due to their automaticity. That is, learners rely on strategies that were effective prior to change, which might be less adaptive following change. If so, learners who attain higher levels of performance during skill acquisition are expected to demonstrate similar patterns. A simpler alternative explanation is that those with higher cognitive ability have more to lose in transition adaptation as a result of higher pre-change performance (Beier & Oswald, 2012). Therefore, it is important to account for pre-change performance when examining the direct effects of other individual differences on post-change performance.

Although knowledge, skills, and abilities are necessary for successful adaptation to change, they are insufficient. Individuals must also be motivated to put effort toward modifying their performance (Chen & Firth, 2014). Several willingness traits are associated with adaptive transfer, including emotional stability, conscientiousness, and mastery goal orientation (Bell & Kozlowski, 2008; Huang et al., 2014; Pulakos et al., 2002; Shoss et al., 2012). Openness to experience in particular is known for having conceptual relationships with complex learning (Huang et al., 2014; von Stumm et al., 2011). Learners with higher levels of openness tend to seek out learning experiences and prefer novelty. However, research on its relationship with adaptive transfer is mixed (Huang et al., 2014; Pulakos et al., 2002). Whereas openness is a broad personality trait, EC is conceptually more proximal to learning behaviors, and is thus more likely to have apparent effects on adaptive transfer.

Following a change in task demands, many strategies practiced before the change are no longer effective. Learners with higher levels of I-type EC are expected to more quickly and effectively explore the post-change task environment, thus generating a broader repertoire of new strategies. Hardy et al. (2017) found that individuals with higher I-type EC generated more potential solutions to ill-defined problems. I-type EC is also related to positive appraisals of complexity (Litman, 2010), which can moderate the stress on cognitive resources initiated by change. D-type EC is also related to tolerance of ambiguity and increased persistence when challenged (Lauriola et al., 2015). Individuals with higher levels of D-type EC are considered less likely to respond maladaptively to change, and more likely to persist in resolving its accompanying discrepancies. D-type EC is more intense than I-type EC (Lauriola et al., 2015), which might be expressed as greater sensitivity to change. Following an unforeseen change in task demands, learners with high levels of D-type EC are expected to allocate effort toward resolving performance decrements. Given the intentionality of their effort, these individuals are expected to employ metacognitive and emotion regulation strategies, which facilitate systematic problem solving. As such, we examined the following hypothesis.

Hypothesis 3: EC is positively related to transition adaptation (i.e., higher initial post-change performance).

Adaptation involves recognizing changes in the environment, modifying existing performance routines, and developing new strategies (Jundt et al., 2015). As in skill acquisition, greater cognitive capacity should correspond to faster rates of learning. One key difference in adaptation is that learners possess more proximal experiences with the learning environment prior to change. Learners might be inclined to rely on procedures that are ill suited to the post-change environment, or draw on a broader range of experiences. Throughout initial skill acquisition, learners may discover task strategies or cues via exploration that are not utilized immediately. Rather, they might be recalled and applied during adaptation, thus resulting in a direct effect of EC on reacquisition adaptation.

Hypothesis 4: EC is positively related to reacquisition adaptation, such that learners with higher EC gain performance at a faster rate following the task change.

Method

The objective of this correlational study was to model the effects of EC on performance during skill acquisition and adaptation. A task-change paradigm and piecewise latent trajectory model were used to estimate change in videogame performance and its relationships with self-reported individual differences. Path estimates were considered statistically significant at an alpha level of .05. This research was approved by the university’s Institutional Review Board and the data were collected as part of a larger research effort. All participants gave informed consent prior to beginning the study.

Sample

Participants were 214 undergraduate students (41.6% female, M_age = 19.20, SD = 1.70, Range = 18–32) at the University of Oklahoma who received research participation credit toward a psychology course. In addition, participants had the opportunity to win entries into a gift card drawing for attaining high performance scores during the study.

Protocol

The game was Unreal Tournament 2004 (UT2004; Epic Games, 2004), a first-person shooter PC game used in previous research on game-based learning (e.g., Hardy et al., 2014; Hughes et al., 2013). UT2004 involves strong cognitive and perceptual-motor demands, making it a suitable environment for studying individual differences in complex skill learning. As a dynamic and fast-paced game, UT2004 shares many characteristics with simulations and games used for training complex skills in real-world applications (Kozlowski et al., 2001). Participants’ objective was to destroy as many computer-controlled opponents as possible while minimizing the destruction of their character. UT2004 was designed to be easy to learn, but difficult to master. Participants were challenged to learn the advantages and disadvantages of each weapon, power-up, and strategy, and quickly decide which to use in the moment. There were unlimited opportunities to explore the map and collect weapons and power-ups scattered throughout. To control their character, each participant used a mouse and keyboard.

Over 4 hours, participants practiced UT2004. Prior to playing the game, participants completed assessments of the individual difference predictor variables. Participants then watched a 15-min presentation explaining the basic game controls, rules, and power-ups, followed by a 1-min practice trial to become familiar with the display, controls, and game map without opponents. Participants completed 28 trials in which they had 4 minutes to obtain the highest score possible. For the first 14 trials (i.e., pre-change), participants competed against two computer-controlled opponents at a difficulty setting of 4 (i.e., “Skilled”) on a 1-to-8 scale. Detailed performance feedback was displayed after each trial.

Following the 14^th trial, many elements of the game were changed without warning. Participants competed against nine computer-controlled opponents at a difficulty setting of 5 (i.e., “Adept”). In addition, the map was much larger, with multiple levels of platforms and edges over which characters could fall to their destruction. These changes corresponded to an increase in complexity, as representative of dynamic, real-world environments (Hughes et al., 2013). Performing in the post-change environment required different strategies for attaining high performance than prior to the change. For example, the post-change map was more open, such that it was more difficult to hold a defensive position while attacking opponents from afar. Participants were debriefed following the 28^th trial.

Measures

Cognitive ability via self-reported ACT scores and videogame experience were measured as control variables. Videogame experience was measured using the scale from Hardy et al. (2014). Two items were scored with a 5-point Likert scale ranging from 1 (not at all) to 5 (daily). “Over the last 12 months, how frequently have you typically played video/computer games?” (M = 2.92, SD = 1.42) and “Over the last 12 months, how frequently have you typically played first-person shooter video/computer games (e.g., Call of Duty, Half-Life, Halo, Unreal Tournament)?” (M = 2.35, SD = 1.33). For the remaining two items, participants reported how many hours per week they typically play videogames (M = 4.61, SD = 6.58) and first-person shooter videogames (M = 2.03, SD = 4.03). Scores for each item were standardized and averaged into a composite score.

EC was measured using Litman’s (2008) I-type (I-EC) and D-type EC (D-EC) scales adapted to the domain of videogames. Five items assessed I-EC (e.g., “I enjoy trying new things when playing videogames;” α = .88) and five items assessed D-EC (e.g., “I spend hours on challenging videogames because I can’t rest until I have figured them out;” α = .83). Responses were scored with a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Consistent with Litman (2008), confirmatory factor analysis supported the correlated 2-factor model of EC (CFI = .968; RMSEA = .071 [CI₉₀ = .041–.099], SRMR = .049) over the 1-factor model (χ²_diff(1) = 80.66, p < .001).

Performance for each trial was calculated using the UT2004 performance index (Epic Games, 2004; Hardy et al., 2014:): the quantity of kills (i.e., number of times a participant destroyed an opponent) divided by the quantity of kills plus deaths (i.e., number of kills plus the number of times a participant’s own character was destroyed) plus rank (i.e., the participant’s rank relative to the opponents in that game). Performance scores across the 28 trials were collapsed into 14 measurement occasions by averaging the scores for each pair of trials to improve model stability (Howe, 2019). Scores were then multiplied by 100 for easier interpretation.

Analysis

We examined performance change using piecewise latent trajectory models, which estimate latent intercepts and slopes corresponding to trajectories before and after specified discontinuities (Bollen & Curran, 2006; Flora, 2008). In the present study, there were 14 repeated measures nested within individuals and one discontinuity corresponding to the task change following the 7th measurement occasion. We estimated both pre- and post-change intercepts and slopes, which were interpreted based on their loadings with observed variables (Flora, 2008). We specified loadings using a relative coding scheme in which post-change latent trajectories are interpreted relative to those of pre-change (Table 1; Bliese & Lang, 2016). With this specification, the pre-change intercept and slope respectively correspond to baseline performance and skill acquisition. Transition adaptation, corresponding to the post-change intercept, is the change in performance relative to the expected level of performance at the 8^th measurement occasion had the task change not occurred. Likewise, the post-change slope representing reacquisition adaptation is the linear rate of performance change relative to skill acquisition. All models were computed using robust maximum likelihood estimation in Mplus Version 8.1 (Muthén & Muthén, 1998–2018).

Table 1.

Factor Loadings of Repeated Measures on Piecewise Latent Trajectory Model Variables.

	Pre-change sessions							Post-change sessions
Latent variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14
Relative coding
Baseline performance	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Skill acquisition	0	1	2	3	4	5	6	7	8	9	10	11	12	13
Quadratic skill acquisition	0	1	4	9	16	25	36	36	36	36	36	36	36	36
Transition adaptation	0	0	0	0	0	0	0	1	1	1	1	1	1	1
Reacquisition adaptation	0	0	0	0	0	0	0	0	1	2	3	4	5	6
Absolute coding
Baseline performance	1	1	1	1	1	1	1	0	0	0	0	0	0	0
Skill acquisition	0	1	2	3	4	5	6	0	0	0	0	0	0	0
Quadratic skill acquisition	0	1	4	9	16	25	36	0	0	0	0	0	0	0
Transition adaptation	0	0	0	0	0	0	0	1	1	1	1	1	1	1
Reacquisition adaptation	0	0	0	0	0	0	0	0	1	2	3	4	5	6

Note. See Bliese and Lang (2016) for details on contrasts for discontinuous models.

Results

Means, standard deviations, and intercorrelations are reported in Table 2 and observed performance scores are displayed in Figure 1. The intraclass correlation coefficient (ICC1) indicated that 72% of the variance in performance resided at the between-person level. Following Bliese and Lang (2016), we systematically built an unconditional model of performance change without any predictors (Figure 2). First, we estimated the latent intercepts followed by linear slopes. Participants started with an average score of 27.48 points (SE = 1.25, p < .001) and gained 5.66 points (SE = .41, p < .001) on average per session. Following the task change, performance immediately declined –19.01 points (SE = .88, p < .001) on average. The rate of performance change was slower in the subsequent post-change trials relative to skill acquisition (η = –4.86, SE = .43, p < .001). All latent variables varied significantly between participants. Lastly, we tested for the presence of quadratic trends in performance, as expected by skill acquisition theory (Fitts & Posner, 1967). Indeed, the rate of performance change decreased over time (η = −.60, SE = .07, p < .001), indicating that participants typically approached a stable level of proficiency by the end of the pre-change sessions. A curvilinear pattern was not detected for post-change performance. Overall, the unconditional model was an excellent fit to the data (CFI = .991, RMSEA = .040 [CI₉₀ = .018–.058], SRMR = .025). Parameter estimates are displayed in Table 3.

Table 2.

Means, Standard Deviations, and Correlations of Study Variables.

Variable	M	SD	1	2	3	4	5	6
1. Gender	—	—	—
2. Cognitive ability	25.94	6.29	−.17*	—
3. Videogame experience	0.00	1.00	−.55***	.17*	.88
4. I-type epistemic curiosity	3.58	0.94	−.52***	.08	.62***	.91
5. D-type epistemic curiosity	2.74	0.93	−.38***	.11	.46***	.67***	.81
6. Average pre-change performance	36.73	18.32	−.76***	.29***	.63***	.61***	.41***	—
7. Average post-change performance	19.08	16.43	−.69***	.30***	.61***	.58***	.39***	.90***

Note.N = 214. Diagonals are Cronbach’s alphas.

p < .05, ** p < .01, *** p < .001, two-tailed.

Table 3

Parameter Estimates for the Unconditional Piecewise Latent Trajectory Model.

Parameter	Estimate	SE
Means
Baseline performance	27.48***	1.25
Skill acquisition	5.66***	0.41
Quadratic skill acquisition	−0.60***	0.07
Transition adaptation	−19.01***	0.88
Reacquisition adaptation	−4.86***	0.43
			Correlations
Variances			1	2	3
1. Baseline performance	291.41***	22.82
2. Skill acquisition	1.75***	0.41	.13
3. Transition adaptation	86.04***	15.80	−.50***	−.77***
4. Reacquisition adaptation	2.41***	0.65	.10	−.63***	.42**

Note.N = 214. k = 2996. Loadings were specified with relative contrasts (Table 1).

p < .01, *** p < .001, two-tailed.

Prior to adding covariates to the unconditional model, we respecified covariances between latent variables as direct paths to address issues regarding interpretations of covariate effects on transition adaptation (Beier & Oswald, 2012). Specifically, we estimated direct effects of the pre-change latent variables (i.e., baseline performance and skill acquisition) on the post-change intercept (Model 1; Table 4). Larger declines in performance following the task change were observed for individuals with higher baseline performance (β = −.41, SE = .08, p < .001) and who acquired skill at a faster rate (β = −.72, SE = .08, p < .001).

Table 4.

Standardized Path Coefficients for Conditional Piecewise Latent Trajectory Models.

		Pre-change				Post-change
		Intercept		Slope		Intercept		Slope
Model	Variable	β	SE	β	SE	β	SE	β	SE
1	Baseline performance					−0.41***	0.08
	Skill acquisition					−0.72***	0.08
2	Baseline performance					−0.30**	0.11
	Skill acquisition					−0.70***	0.08
	Gender	−0.58***	.05	0.09	.08	−0.17	0.13	0.12	.13
	Cognitive ability	0.13**	.05	−0.04	.05	0.14	0.13	0.09	.09
	Videogame experience	0.29***	.06	−0.04	.08	0.07	0.12	0.18	.12
3	Baseline performance					−0.33**	0.11
	Skill acquisition					−0.71***	0.08
	Gender	−0.53***	.05	−0.13	.13	0.09	0.08	0.07	.14
	Cognitive ability	0.14**	.05	0.15	.13	−0.03	0.05	0.08	.09
	Videogame experience	0.21**	.07	0.00	.14	−0.06	0.08	0.26	.13
	Interest-type EC	0.23**	.07	0.14	.20	0.10	0.09	−0.25	.18
	Deprivation-type EC	−0.05	.07	0.00	.16	−0.06	0.08	0.10	.17

Note.N = 214. k = 2996. Gender was coded 0 = Male; 1 = Female. EC = epistemic curiosity; pre-change intercept = baseline performance; pre-change slope = skill acquisition.

p < .01, *** p < .001, two-tailed.

To gather a comprehensive perspective of transition and reacquisition adaptation, we also estimated an equivalent model in which the factor loadings were respecified with an absolute coding scheme (Table 1). With this specification, transition and reacquisition adaptation are respectively interpreted as the absolute level of performance and rate of performance change following the task change. When re-specifying the model with absolute coding, baseline performance (β = .91, SE = .03, p < .001) and skill acquisition (β = .18, SE = .08, p = .024) were positively related to absolute levels of performance immediately following the task change. Together, these results indicate that participants with higher performance during the pre-change sessions experienced a larger post-change decline relative to others, but still maintained higher absolute levels of performance.

Next, the direct effects of gender, cognitive ability, and videogame experience on the pre- and post-change intercepts and slopes were added to the model. As shown in Model 2 (Table 4), gender was negatively related to the baseline performance, with lower scores for females (β = −.58, SE = .05, p < .001). Gender did not yield any other statistically significant effects. Both GMA (β = .13, SE = .05, p < .01) and videogame experience (β = .29, SE = .06, p < .001) were positively related to baseline performance, but neither was significantly related to the remaining latent trajectories. Cognitive ability and videogame experience were not significantly related to skill acquisition and not significantly related to transition nor reacquisition adaptation after controlling for baseline performance and skill acquisition.

We tested hypotheses regarding the effects of EC on performance and a research question regarding the extent to which effects differ between I- and D-EC. A precondition for testing the hypotheses is determining whether EC was considerably related to performance. Both I- and D-EC were significantly correlated with performance before (r = .61 and r = .58 respectively) and after the task change (r = .41 and r = .39 respectively). When adding EC to the piecewise latent trajectory model, I-EC (β = .23, SE = .07, p = .002) but not D-EC (β = −.05, SE = .07, p = .430) was significantly and positively related to baseline performance. As such, the data supported Hypothesis 1 with respect to I-EC but not D-EC. Neither EC variable was significantly related to the remaining latent trajectories (Model 3; Table 4), which failed to support Hypotheses 2-4. This final model was a slightly better fit for the data than the unconditional model (CFI = .994, RMSEA = .029 [CI₉₀ = .000–.045], SRMR = .027). Addressing the research question, I-EC effects were consistently larger than those of D-EC.

Supplemental Analysis

In addition to estimating the direct effects of EC on latent trajectories, we were interested in the indirect effects of EC on performance from a cross-sectional perspective. We tested the indirect effect of I-EC on post-change performance mediated by pre-change performance while controlling for D-EC, cognitive ability, and videogame experience. We used Hayes’ (2017) bootstrapping approach with the PROCESS macro in SPSS to conduct the analysis. Indeed, the relationship between I-EC and the post-change performance was fully mediated by pre-change performance. The indirect effect was positive and statistically significant (ab = 0.21, [CI₉₅ = .10, .31], p = .001) while the direct effect was nonsignificant (c’ = 0.03, p = .548).

Discussion

The present study contributes to a growing literature on learner-controlled, game-based learning by examining the role of individual differences throughout the learning process. We specifically focused on the role of EC, which was hypothesized to have positive effects on the acquisition and adaptation of a complex skill. Using a task-change paradigm, we disentangled performance trajectories before and after a manipulated change in task demands. Piecewise latent trajectory models estimated direct effects on adaptive transfer while controlling for pre-training proficiency and performance attained prior to the task change. Contrary to our predictions, EC did not have direct effects on transition or reacquisition adaptation. The only robust effect beyond the control variables was that of I-EC on baseline performance. However, the overall pattern of results suggests that EC is a distal predictor of adaptive transfer.

Epistemic Curiosity and Learning

I-EC positively related to the level of performance following the task change, but pre-change performance fully mediated this relationship. Learners reporting higher levels of I-EC maintained higher performance, but did not learn at a faster rate. Contrary to the assumption that exploration during pre-change sessions would produce a broader repertoire of task strategies useful in the post-change sessions, there were no direct effects of I-EC on adaptive transfer. Rather, experimenting with strategies could be more important in the earlier phases of learning. Prior studies have observed that most exploratory behavior occurs in the beginning of skill acquisition and decreases over time (Hardy et al., 2014, 2018). Additionally, I-EC might facilitate adaptive transfer indirectly via increased cognitive engagement (Smillie et al., 2016). Learners with higher I-EC appear more likely to allocate effort toward learning, regardless of the direction of that effort.

Contrary to our predictions, D-EC lacked a consistent relationship with performance. Although zero-order correlations with performance were strong and positive, these effects were no longer present when controlling for gender, GMA, and videogame experience. This indicates that sensitivity to detecting information-knowledge gaps did not have incremental validity for predicting game-based learning. Past research has also observed smaller effects of D-EC compared to I-EC (Hardy et al., 2017; Powell et al., 2017). The enjoyment of acquiring knowledge is perhaps a stronger motivator than striving to resolve gaps in knowledge. Although D-EC shares many characteristics with I-EC, it uniquely relates to factors that hinder performance, like a tendency to focus on negative outcomes (Lauriola et al., 2015). This echoes research on approach and avoidance motivation, which suggests that approaching positive outcomes facilitates learning more than avoiding negative outcomes (Elliot, 1999).

Learners with higher I-EC, but not necessarily higher D-EC, are more likely to benefit from game-based learning environments. By examining the effects on different performance trajectories, this study provided insight into how EC is associated with learning. The effect of I-EC was uniform throughout the different phases of skill acquisition and adaptation, such that learners with higher I-EC maintained higher levels of performance overall. The size of the effect was comparable with that of videogame experience. This implies that individual differences in EC might resemble those of other constructs related to sustaining high levels of task engagement (Mussel, 2013). None of the hypotheses inspired by curiosity theory were supported, indicating that self-regulatory effects of EC might operate at different levels of analysis. Although D-EC had negligible effects on performance, sensitivity to information-knowledge gaps might play an important role as a self-regulatory variable at the within-person level (Hardy et al., 2019). Nevertheless, the results of the present study suggest that game-based learning environments should be designed more so to capitalize on I-EC rather than D-EC, for example by emphasizing the enjoyment of trying new strategies and tactics more so than addressing weaknesses or gaps in current understanding and performance.

Cognitive Demands of Adaptive Performance

These results have implications for the study of individual differences in the capacity for adapting to change. Previous research suggests that individuals with higher cognitive ability are quicker to become entrenched in performance routines, which temporarily hinders adaptation (Lang & Bliese, 2009). In the present study, controlling for pre-change performance accounted for what would otherwise be an effect of cognitive ability on the decline in performance following change. This suggests that high-ability individuals have relatively “more to lose” under changing conditions and that individual differences in performance are maintained, if not elevated with practice (Ackerman, 2007; Beier & Oswald, 2012). Neglecting to control for pre-change performance with relative contrasts can make variables that facilitate learning appear maladaptive. Moreover, reporting discontinuous models specified in terms of absolute and relative change assists with model interpretation (Bliese & Lang, 2016).

By disentangling effects on different phases of learning, the present study contributed toward understanding which individuals are most likely to benefit from game-based learning environments. Cognitive ability, videogame experience, and EC are important for performing under conditions of high complexity with minimal guidance. In the present study, the effects of these variables on performance were uniform, such that effects were not moderated by the phase of learning. Indeed, the most robust predictor of successful adaptation to change is the level of knowledge and skill acquired prior to the disruption. Much remains to be learned about the extent to which processes of adaptation differ from initial skill acquisition. As the workplace continues to become increasingly complex, individual differences that foster adaptability will become increasingly important.

Limitations and Future Directions

There are some limitations to consider when interpreting these results. Task characteristics and study design are important qualifiers for interpreting effects on learning. Although UT2004 is representative of game-based training used in real-world applications, it distinctively involves a fast-paced environment with substantial perceptual-motor demands (Hughes et al., 2013). Research has shown that motivational variables can have negligible or even detrimental effects on performance for such tasks because they rely heavily on implicit learning (DeShon & Alexander, 1996; DeShon et al., 1996). Performance incentives were used to encourage engagement in this study, but may have constrained the variance in performance attributable to intrinsic motivation by fostering more of a performance- versus learning-oriented environment (Kozlowski & Bell, 2006). Future experimental designs could uncover how EC effects are moderated by contextual features. Although our design was effective at distinguishing phases of learning throughout a short period, much remains to be learned about the role of EC for learning over longer periods, such as that of a career.

Our measurement of EC covered just a small range of the broader network of constructs related to dealing with complexity (Mussel, 2013; von Stumm & Ackerman, 2013). Measuring EC as a domain-specific predisposition instead of a general personality trait decreased the likelihood of confounding these constructs. However, that performance effects were mostly limited to I-EC raises questions about the distinction between curiosity and interest (Grossnickle, 2016). The extent to which curiosity and interest are distinct in terms of affective and cognitive processes remains to be understood (Shin & Kim, 2019). Whereas the present study investigated between-person differences in EC, future studies ought to highlight the within-person dynamics and person-task interactions (Hardy et al., 2019). Understanding how EC changes over time and across situations can provide insight into self-regulated learning processes and inform the use of game design for improving learning outcomes. As technology continues to advance and become more affordable, game-based learning environments have potential for facilitating learning that generalizes beyond instructional settings.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Jonathan T. Huck, MS, is a doctoral candidate in the Industrial and Organizational Psychology program at the University of Oklahoma. His research focuses on the development of complex skills for knowledge workers and technological applications for improving assessment and training outcomes.

Contact: huck@ou.edu

Eric Anthony Day, PhD, is professor and chair of the Department of Psychology at the University of Oklahoma, USA. His research interests span personnel psychology and organizational behavior, including topics in training and development, complex skill acquisition and adaptability, group dynamics, and leadership.

Contact: eday@ou.edu

Li Lin, PhD, is a senior analyst in Global Talent Management at PepsiCo, Inc. She specializes in using quantitative analytic techniques to understand business/organizational needs and provide strategic insights for organizational decisions. Her interests include employee turnover, workforce planning, career development, measurement and assessments.

Contact: li.lin-1@ou.edu

Ashley G. Jorgensen, MS, is a doctoral candidate in the Industrial and Organizational Psychology program at the University of Oklahoma. Her research interests include emotion variability and regulation, complex skill acquisition, and adaptive performance.

Contact: jorgensen@ou.edu

Joseph Westlin, PhD, is an assistant professor of Management in the Huether School of Business at the College of Saint Rose. His research interests include complex skill acquisition, employee absenteeism, and leadership.

Contact: westlinj@mail.strose.edu

Jay H. Hardy III, PhD, is an assistant professor of Management in the College of Business at Oregon State University. His research focuses on (a) understanding how self-regulated learning processes can be leveraged for improving dynamic training and development interventions, (b) exploring the implications of applicant experience for shaping applicant behavior, and (c) applying simulation and computational modeling methodologies to understanding practical HR phenomena.

Contact: jay.hardy@oregonstate.edu

References

Ackerman

P. L.

(1992). Predicting individual differences in complex skill acquisition: dynamics of ability determinants. Journal of Applied Psychology, 77, 598–614.

Ackerman

P. L.

(1988). Determinants of individual differences during skill acquisition: Cognitive abilities and information processing. Journal of Experimental Psychology: General, 117, 288–318. http://doi.org/10.1037/0096-3445.117.3.288

Ackerman

P. L.

(2007). New developments in understanding skilled performance. Current Directions in Psychological Science, 16, 235–239. https://doi.org/10.1111/j.1467-8721.2007.00511.x

Ackerman

P. L.

Kanfer

Goff

(1995). Cognitive and noncognitive determinants and consequences of complex skill acquisition. Journal of Experimental Psychology: Applied, 1, 270–304. https://doi.org/10.1037/1076-898X.1.4.270

Anderson

J. R.

(1982). Acquisition of cognitive skill. Psychological Review, 89, 369–406.

Barnett

S. M.

Ceci

S. J.

(2002). When and where do we apply what we learn?: A taxonomy for far transfer. Psychological Bulletin, 128, 612–637.

Beier

M. E.

Oswald

F. L.

(2012). Is cognitive ability a liability? A critique and future research agenda on skilled performance. Journal of Experimental Psychology: Applied, 18, 331–345. https://doi.org/10.1037/a0030869

Bell

B. S.

Kanar

A. M.

Kozlowski

S. W.

(2008). Current issues and future directions in simulation-based training in North America. The International Journal of Human Resource Management, 19, 1416–1434. https://doi.org/10.1080/09585190802200173

Bell

B. S.

Kozlowski

S. W.

(2002). Adaptive guidance: Enhancing self-regulation, knowledge, and performance in technology-based training. Personnel Psychology, 55, 267–306. https://doi.org/10.1111/j.1744-6570.2002.tb00111.x

10.

Bell

B. S.

Kozlowski

S. W. J.

(2008). Active learning: Effects of core training design elements on self-regulatory processes, learning, and adaptability. Journal of Applied Psychology, 93, 296–316. https://doi.org/10.1037/0021-9010.93.2.296

11.

Berlyne

D. E.

(1954). A theory of human curiosity. British Journal of Psychology, 45, 180–191. https://doi.org/10.1111/j.2044-8295.1954.tb01243.x

12.

Bliese

P. D.

Lang

J. W. B.

(2016). Understanding relative and absolute change in discontinuous growth models: Coding alternatives and implications for hypothesis testing. Organizational Research Methods, 19, 562–592. https://doi.org/10.1177/1094428116633502

13.

Bollen

K. A.

Curran

P. J.

(2006). Latent curve models: A structural equation perspective. John Wiley.

14.

Cannon-Bowers

J. A.

Bowers

(2010). Synthetic learning environments: On developing a science of simulation, games, and virtual worlds for training. In Kozlowski

Salas

(Eds.), Learning, Training, and Development in Organizations (pp. 229–263). Mahwah, NJ: Lawrence Erlbaum Associates.

15.

Carolan

T. F.

Hutchins

S. D.

Wickens

C. D.

Cumming

J. M.

(2014). Costs and benefits of more learner freedom: Meta-analyses of exploratory and learner control training methods. Human Factors, 56, 999–1014. https://doi.org/10.1177/0018720813517710

16.

Chen

Firth

B. M.

(2014).The motivational underpinnings of adaptability. In Chan

(Ed.), Individual adaptability to changes at work: New directions in research (pp. 18–35). New York, NY: Routledge.

17.

Clark

D. B.

Tanner-Smith

E. E.

Killingsworth

S. S.

(2016). Digital games, design, and learning: A systematic review and meta-analysis. Review of Educational Research, 86(1), 79–122. https://doi.org/10.3102/0034654315582065

18.

Collins

R. P.

Litman

J. A.

Spielberger

C. D.

(2004). The measurement of perceptual curiosity. Personality and Individual Differences, 36, 1127–1141.

19.

Dörner

(1980). On the difficulties people have in dealing with complexity. Simulation & Games, 11, 87–106.

20.

DeShon

Alexander

(1996). Goal setting effects on implicit and explicit learning of complex tasks. Organizational Behavior and Human Decision Processes, 65, 18–36.

21.

DeShon

R. P.

Brown

K. G.

Greenis

J. L.

(1996). Does self-regulation require cognitive resources? An evaluation of resource allocation models of goal setting. Journal of Applied Psychology, 81, 595–608.

22.

Elliot

A. J.

(1999). Approach and avoidance motivation and achievement goals. Educational Psychologist, 34, 169–189. https://doi.org/10.1207/s15326985ep3403_3

23.

Epic Games. (2004). Unreal Tournament 2004. Atari.

24.

Espejo

Day

E. A.

Scott

(2005). Performance evaluations, need for cognition, and the acquisition of a complex skill: An attribute–treatment interaction. Personality and Individual Differences, 38(8), 1867–1877. https://doi.org/10.1016/j.paid.2004.10.003

25.

Fayn

Silvia

P. J.

MacCann

Tiliopoulos

(2017). Interested in different things or in different ways? Journal of Individual Differences, 38(4), 265–273.

26.

Fitts

P. M.

Posner

M. I.

(1967). Human performance. Monterey, CA: Brooks/Cole.

27.

Flora

D. B.

(2008). Specifying piecewise latent trajectory models for longitudinal data. Structural Equation Modeling: A Multidisciplinary Journal, 15, 513–533. https://doi.org/10.1080/10705510802154349

28.

Grossnickle

E. M.

(2016). Disentangling curiosity: Dimensionality, definitions, and distinctions from interest in educational contexts. Educational Psychology Review, 28, 23–60. https://doi.org/10.1007/s10648-014-9294-y

29.

Hardy

J. H.

Day

E. A.

Arthur

(2019). Exploration-exploitation tradeoffs and information-knowledge gaps in self-regulated learning: Implications for learner-controlled training and development. Human Resource Management Review, 29, 196–217. https://doi.org/10.1016/j.hrmr.2018.07.004

30.

Hardy

J. H.

Day

E. A.

Hughes

M. G.

Wang

Schuelke

M. J.

(2014). Exploratory behavior in active learning: A between- and within-person examination. Organizational Behavior and Human Decision Processes, 125, 98–112. https://doi.org/10.1016/j.obhdp.2014.06.005

31.

Hardy

J. H.

Day

E. A.

Steele

L. M.

(2018). Interrelationships among self-regulated learning processes: Toward a dynamic process-based model of self-regulated learning. Journal of Management, 45, 3146–3177. https://doi.org/10.1177/0149206318780440

32.

Hardy

J. H.

Ness

A. M.

Mecca

(2017). Outside the box: Epistemic curiosity as a predictor of creative problem solving and creative performance. Personality and Individual Differences, 104, 230–237. https://doi.org/10.1016/j.paid.2016.08.004

33.

Harrison

S. H.

Sluss

D. M.

Ashforth

B. E.

(2011). Curiosity adapted the cat: The role of trait curiosity in newcomer adaptation. Journal of Applied Psychology, 96, 211–220. https://doi.org/10.1037/a0021647

34.

Hassan

M. M.

Bashir

Mussel

(2015). Personality, learning, and the mediating role of epistemic curiosity: A case of continuing education in medical physicians. Learning and Individual Differences, 42, 83–89. https://doi.org/10.1016/j.lindif.2015.07.018

35.

Hayes

A. F.

(2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Publications.

36.

Hoffman

R. R.

Ward

Fiore

S. M.

Feltovich

P. J.

DiBello

Andrews

D. H.

(2014). Accelerated expertise: Training for high proficiency in a complex world. Psychology Press.

37.

Howe

(2019). General mental ability and goal type as antecedents of recurrent adaptive task performance. Journal of Applied Psychology, 104, 796–813. https://doi.org/10.1037/apl0000379

38.

Huang

J. L.

Ryan

A. M.

Zabel

K. L.

Palmer

(2014). Personality and adaptive performance at work: A meta-analytic investigation. Journal of Applied Psychology, 99, 162–179. https://doi.org/10.1037/a0034285

39.

Hughes

M. G.

Day

E. A.

Wang

Schuelke

M. J.

Arsenault

M. L.

Harkrider

L. N.

Cooper

O. D.

(2013). Learner-controlled practice difficulty in the training of a complex task: Cognitive and motivational mechanisms. Journal of Applied Psychology, 98, 80–98. https://doi.org/10.1037/a0029821

40.

Johnson-Laird

P. N.

(1980). Mental models in cognitive science. Cognitive Science, 4(1), 71–115. https://doi.org/10.1016/S0364-0213(81)80005-5

41.

Jundt

D. K.

Shoss

M. K.

Huang

J. L.

(2015). Individual adaptive performance in organizations: A review. Journal of Organizational Behavior, 36(S1), S53–S71. https://doi.org/10.1002/job.1955

42.

Kaber

D. B.

Draper

D. V.

Usher

J. M.

(2002). Influence of individual differences on application design for individual and collaborative immersive virtual environments. In Hale

K. S.

Stanney

K. M.

(Eds.), Handbook of virtual environments: Design implementation, and applications (pp. 379–402). CRC Press.

43.

Kalyuga

(2009). Instructional designs for the development of transferable knowledge and skills: A cognitive load perspective. Computers in Human Behavior, 25, 332–338. https://doi.org/10.1016/j.chb.2008.12.019

44.

Kanfer

Ackerman

P. L.

(1989). Motivation and cognitive abilities: An integrative/aptitude-treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74, 657–690. https://doi.org/10.1037/0021-9010.74.4.657

45.

Kang

M. J.

Hsu

Krajbich

I. M.

Loewenstein

McClure

S. M.

Wang

J. T.

Camerer

C. F.

(2009). The wick in the candle of learning: Epistemic curiosity activates reward circuitry and enhances memory. Psychological Science, 20, 963–973. https://doi.org/10.1111/j.1467-9280.2009.02402.x

46.

Kashdan

T. B.

Gallagher

M. W.

Silvia

P. J.

Winterstein

B. P.

Breen

W. E.

Terhar

Steger

M. F.

(2009). The curiosity and exploration inventory-II: Development, factor structure, and psychometrics. Journal of Research in Personality, 43, 987–998. https://doi.org/10.1016/j.jrp.2009.04.011

47.

Keith

Wolff

(2014).Encouraging active learning. In Kraiger

Passmore

dos Santos

N.R.

Malvezzi

(Eds.), The Wiley-Blackwell handbook of the psychology of training, development, and performance improvement (pp. 92–116). John Wiley & Sons: Chichester, UK.

48.

Kim

Lee

(2017). Mobile gamer’s epistemic curiosity affecting continuous play intention. Focused on players’ switching costs and epistemic curiosity. Computers in Human Behavior, 77, 32–46. https://doi.org/10.1016/j.chb.2017.08.023

49.

Kirschner

P. A.

Sweller

Clark

R. E.

(2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41, 75–86. https://doi.org/10.1207/s15326985ep4102_1

50.

Koo

D.-M.

Choi

Y.-Y.

(2010). Knowledge search and people with high epistemic curiosity. Computers in Human Behavior, 26, 12–22. https://doi.org/10.1016/j.chb.2009.08.013

51.

Kozlowski

S. W. J.

Bell

B. S.

(2006). Disentangling achievement orientation and goal setting: Effects on self-regulatory processes. Journal of Applied Psychology, 91, 900–916. https://doi.org/10.1037/0021-9010.91.4.900

52.

Kozlowski

S. W. J.

Toney

R. J.

Mullins

M. E.

Weissbein

D. A.

Brown

K. G.

Bell

B. S.

(2001). Developing adaptability: A theory for the design of integrated-embedded training systems. In Salas

(Ed.), Advances in human performance and cognitive engineering research (pp. 59–123). Elsevier Science; JAI Press.

53.

Landers

R. N.

Armstrong

M. B.

(2017). Enhancing instructional outcomes with gamification: An empirical test of the technology-enhanced training effectiveness model. Computers in Human Behavior, 71, 499–507. https://doi.org/10.1016/j.chb.2015.07.031

54.

Lang

J. W. B.

Bliese

P. D.

(2009). General mental ability and two types of adaptation to unforeseen change: Applying discontinuous growth models to the task-change paradigm. Journal of Applied Psychology, 94, 411–428. https://doi.org/10.1037/a0013803

55.

Lauriola

Litman

J. A.

Mussel

De Santis

Crowson

H. M.

Hoffman

R. R.

(2015). Epistemic curiosity and self-regulation. Personality and Individual Differences, 83, 202–207. https://doi.org/10.1016/j.paid.2015.04.017

56.

Litman

J. A.

(2005). Curiosity and the pleasures of learning: Wanting and liking new information. Cognition and Emotion, 19(6), 793–814. https://doi.org/10.1080/02699930541000101

57.

Litman

J. A.

(2008). Interest and deprivation factors of epistemic curiosity. Personality and Individual Differences, 44, 1585–1595. https://doi.org/10.1016/j.paid.2008.01.014

58.

Litman

J. A.

(2010). Relationships between measures of I- and D-type curiosity, ambiguity tolerance, and need for closure: An initial test of the wanting-liking model of information-seeking. Personality and Individual Differences, 48, 397–402. https://doi.org/10.1016/j.paid.2009.11.005

59.

Litman

J. A.

Crowson

H. M.

Kolinski

(2010). Validity of the interest- and deprivation-type epistemic curiosity distinction in non-students. Personality and Individual Differences, 49, 531–536. https://doi.org/10.1016/j.paid.2010.05.021

60.

Litman

J. A.

Jimerson

T. L.

(2004). The measurement of curiosity as a feeling of deprivation. Journal of Personality Assessment, 82, 147–157. https://doi.org/10.1207/s15327752jpa8202_3

61.

Litman

J. A.

Mussel

(2013). Validity of the interest- and deprivation-type epistemic curiosity model in Germany. Journal of Individual Differences, 34, 59–68. https://doi.org/10.1027/1614-0001/a000100

62.

Liu

(2012). Task complexity: A review and conceptualization framework. International Journal of Ergonomics, 42, 553–568. https://doi.org/10.1016/j.ergon.2012.09.001

63.

Loewenstein

(1994). The psychology of curiosity: A review and reinterpretation. Psychological Bulletin, 116, 75–98. https://doi.org/10.1037/0033-2909.116.1.75

64.

Mussell

(2010). Epistemic curiosity and related constructs: Lacking evidence of discriminant validity. Personality and Individual Differences, 49, 506–510.

65.

Mussel

(2013). Intellect: A theoretical framework for personality traits related to intellectual achievements. Journal of Personality and Social Psychology, 104, 885–906. https://doi.org/10.1037/a0031918

66.

Mussel

Spengler

(2015). Investigating intellect from a trait activation perspective: Identification of situational moderators for the correlation with work-related criteria. Journal of Research in Personality, 55, 51–60.

67.

Muthén

L. K.

Muthén

B. O.

(1998–2018). Mplus user’s guide: Statistical analysis with latent variables.

68.

Naylor

F. D.

(1981). A state-trait curiosity inventory. Australian Psychologist, 16(2), 172–183.

69.

Orvis

K. A.

Horn

D. B.

Belanich

(2009). An examination of the role individual differences play in videogame-based training. Military Psychology, 21, 461–481.

70.

Powell

Nettelbeck

Burns

N. R.

(2017). The incremental validity of intellectual curiosity and confidence for predicting academic performance in advanced tertiary students. Personality and Individual Differences, 116, 51–56. https://doi.org/10.1016/j.paid.2017.04.011

71.

Pulakos

E. D.

Schmitt

Dorsey

D. W.

Arad

Borman

W. C.

Hedge

J. W.

(2002). Predicting adaptive performance: Further tests of a model of adaptability. Human Performance, 15, 299–323.

72.

Pulakos

E. D.

Arad

Donovan

M. A.

Plamondon

K. E.

(2000). Adaptability in the workplace: Development of a taxonomy of adaptive performance. Journal of Applied Psychology, 85, 612–624. https://doi.org/10.1037/0021-9010.85.4.612

73.

Reio

T. G.

Wiswell

A. K.

(2000). Field investigation of the relationship among adult curiosity, workplace learning, and job performance. Human Resource Development Quarterly, 11, 5–30.

74.

Richards

J. B.

Litman

J. A.

Roberts

D. H.

(2013). Performance characteristics of measurement instruments of epistemic curiosity in third-year medical students. Medical Science Educator, 23, 355–363.

75.

Sacau

Laarni

Hartmann

(2008). Influence of individual factors on presence. Computers in Human Behavior, 24, 2255–2273. https://doi.org/10.1016/j.chb.2007.11.001

76.

Shin

D. D.

Kim

(2019). Homo curious: Curious or interested? Educational Psychology Review, 31, 853–874. https://doi.org/10.1007/s10648-019-09497-x

77.

Shoss

M. K.

Witt

L. A.

Vera

(2012). When does adaptive performance lead to higher task performance? Journal of Organizational Behavior, 33, 910–924. https://doi.org/10.1002/job.780

78.

Silvia

P. J.

(2008). Interest—The curious emotion. Current Directions in Psychological Science, 17, 57–60. https://doi.org/10.1111/j.1467-8721.2008.00548.x

79.

Sitzmann

(2011). A meta-analytic examination of the instructional effectiveness of computer-based simulation games. Personnel Psychology, 64, 489–528. https://doi.org/10.1111/j.1744-6570.2011.01190.x

80.

Smillie

L. D.

Varsavsky

Avery

R. E.

Perry

(2016). Trait intellect predicts cognitive engagement: Evidence from a resource allocation perspective. European Journal of Personality, 30, 215–226. https://doi.org/10.1002/per.2059

81.

Sweller

(2006). How the human cognitive system deals with complexity. In Elen

Clark

R. E.

(Eds.), Handling complexity in learning environments: Theory and research (pp. 13–26). Elsevier.

82.

Tenison

Anderson

J. R.

(2016). Modeling the distinct phases of skill acquisition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42, 749–767. https://doi.org/10.1037/xlm0000204

83.

Vogel

J. J.

Vogel

D. S.

Cannon-Bowers

Bowers

C. A.

Muse

Wright

(2006). Computer gaming and interactive simulations for learning: A meta-analysis. Journal of Educational Computing Research, 34, 229–243. https://doi.org/10.2190/FLHV-K4WA-WPVQ-H0YM

84.

Völkle

M. C.

Wittmann

W. W.

Ackerman

P. L.

(2006). Abilities and skill acquisition: A latent growth curve approach. Learning and Individual Differences, 16, 303–319. https://doi.org/10.1016/j.lindif.2006.01.001

85.

von Stumm

Ackerman

P. L

. (2013). Investment and intellect: A review and meta-analysis. Psychological Bulletin, 139(4), 841–869. https://doi.org/10.1037/a0030746

86.

von Stumm

Hell

Chamorro-Premuzic

. (2011). The hungry mind: Intellectual curiosity is the third pillar of academic performance. Perspectives on Psychological Science, 6, 574–588. https://doi.org/10.1177/1745691611421204

87.

Wilson

K. A.

Bedwell

W. L.

Lazzara

E. H.

Salas

Burke

C. S.

Estock

J. L.

Conkey

(2009). Relationships between game attributes and learning outcomes: Review and research proposals. Simulation & Gaming, 40, 217–266. https://doi.org/10.1177/1046878108321866

88.

Wittmann

W. W.

Süß

H.-M.

(1999). Investigating paths between working memory, intelligence, knowledge, and complex problem-solving performances via Brunswik symmetry. In Ackerman

P. L.

Kyllonen

P. C.

Roberts

R. D.

(Eds.), Learning and individual differences: Process, trait, and content determinants (pp. 77–108). American Psychological Association.

89.

Wood

R. E.

(1986). Task complexity: Definition of the construct. Organizational Behavior and Human Decision Processes, 37, 60–82. https://doi.org/10.1016/0749-5978(86)90044-0

90.

Wouters

van Oostendorp

(2013). A meta-analytic review of the role of instructional support in game-based learning. Computers & Education, 60, 412–425. https://doi.org/10.1016/j.compedu.2012.07.018

91.

W.-H.

Hsiao

H.-C.

P.-L.

Lin

C.-H.

Huang

S.-H.

(2012). Investigating the learning-theory foundations of game-based learning: A meta-analysis. Journal of Computer Assisted Learning, 28, 265–279. https://doi.org/10.1111/j.1365-2729.2011.00437.x