• No results found

Curious Omosa: Does player satisfaction increase the more they learn about their game environment?

N/A
N/A
Protected

Academic year: 2021

Share "Curious Omosa: Does player satisfaction increase the more they learn about their game environment?"

Copied!
20
0
0

Loading.... (view fulltext now)

Full text

(1)

Master’s Thesis, 15 ECTS

Master’s Program (one year) in Cognitive Science, 60 ECTS Spring 2020

Supervisor: Dr. Linus Holm

Curious Omosa: Does

player satisfaction

increase the more they

learn about their game

environment?

(2)
(3)

Abstract

The science of curiosity is not fully understood, yet it seems to be a key component of nature which drives both humans and animals to seek out new information. Humans actively seek out to solve problems for the sake of solving them, with evidence suggesting that the seeking and obtaining of new knowledge is itself inherently rewarding. This study uses new methods to collect data to investigate how humans react when presented with novel environments and a problem to solve. Information gain was tracked using Shannon’s entropy, a measure of how effective a communication is at communicating its message across. The study investigates if participants feelings of satisfaction will increase the more information they receive, as measured by a change in Shannon’s entropy. A total of 44 participants with complete data were recruited accross two conditions A and B, with A containing a complete knowlege graph to determine what knowledge is gained through interactions with the environment and B containing more uncertainty so that the participant can be observed building their own knowledge-graph. Participants entered a virtual enviroment named Omosa where they were told about a mystery involving the deaths of herbivores on the island. Participants were given free reign to explore and investigate for a minimum 6 minutes. In increments of 90s, participants were asked questions about what they thought was killing the herd and how confident they were of their answer. After 6 minutes final questions were presented collecting player satisfaction and trait curiosity. Additional meta-data including trajectory and interactions were also collected. No significant results were gleaned due to high drop out and incomplete data. Methodology could be altered in future renditions to increase participation and reduce drop out.

Keywords: curiosity, Shannon’s entropy, virtual environment Sammanfattning

Vetenskapen om nyfikenhet förstås inte helt, men det verkar vara en nyckelkomponent i naturen som driver både människor och djur att söka ny information. Människor försöker aktivt lösa problem för att lösa dem, med bevis som tyder på att att söka och få ny kunskap i sig är givande i sig. Denna studie använder nya metoder för att samla in data för att undersöka hur människor reagerar när de presenteras för nya miljöer och ett problem att lösa. Informationsvinster spårades med hjälp av Shannons entropi, ett mått på hur effektiv en kommunikation är för att kommunicera sitt budskap. Studien undersöker om deltagarnas känslor av tillfredsställelse kommer att öka mer information de får, mätt med en förändring i Shannons entropi. Totalt rekryterades 44 deltagare med fullständig data enligt två villkor A och B, där A innehöll en fullständig kunskapsgraf för att bestämma vilken kunskap som erhålls genom interaktioner med miljön och B som innehåller mer osäkerhet så att deltagaren kan observeras bygga sin egen kunskaps grafen. Deltagarna gick in i ett virtuellt miljö med namnet Omosa där de fick höra om ett mysterium som involverade djur av växtätare på ön. Deltagarna fick fri tid att utforska och undersöka i minst 6 minuter. I steg från 90-talet ställdes deltagarna frågor om vad de trodde dödade besättningen och hur säkra de var på svaret. Efter 6 minuter presenterades de sista frågorna för att samla spelarnas nöjdhet och dragkänslighet. Ytterligare metadata inklusive bana och interaktioner samlades också in. Inga signifikanta resultat samlades in på grund av högt bortfall och ofullständig data. Metodik kan ändras i framtida versioner för att öka deltagandet och minska bortfallet.

(4)

Curious Omosa: Does player satisfaction increase the more they learn about their game environment?

Curiosity is considered to be a basic component of human nature and is therefore deeply ingrained into our cognition. It leads us to insatiably seek out and consume

information on a massive scale, so much so that it even goes so far as driving much of our global economy as well as learning and foraging behaviours in animals (Kidd & Hayden, 2015). Despite this, the cognitive, biological and neurological mechanisms underpinning curiosity are still not mapped and well understood (Holm et al., 2019; Kidd & Hayden, 2015). One of the key issues within curiosity research is that there is a lack of universally accepted definition for what exactly curiosity is, and how it phenomenally presents itself. This often results in multitudes of studies which do not use the term curiosity and instead focus on examining individual behaviours that are associated with curiosity such as exploration, play and self-reporting of information desire, amongst others (Kang et al., 2009).

Traditionally, scholars approached the concept of curiosity through the lens of philosophy and religion. The philosopher Cicero for example described curiosity as “an intrinsic motivated desire for information without reward” (Cicero, 1813; Posnock, 1991; Shah et al., 2018).On the other hand, the theologian St. Augustine of Hippo, in his

autobiography “Confessions” describes curiosity as “a passionate and vain longing for

knowledge” often using the term “ocular lust” to emphasise how powerful curiosity can be as a drive in human behaviour (Augustine, 1943; Loewenstein, 1994). The German philosopher and anthropologist Ludwig Feurbach viewed curiosity as being “a hunger which could cause painful feelings of dissatisfaction should it go unsated.” (Loewenstein, 1994). Similarly, the German philosopher Kant also described curiosity as a “hunger for knowledge”. With the advent of psychology, psychologists also began to look at curiosity and its associated

behaviours, with Freud referring to curiosity as a ”thirst for knowledge” (Aronoff, 1962) and James (1983) referring to it as an ”impulse towards better cognition” in that it is a drive to understanding the things unknown to you (Kidd & Hayden, 2015); James (1983) noted that particularly in children there is a drive that guides them towards objects with novel sensory charecteristics such as those with bright colours or those which elicited a startle response. Behaviourists also took an interest in curiosity, especially the range of behaviours that could be associated with it; particularly exploratory behaviours in animals. Pavlov for example in 1927 observed that dogs would actively turn towards unusual stimuli such as sounds, sights or smells, concluding that this was a reflex which he named the ”what-is-it? Reflex”.

(5)

dimension was defined as a curiosity that is inspired through the desire or need for cognitive and/or perceptual stimulation; such as the alieviation of discomfort associated with boredom.

More recent studies on curiosity such as Loewenstein (1994) and Kaplan and Oudeyer (2004) tend to lean towards a definition of curiosity whereby it is a special form of

information-seeking behaviours which seem to be intrinsically motivated. Oudeyer and others (2016) suggest that the satisfaction of curiosity itself tends to be inherently rewarding to Humans, thereby incentivising people to actively seek out new information for its own sake. This stance is backed by research such as Kang et al. (2009), who used trivia questions and the answers to said trivia questions in order to charecterize their participants’ curiosity. Interestingly, they found that curiosity towards the answers to trivia formed an inverted u-shaped curve in relation to the participants confidence about knowing the answer; having little to no curiosity when being completely confident that they do not know the answer, yet becoming increasingly curious when having some idea of an answer but with little to no confidence in being correct. This effect on their curiosity was so effective, that many

participants were driven to the point where they were willing to pay (in the form of forgoing part of their participation incentive) in order to get an answer to these questions then and there, rather than waiting until the end of the session where answers would have been revealed freely.

Just like the definition of curiosity, the function that it fulfills is not properly understood. One of the more popular theories is ”Gap Theory” (Loewenstein, 1994). Lowenstein describes curiosity as ”a cognitive state of deprivation arising from the perception of a potential lack in knowledge and understanding” The theory suggests that curiosity is an intrisic drive which exists and functions in parallel with the other drive-based states such as for example thirst, hunger or the desire to mate. This in turn motivates the organism to drink, eat and procreate respectively, thereby satiating the drive-state and resulting in a drop in motivation to perform the relevent behaviour. Loewenstein (1994) suggests that an initial small piece of information functions to prime the curiosity drive-state thereby greatly increasing the motivation to perform curiosity-satiating tasks such as

information-seeking behaviours. This increase in motivation results in large amounts of information being consumed by the organism. As more information is consumed, the curiosity-state is satiated resulting in a reduction in motivation and therefore a reduction in desire to perform infromation-seeking behaviour.

The vast majority of studies aiming to investigate curiosity assign arbitary values to information in simple decision games or by quantifying curiosity in relation to trivia questions. However, these studies have not been able to specifically measure information seeking behaviours, as there is currently little previous information available regarding information seeking behaviours and how these may be applied to real life situations.

(6)

Wang & Hayden, 2019). For this study, it is hypothesised that, as satiation of curiosity itself can be considered inherently rewarding, it is expected that participants feelings of satisfaction should increase the more that their knowledge is updated through interaction with their environment.

As the use of novel virtual environments in curiosity research is in itself new, the present study aims to examine how humans react when presented with a novel environment and how, when presented with a problem within said environment, they might go about solving it. The current study should provide a valuable testing ground for refining future methodologies involving virtual environments and curiosity, whilst also providing a rich source of novel data that might dictate the direction of any similar future research. In terms of methodology, two formats of the game were produced with the only difference being the structure of the questionnaire presented whilst playing and therefore having differing amounts of information available to the participant. This allowed a between-groups design method, allowing a comparison between group A, which had prior exposure of the

knowledge graph thereby limiting their options of potential causes for the dilemma presented and group B, which was more open-ended and allowed observations to be made regarding to how the participant builds their own knowledge graph and to what extent this might

contribute to the building of a storyline. It is expected that participants exposed to condition B should be more curious due to being more uncertain about their environment when compared to condition A.

Method Participants

(7)

Instruments and Materials Omosa Virtual Environment

The virtual environment of the “Omosa Island” was produced using the Unity video game engine, version 5.6.1f1 (Unity Technologies, 2020) and was designed to run within a web browser on basic, commonly used hardware. The environment is custom built for this experiment and consists of very basic exploration-based, role-playing elements. Two versions were programmed named Omosa A and Omosa B; with the only difference being the questionnaire presented to the participants. The user interface is from a first-person perspective and contains a small character portrait in the right-hand corner. A circular mini map is present in the top left-hand corner. Above is a backpack function, which users can click to investigate objects that they may have picked up. These are accessible in a library like format, but must be picked up from the environment before they are able to be read; with the exception of the welcome note from the Omosan council, which the player starts with by default. To the left of the mini-map is an interactable log-out button, whereby the player can save their progress and take a break. To the right there is a help function, which when clicked gives basic information regarding movement and controls. To the bottom of the mini map there is a map, whereby when clicked by the player, opens a full-size map which can be used to teleport around the island by clicking where they wish to go. Controls consisted of WASD (Up, Left, Down, Right) for moving, the spacebar for jumping and the camera was moved with the mouse. Meta-data such as location and environment interactions were also collected as part of the game, with the data being stored online within a secure database.

The instance-graph of the environment consisted of a landmass, which was surrounded by water. The land mass was mostly desert and grasslands, with patches of trees both alive and desolate. Scattered upon these land masses were herds of Yernt; a deer-like herbivore fabricated to be part of the island’s plotline. There were two main locations that the participants were encouraged to explore, those being the ‘Research Lab’ and the ‘Village’; however, the participants were not subjected to any restrictions and could explore the whole island freely. The Research Lab area consisted of various stationary environmental object such as desks, computers and chairs and other instruments typically expected to be in a laboratory. One interactable resident or Non-Playable Character (NPC) called Charlie the Ecologist was present in this area, with two additional interactable objects the computer and a collectable Virologists Notes. On the other hand, the Village area consisted of various tents pitched in an omega-shaped formation, with a fire in the center. These tents contained primarily non-interactable items such as bows or crockery. There were three interactable NPCs present within the village prior. Lyina the farmer at the entrance to the village. Pedro the microbiologist sat at a table to the right of the center of the village and Zafirah the nutritionist at the back end of the village. There were also three interactable objects, these were the Ecologist’s notes (available from the nutritionist’s desk), Microbiologist’s notes (available from the microbiologist’s desk) and the Fauna notes (available from the tent behind the nutritionist).

(8)

player’s inventory by default, welcoming visitors and also setting the scene. Charlie’s computer is interactable with three spreadsheet files available for view however these are only numbers supporting that the Yernt population is falling. Behind Charlie, there is the Virologist’s notes which explain that there has been an unknown virus found in the Yernt herd that seems to be related to another intestinal virus found in other animals. It goes on to explain that it is prevalent within the herd but does not seem to be causing an increase in mortality. Within the village, the player encounters the farmer Lyina who mentions that they have seen more dead Yernt than usual and that they have seen predators attacking them, though this is not supported by the scientists. To the right hand-side of the village they encounter Pedro the Microbiologist who suggests that it may be bacterial contamination in the water, they point the player to their notes on the table and also reiterate that a farmer has suggested predation to be the cause. The microbiologist’s notes suggest that there is bacterial contamination in the water that would kill most animals but that the Yernt herd seem to avoid it; this is also accompanied by population data in the form of a graph and tabulated data. Near the north side of village, the player encounters Zafirah the nutritionist, who lets the player know that there has been a drought, which in turn has reduced the number of plants available on the island, including a particular bean containing vital nutrients that the Yernt need, resulting in a reduction in the population growth this year and if the problem is not rectified then there will be likely widespread malnutrition. Zafirah points the player to the Ecologist’s notes placed on her desk. These support Zafirah’s observations regarding the reduction in plant diversity due to drought and that it is likely responsible for causing malnourishment within the Yernt herd, causing them to die and reduce in number. The final interactable document available to the player within the Omosa Village is the book on local Omosan Fauna. This contains observations regarding two species of animal that live on Omosa, including the Yernt, as well as the “Tooru”, a leopard-like predator that hunts and eats the Yernt.

Welcome letter

The questionnaires used were produced in and hosted by Qualtrics (2020). A welcome letter was produced in html to inform the participants regarding consent and ethical considerations.

Demographics Questionnaire

The demographic questionnaire collected data regarding the participant’s age, gender, how many hours they play videogames per week, nationality, and the highest level of education they had completed.

Internal Questionnaire

For both Omosa A and B an internal questionnaire was presented to participants every 90 seconds, which they would have to complete before continuing playing There was a timer present at the top of the web page counting down from 6 minutes, which was the minimum amount of time required for the participant to play the game; after expiring, it would then be possible for the participant to click a hyperlink taking them to the concluding questionnaire. These internal questionnaires were coded into the game and were hooked into an online database using Qualtrics (2020) technology.

(9)
(10)

Concluding Questionnaire

The concluding questionnaire collected data regarding the participants enjoyment of the experiment, how curious the participant was to learn of the cause of the herbivore deaths and final confidence ratings. These were presented as radio buttons ranging from 1 through 7 with 1 being Strongly Disagree and 7 being Strongly Agree. Trait curiosity was measured using the Curiosity and Exploration Inventory II (CE2) (Kashdan et al., 2009).

Procedure

(11)

Results

Data was analyzed using a custom written program within MATLAB (The Mathworks Inc., 2020). CSV files containing the downloaded data were read by the program and reorganized according to their subject identifier (ID). Object names and NPC names within the data were renamed for convenience. Confidence rating values for participants that did not answer subsequent requests for confidence ratings were recoded so that said subsequent answers were the same as previous answers; this was implemented as entropy was defaulting to maximum in these cases. Database entries without user-IDs were purged, as these could not be traced back to other data points along with irrelevant and nonsensical data. Players who failed to complete >34% of the survey were also purged from the database as this indicated that they did not play the game. The time entries were then recomputed to standardize the recorded game time across all participants and information sources; this was as time zone influenced the times recorded and made them unable to be compared. Time taken within game was also influenced by loading speed, with some individuals loading quicker or slower depending on their internet speeds. In order to make participant timings comparable, all times were recomputed to begin from the moment that the person spoke to Charlie the Ecologist as the game is constrained to immediately talk to this NPC upon loading and is therefore a valid starting point for all participants.

(12)

Figure 3

Pearson’s Correlation Coefficient Matrix

Note. N= 36. Shows Pearson’s R correlation statistics for combined Omosa A and Omosa B variables

Between group T-tests were performed to discern whether participants in the less structured version of the game (Version B) were more curious and therefore more satisfied with it. Satisfaction ratings were compared (M = 2.37 (B), 2.54 (A), t (36) = -0.63 p = .53) There was no significant difference found between the two versions. Task curiosity ratings were compared (M = 3.83 (B), 5.00 (A), t (35) = -1.87 p = .07). A significant difference was found, with participants in condition B displaying greater task curiosity than in condition A. Game lengths were compared (M = 7.95 (B), 10.0 (A), t (34) = -0.78 p = .44). No significant difference in game length was observed between game versions. Final confidence ratings were compared (M = 2.84 (B), 2.92 (A), t (36) = -0.13 p = .89). There was no significant difference observed between the two conditions.

(13)

Figure 4

Game Length vs. Ratings of Satisfaction, Curiosity and Final Confidence

(14)

Information gain was characterized by calculating the change in Shannon’s entropy over time and calculating which objects produce the greatest changes in entropy. A histogram was produced (Figure 5) showing how much information (entropy change) was conveyed when interacting with objects and NPCs. The histogram shows that interacting with the microbiologist Pedro caused the greatest change in entropy (0.65) when spoken to. Likewise, reading the Virologists notes caused a noticeable change in entropy (0.85)

Figure 5

Histogram showing information gain according to NPC/Object interacted with in Omosa A

Note. The Histogram shows information gain as the total change in entropy after interaction with said NPC or object using data from Omosa A. Objects or NPC with larger information gains through total change in entropy are thought to be the most informative.

Data regarding where the participant went geographically within the game was then analyzed. Participants who played the game had their coordinates monitored for the duration of their time within the Omosa environment in the form of X (Latitude) Y (Height) and Z (Longitude). X and Y coordinates were taken for participants and were visualized graphically (Figure 6.) for both Omosa A and Omosa B conditions. As there was clear overlap, with many participants spending time in the same places, a heat map was produced so that

(15)

Figure 6

Scatter graphs showing X and Y coordinates for participants in Omosa A and Omosa B.

Note. Scatter graphs show participant’s spatial distribution in Omosa A (blue) and Omosa B (red). Small green circles in Omosa A represent the coordinates of where the NPCs were present in the game. A third graph, a heatmap shows the concentration of participants who visited area clusters when data for Omosa A and Omosa B are superimposed over one another.

Discussion

The main purpose of performing this study was in order to explore a new

methodology for obtaining real-time curiosity data, whilst potentially collecting information that can be used to determine how humans behave when presented with a novel environment, with a potentially curiosity inducing mystery to be solved at their discretion. Unfortunately, due to the low participant numbers and the high drop-out rates, the data is not sufficient, and it is therefore not possible to make any sweeping generalizations or conclusions.

(16)

them, more important tasks, tasked on satiating their curiosity drive primed by other sources. Another interesting correlation is the weak positive correlation observed between game length and task curiosity. This suggests that that as game length increases, so does the participant’s curiosity towards the task at hand. It could be concluded that this was due to the person’s natural curiosity being primed by the task, leading them to seek and satiate said curiosity by trying to solve the task. Another potential explanation however that is just as likely, is that the participant has committed to playing the game until its finish, and that being curious about the outcomes of the task is just a by-product of playing the game.

The hypothesis for the present study, was that participants feelings of satisfaction should increase the more that their knowledge is updated, and curiosity satiated through interaction with their environment. There is a weak positive relationship between task curiosity and game satisfaction that may back up this hypothesis. This correlation suggests that the participant is more satisfied the more curious they are about the experimental task. This is as expected and is somewhat supported by previous studies, which suggest that curiosity is itself inherently rewarding. However, as previously mentioned the low participant numbers and therefore lack of data and statistical power means that the null hypothesis that there is no difference between participant satisfaction and curiosity must be accepted until proven otherwise.

When game length in the form of number of prompts is visualized graphically (Figure 4.), there are additional relationships which can be observed. Firstly, when game length is plotted against satisfaction ratings, it seems that the longer that participants play the game, the more satisfied they are with the outcome of the experiment. This could be because their curiosity has not been satisfied sufficiently to satiate them and provide the pleasure that satiation would provide. Alternatively, it could be that the stimulation from the game is insufficient and that as participants play, they may become bored and therefore dissatisfied with the outcome of the game due to lack of commitment. When plotted against task curiosity, there is a weak positive relationship with game length. This suggests that participants may being primed into becoming curious the more information they get as playing the game and may be indicative that the experiment is sufficient in providing the small amount of information suggested by Loewenstein (1994) to prime a curiosity-state. Finally, when game time and confidence rating is plotted there is also a weak positive relationship present. This may be indicative that there is sufficient information gain being communicated through the various NPCs and objects within the game, as the longer that the game is played, the more confident the player becomes in what might be the cause.

In terms of information gains, (Figure 5.) the information communicated through the different NPCs and objects within the game are more than sufficient for providing the participant with some idea to what might be the cause of the Yernt herd deaths. The pieces of information that were designed to eliminate certain causes, for example both the virologist’s notes and the ecologist’s (Charlie’s) notes eliminated viruses and bacteria as being potential causes and therefore functioned as designed. This also was backed up by Pedro’s character dialogue which communicates the same information regarding bacterial infection not being the cause of the herd’s demise.

(17)

heatmap, which combined location data from both Omosa A and Omosa B. This is consistent with the type-graph of the Omosa island, which only had two locations containing objects and NPCs which the participant were nudged towards exploring so that they might solve their task. However, in some instances players did not immediately flock to these areas. Many seem to opt into exploration strategies, often exploring areas of the island which were not suggested yet were still made available to them. In addition, the warping mechanic which allowed a type of fast travel by clicking on sections of the map also allowed players to explore areas such as the sea which the programmer had not intended. It seemed that perhaps there were three or more strategies that might be employed by participants when presented with a novel environment. In the first strategy, the participant reads through the dialogue and systematically completes the task at hand by searching for information in the research area and village area as suggested. Upon completion through finding out this information, they simply exit the game and complete the questionnaire. In the second strategy, the participant will skim over the dialogue or skip it altogether and head straight out into the open world either running on foot or through teleportation. They will then run around for some time, whilst seemingly exploring their new environment, these players will then quit the game and complete the questionnaire having not actually completed the task at hand. The third strategy combines aspects of the two previously mentioned. The player will both explore, seemingly at random, but also pursue the task through reading objects and talking to NPCs.

In terms of comparing Omosa A and Omosa B to determine which evoked greater curiosity in participants, Omosa A was more structured, containing sliding bars representing each of the potential causes for the Yernt herd dying; thereby showing what information is picked up by the participant at the cost of disclosing the knowledge graph. Omosa B on the other hand asked for specific names of NPCs, which resulting in an increase in uncertainty experienced by the participants as unless they have played through the game and have interacted with NPCs and objects, they would not know what potential causes were available as options, which in turn allows us to make observations as they build their own knowledge graph. The between-group T-test for task curiosity ratings between Omosa A and Omosa B was significantly different, suggesting that participants were more curious in condition A than condition B. As a result of both conditions producing similar results, it may be considered valid to use findings from A to interpret the states and behavior of group B. This can be considered a strength of this study and its results, should larger and new data collections be performed with richer game contents.

(18)

and due to time and money constraints could not be updated to newer versions of unity that may have been able to support mobile platforms. It can be concluded therefore that the Unity not working, or not functioning as intended for many participants was responsble for the low turn out numbers. Not only this the added hassle of having to download an additonal web browser in order to participate in an experiment with no renumeration may have been too much for some people’s patience.

Advertisement of the study was performed mostly through the use of social media. Present day social media apps tend to push through notification such as messages or posts from friends in a real-time basis and , due to their portability and ability to pick up telephone based internet they are perfect for carrying on one’s person. It could be posited therefore that a good majority of potential participants would recieve the invite to this study in real-time as it is sent and either attempted to participate there and then, only to conclude that they were unable to on their telephone and therefore giving up on the experiment and moving on. Or, alternatively they could have made a mental note to log back in to their social media on a computer at a later date, only to then forget. There was also no renumeration available and this could have influenced some participants to not take part as it was somewhat time consuming, although no more than about 15 minutes of their time.

(19)

References

Aronoff, J. (1962). Freud's conception of the origin of curiosity. The Journal of Psychology, 54(1), 39-45.

Augustine, S. (1943). 398 AD. Confessions. Trans. FJ Sheed. New York: Sheed and Ward. Berlyne, D. E. (1954). Knowledge and stimulus-response psychology. Psychological Review,

61(4), 245-254. https://doi.org/10.1037/h0055729

Berlyne, D. E. (1966). Curiosity and Exploration. Science, 153(3731), 25-33. https://doi.org/10.1126/science.153.3731.25

Cicero, M. T. (1813). De finibus bonorum et malorum (Vol. 3). Weidmann.

Golman, R., & Loewenstein, G. (2015). Curiosity, information gaps, and the utility of

knowledge. Information Gaps, and the Utility of Knowledge (April 16, 2015), 96-135. Holm, L., Wadenholt, G., & Schrater, P. (2019, 2019/08/02). Episodic curiosity for avoiding

asteroids: Per-trial information gain for choice outcomes drive information seeking. Scientific Reports, 9(1), 11265. https://doi.org/10.1038/s41598-019-47671-x James, W. (1983). Talks to teachers on psychology and to students on some of life's ideals

(Vol. 12). Harvard University Press.

Kang, M. J., Hsu, M., Krajbich, I. M., Loewenstein, G., McClure, S. M., Wang, J. T.-y., & Camerer, C. F. (2009). The wick in the candle of learning: Epistemic curiosity activates reward circuitry and enhances memory. Psychological science, 20(8), 963-973.

Kaplan, F., & Oudeyer, P.-Y. (2004). Maximizing learning progress: an internal reward system for development. In Embodied artificial intelligence (pp. 259-270). Springer. Kashdan, T. B., Gallagher, M. W., Silvia, P. J., Winterstein, B. P., Breen, W. E., Terhar, D.,

& Steger, M. F. (2009). The curiosity and exploration inventory-II: Development, factor structure, and psychometrics. Journal of research in personality, 43(6), 987-998.

Kidd, C., & Hayden, B. Y. (2015). The Psychology and Neuroscience of Curiosity. Neuron, 88(3), 449-460. https://doi.org/10.1016/j.neuron.2015.09.010

Loewenstein, G. (1994). The psychology of curiosity: A review and reinterpretation. Psychological bulletin, 116(1), 75.

(20)

Posnock, R. (1991). The trial of curiosity: Henry James, William James, and the challenge of modernity. Oxford University Press on Demand.

Qualtrics. (2020). Leading Experience Management & Survey Software. Retrieved 17/05/2020 from https://www.qualtrics.com

Shah, P. E., Weeks, H. M., Richards, B., & Kaciroti, N. (2018, 2018/09/01). Early childhood curiosity and kindergarten reading and math academic achievement. Pediatric

Research, 84(3), 380-386. https://doi.org/10.1038/s41390-018-0039-3

Shannon, C. E. (1948). A mathematical theory of communication. The Bell system technical journal, 27(3), 379-423.

The Mathworks Inc. (2020). MATLAB version 9.8.0.1359463 (R2020a) Update 1. In https://www.mathworks.com/

Unity Technologies. (2020). Unity Real-Time Development Platform | 3D, 2D VR & AR Visualizations. Retrieved 17/05/2020 from https://unity.com/

References

Related documents

[r]

Instead of visiting a physical location, the user could place a map that looks like a vinyl record on the device and starts to play stories based on the location of the needle..

I allmänna råd för förskolan (Skolverket, 2005) finns beskrivet hur man kan arbeta med dokumentation för att främja barns utveckling ”ett sätt att öka medvetenheten om det egna

R2 : - Ja vi pratade ju med Namn från MRS och hon påpekade också det att det vore bra att ha något, för de får också väldigt tunga modeller när de ska göra kataloger och

According to the socio-educational motivation theory, instrumental and integrative motivation can be equally compelling, which this study confirms (Gardner 1985:55).

The aim of this study is twofold: firstly, to analyze students’ texts to see if the students, with the help of scaffolding, can develop their writing skill and get higher

50 Swedish elites compiled these ballads in visböcker (“songbooks”), many of which provide source material for Sveriges medeltida ballader. 51 For Sweden, interest in recording

Trader Agent: This agent is responsible for making compelling trade offers to other players in exchange of resources needed by the player to build settlements that could give