RoRk at Work

Une amie vient de me transmettre de la pub pour un atelier/conférence intitulé “les ateliers des médias 2007″. Il s’agit de quelques invités de la presse et des médias (le guest semble être Karl Zéro) qui parleront de buzzwords du genre “marketing viral”, “blog” (c’est mort  comme buzzword les gars), “second life”, “e-paper”… Tout ça durant le salon du livre à Genève le mois prochain.

Bref, mon but n’est pas de faire de la pub. Ce qui m’a choqué au point de me faire bloguer (c’est dire…) c’est que malgré le thème de l’atelier et le discours qu’on recherche… Il n’y a pas de page web, pas de blog, rien… Si par hasard on voulait passer l’info plus loin, on se retrouve à forwarder le .pdf!

J’ai trouvé cela splendide, et très révélateur du niveau de compréhension de ces “buzzwords” par les grosses boîtes de comm et autre qui sont impliquées dans ce projet.

Bien entendu, je met le pdf à disposition (au moins comme ça il sera dispo sur le web tiens).

I just played with Alexa which seems to be another tool to know about trafic ranking of websites. And I tried for my university’s website (view complete stats). I was shocked when I saw this:

Where do people go on unige.ch?

• unige.ch - 34%
• tecfa.unige.ch - 23%
• cui.unige.ch - 4%
• wadme.unige.ch - 4%

Come on! 23 % of webtrafic on the university servers go to tecfa? This means something like 23 mio hits every day? No wonder if we are high on google, or maybe is it only because we are high on google…
If that’s accurate, we definitely should think about advertising, this could pay our coffee or even another research associate

P.S. K, I’m joking okay??

Wii usa features several movies of several populations of users playing with the wii. It’s of course a marketing move and I hope these people are not paid actors, because I liked it.

In fact I was wondering whether or not I should record faces from my players in my next experiment. I wondered if the information would be rich enough and usable in a kind of “game engagement” assessment. It seems that the wii triggers facial expressions, but could I use these in a detailed analysis?

I like to compare the kind of expressions of players on their own or in couple/in a group. The social settings completely changes what’s expressed, this is very obvious. But I’m happy to see that even alone, things are expressed: mostly surprise. I hope nintendo did not cut in thousands of footage hours to get these.

Clickable culture reports about a new trend in Wow: corpse graffiti. the idea is to create a new avatar, to name it with a strange (but carefully chosen) name, and to have it die in some very specific (and carefully chosen) place in the world. The corpse, and therefore the name, will stay there for a long time, allowing everyone to read the message.

A wow forum thread reports some of the best examples : corpses named “welcome” or “please wipe your feet” in front of a city’s entrance, another named “trout” in a river or “firesucks” on top of a torch. But you also can find references: lots of gnome corpses named around “oompaloompa”, a floating “jeff buckley” and so on.

it made me smile, since geolocated messages are a research trend right now. I played with stamps in Geneva, but here it’s different, really more like a graffiti: you never asked for this information, but the message in the place makes it relevant and you are happy to have seen it. And that’s because you were in the place also, located with the message.
Another creative trend in Wow. Boy, players surely are bored with the game, but they keep playing!

Long time no see. I have no excuse, if you need one go there and choose. I don’t think I’m a blogger, I just happen to write on a blog…
So, I’m working for tecfa since 3 years now. Let’s take some time to look back and then look forward…

The cleap project : yeah that’s how I started here (I was SO young!), a FNS grant to do nothing but research on a topic which seemed fun when I began. 3 years later, after 4 experiments and almost 300 participants, the project is finished (for 6 month now). We spoke about it in 7 international conferences (I did 4 of them) + 2 “real” posters. 2 articles are waiting for approval (for almost 1 year concerning the one I wrote, thank you for this -ever delayed- special issue proposal), and I need to write the last one. This project is the start of my PhD and my first real experience of research (theory, experimental verification and iterate !). I also managed to obtain 2 Masters by presenting parts of this research.

Good: I learned a lot, experimental methods of course, but also the job of a researcher (read, do, write, talk). Tecfa is a great environment for a first job in research, I also made myself known a little bit in the european educational multimedia research world.
Not so good: I’m still tied to this research. It has to be part of my Phd and I want to do something else now, I have to find a link between the two and this restricts me. I still have a journal paper to write about the two last experiments.

Mediaplus: The last 6 months I have been paid by the university to prepare a course for all first year in psychology and educational sciences (around 400 people) in order to have eneryone able to efficiently use computers. I’m done with this next week, but I still have to write a report.
Good: I did some research, interviews-based, It was nice to change paradigm. I also had to dig data from the program used last year, I had to deal with the logfiles I had (and could not change) in order to show things, that’s odd for an experimentalist working with carefully designed variables and measures. The best part is that it was applied, so I had to figure out ways to make things work this year.

Not so good: We had to do with one platform (that is crappy imo), to teach microsoft-only content (that is limitative and ethically questionable imo). Also decisions took long to be taken and everything changed the week before the start of the academic year. I also think i could have improved the project by being more proactive and more upseting people’s routines… I have that problem that I stay too much at my desk.

Teaching assistant: Since six months but fulltime from now on. This is rather an easy task, I don’t have to give the lessons myself (most of the time), I know what we are talking about (user-centered design and methodology), and it’s nice to make student’s projects work.

Good: I also like to be the “assistant” and to make things work. Yesterday I followed Charline Poirier’s user-centered design lessons and I remembered how much I liked this course when I followed it 4 years ago. I think I should do UCD/ethnographic research again…

Not so good: Maybe this job isn’t demanding enough, I am afraid to feel secure (even if i can’t stay more than 2 years), and start procrastinating (I mean, more than I already do).

NeglectCity :I have a small project with two friends in neuropsychology. They work with patients with hemispatial neglect and I used a video game to quickly mod a detailled environnment and create a small movie that could be used in the diagnostic of this condition. We now are validating the procedure (well, they are).

I’m working on an english proficiency certification (even if you can’t see it with my typos and stuff on this blog), and wanted to go one year in Oxford with a scholarship but its regulation excludes visiting PhD’s.

SO…
Where’s the problem? I have a job, a well-started PhD and side-projects to keep me awake…

Well, the problem comes from what I would like to do with my life! I don’t see myself in educational multimedia for a career, I like it allright and appreciate very much the people I met in this small world but I want to work in games when I’m grown up!

I like research, users and games and I want to do something mixing all these. Selling services like the ones xeodesign seem to provide could be really great: providing game companies several research methodologies from ethnology and psychology in order to help them design great games. This sounds nice! Being an academic is nice also, and nothing prevents me from doing both. However I have things to do in order to be an efficient research provider for the game industry…

1. Experience a wide range of research paradigms and methodologies: I’m quite good with cognitive/experimental designs, I have experience of interviews, user-based design, and usability testing. I guess it’s a good start but if I want to be a reference in methods, i need to know more and to use all of them!
2. Get involved with game companies; this definitely is my problem, I know nobody. It’s nice to follow what’s happening but I have only a small idea of what designing a game involves. What are the questions in the industry today? I totally missed this aspect until now…
3. Write a PhD: I still have two years to write my PhD, I think of this as a chance since I have two years to realize the two previous points while being paid by the university for it… I can get involved in projects and help industries without asking any money (well, of course i can’t do it fulltime).

So, My PhD should be:

• Academically irreproachable (of course)
• About video games
• Applicable/of interest for the industry
• Using several methodologies
• In english
• Linked with the 4 experiments I already have !

For sure, this should be an old post…

I already have a mixed-paradigm experiment in my mind to be realised before christmas, I’ll certainly blog about it (althrough i’m not a blogger).

References: I write this out of several small papers from Nicole Lazzaro.

Lazzaro, N. (2004). Why we play games: Four keys to more emotion without story. pdf, Online 5 july 2006.

Lazzaro, N. (2005). Why we play games together: the people factor. Presented at the Game Developer Conference Europe, 31 august - 1 sept 2005. London, UK. pdf, Online 5 July 2006. Webcast (registration): http://www.pqhp.com/cmp/gdctv/

Lazzaro, N. & Keeker, K. (2004). What’s my method? a game show on games. Paper presented at the CHI 2004 conference, April 24-29. Vienna, Austria. (ACM press DOI).

I can’t find articles presenting the researches in details, only these papers stating that lots of researches have been caried out (xeodesign is a company, not an academic lab). So I don’t know exactly how they got all these results. The first paper talks about 45 hours of video on 30 adults playing their favourite game and on 15 friends and family members nearby. Questionnaires and verbal/non-verbal cues were also coded. But I cite: “It is beyond the methodology and scope of this study to produce statistically significant results; …” (Lazzaro, 2004, p. 2). So, how should we interpret the model presented? I really would like to know more about their methodology…

Some points, the first 2 papers about how to unlock emotions in games:

• There are 4 keys to unlock emotion into a game: Hard fun, Easy fun, Altered states, The people factor. These are not categories but elements to describe the potential enjoyment to get from a game. Combining these factors makes the game more enjoyable.
• Hard fun: emotions coming from meaningful challenges, mastering the game. It is the poursuit of a goal through the strategic use of skills, the reward is acomplishment. Emotions involved are frustration and fiero (personal triumph). The balance of game difficulty is crucial for this kind of play.
• Easy fun: passing time, grab and maintain attention. The player’s curiosity is awaken and nourished through immersion in the game, the goal is to fill attention with something new. Exploration, adventure, story and narrative are core elements.
• Altered states: perceptive, behavioural and cognitive elements to create emotions. The pleasure of felling internal changes through play. The social context can be very important for this kind of experience. Emotions involved are mostly excitement and relief. Avoiding boredom is already part of this category. The richness of graphics, audio and behaviors involved in the game, the richness of interactions, are necessary elements to let player enjoy this kind of experience.
• People factor: Opportunities for competition, cooperation and spectacle. Some players play games they don’t like just to be with their friends. Amusement, Schadenfreude and Naches (a kind of parental pride, read Ekman) are the main emotions involved. Such games are tools for social interaction. Of course multiplayer games are the first to use this key but not only (the mere presence of high scores can make a game get this key). Games fun to watch also have this key (I remember my roommate watching me play Baldur’s Gate and half-life for hours, don’t ask me why but he’s a cop now…).
• In the second paper (for GDCE) the author went further to define the “people factor” key. Seven ways to create more emotion from this key: 1. Support player to player interaction (opportunities to interact). 2. Put on a spectacle (make it fun to watch). 3. Tools to communicate emotion (see the other’s, share yours). 4. Emotional non-player characters (all npc should be able react to player’s decisions). 5. Emotionally expressive tools and objects (more possible emotional exchange, objects mean something when they are used). 6.Emotion cycles, feedback, chains (I don’t really get this one). 7. Save money (test the product early and often, know the users).
• Emotions involved in play: the authors claim to have observed many types of emotions from their videos during game play (facial gestures, body language and verbal comments): Fear, surprise, disgust, naches/kvell, fiero, schadenfreude and wonder. Themes and triggers for each of them are listed in the paper.
• They observed a big contrast in terms of quantity and intensity of emotions between gamers playing alone or with other players. Behaviours, rituals and emotions are added in multiplayer condition that enrich the play and the emotions.

This result has to be linked with the results of Mandryk, Inkpen, & Calvert (2006), in their second experiment, they observed very different emotional responses (in terms of physiological arousal measures) when players were playing alone or against a friend. I reported about this article two months ago.

• Again, the authors cite the concept of “flow” as very apt to describe what happens when playing a game. So thanks again Mihaly.

The CHI paper about games and productivity softwares

• The authors underline the importance of more methods to analyze game characteristics. Enjoyment and fun of the player is definitely a good indicator (more than market surveys and product sales). And this kind of enjoyment research could also be applied to productivity software.

…and also educational media, since it will be my missing link for my PhD! My past research on instructional design and the use of animated pictures for learning needs something to be attached to my present study on games, enjoyment and emotional involvement.

• Games and productivity software have different goals, entertainment vs task completion, intrinsic vs external reward, etc. In particular the satisfaction issued of the use of a productivity software is different from “having fun”, very different! 100% of success is mandatory for productivity tasks but makes a game boring.
• Traditional usability improved the players’ experience on games. But other aspects are beyond the scope of these methods.
• Games must be usable: efficiency, effectiveness, satisfaction… these kind of methods can apply to games
• Games must be challenging: (this goes agains usability since it makes things harder), what are the methods to improve such dimensions as challenge, strategy and problem solving?
• Games must be fun: how can we measure this? accomplishment of course, but also aestethics, humor…
• Qualitative methods are vital, biometric methods (psychophysiological) can be very helpful empirical measures, and qualitative observations need to close the methodological triangle.

Well, I would have loved to be there in 2004 to listen to this talk since such methods are exactly the ones I am trying to put together in my research! I am wondering which scales they are using exactly and to what point these different approaches can give the same information. How they can be used together, in interaction. I also need to define my first factors! Why didn’t these dudes write journal papers??

Reference: Sweetser, P. & Wyeth, P. (2005). GameFlow: A Model for Evaluating Player Enjoyment in Games. ACM Computers in Entertainment, 3(3), 1-24.

Main points:

• Most of literature about designing and evaluating games is centered on interface, mechanics and gameplay issues. However, there is no integrated model/method to evaluate player enjoyment in games.

I like this idea of enjoyment as the “outcome” of a game, if we really can make it measurable, we will have a great dependent variable for lots of experiments…

• Literature about media enjoyment developed several non-integrated models: attitude, parasocial interaction, transportation theory, disposition and cognitive evaluation. The flow theory of optimal experience is used here since it has the more broadest approach.
• Activities succeptible to create a flow experience are goal-directed, bounded by rules, require mental energy and appropriate skills.
• Flow is an end in itself, it is an autotelic activity, people will continue the activity for its own intrinsically rewarding sake. This is particularly true when we think of games.

This could explain why educational games are always bad games! since players have to perform them for something else than their own sake… So, educational games are not games? or will never be enjoyable games??

• From a literature review on usability and user experience in games, the authors builded a model of enjoyment in games (the GameFlow model). With eight core elements: concentration, challenge, skills, control, clear goals, feedback, immersion, and social. Authors discuss the mapping of their elements to the ones of the flow theory (it is close).
• Concentration: needs in terms of workload and attentional ressources. more concentration needed means more absorbtion. So a good game requires a lot of attention, no energy is left to the player to process any other information. There are some exceptions to this rule (stimulis must be game related and respect limited cognitive capacities, etc.).
• Challenge: very important for game design, the small path between frustration/anxiety and boredom/apathy. Satisfaction comes from accomplishing difficult tasks (in a rather large meaning). Challenge has an intrinsic reward (in a way if you don’t want to “succeed”, there is no challenge anyway). The difficulty is to meet the level of challenge of every player, concerning this, pace is an important aspect.

This accomodation between anxiety and apathy is very often cited in game design writings but I don’t know any work really measuring it with an experimental setting (with scales of course, but also physiological measures and “time to decide to stop playing”).

Concerning the level of difficulty, I would love to see a study comparing instructional design and game design researches concerning this cautious increase. Are there similar influences on learning and enjoyment?

• Player skills: Their development and mastery has to be supported in a game. Perceived player skills must match the game’s challenge (beyond a minimum threshold). Learning to play the game is important and can be done in several ways (tutorials, ingame tips, on demand or not, conventions of the genre…), recommendations are discussed.
• Control: the possibility to translate the player’s intention into in-game behavior. Controls should be simple and minimal, but expandable/customisable and allow different play styles. Mistakes Control over the game shell should also be possible (I would call this, play control): quit at any time, save, etc. The possibility to make mistakes should be avoided in games (that is very different from educational materials!). Last, the level of control is also the possibility to play the game you want (closer to a simulation), more freedom, agency, less feeling of linearity and “triggered click zones”. This is supposed to be better and more replayable.
• Clear goals: this is a prerequisite directly imported from the flow theory. Goals have to be present but also very clear to the player. Multiple goals are also often used.
• Feedback: immediate feedbacks are needed in order to achieve flow, so they have to be as frequent as possible. The authors cite several references about the need to know the distance and direction to the goal at any time.

On a personal note I hate games that give you total immediate feedbacks, it is so much in opposition with the need of agency developed in the “control” element. When you always know what to do, when and how long it will take you completely loose the feeling of control over your gameplay and just follow the arrows. The puzzle part disapears very quickly with too much awareness of goals and goal states (but maybe I’m a freak).

• Immersion: deep and effortless immersion is expected. Immersion, engagement and absorption are at the core of positive game experience. Player become less aware of their surroundings. Emotional investment is also high, due to the time, effort and attention devoted. To achieve this, several ways are presented, amon them, personal involvement (emotionaly and visceraly), forget the medium, linger, multimedia, narratives…

Ok, now what is the difference between immersion, engagement and absorption? (I think I could guess something but I need something more consistent). But isn’t immersion the flow state we are looking for? It seems to me that all the other elements presented tend to create the best conditions for immersion. We are not in a component model here, and the relations between the elements is very unclear!

• Social interaction: not an element of flow, even a possible source of interruption for immersion. Not a property of the task, but the task allows social interaction. Interacting with other people is described as enjoyable and to be supported in games. The virtual community created by a game attracts a lot of players.

Reading this paper, I’m amazed of the number of recommendations for effective game design. However, I see very few global regrouping and general rules. Maybe this is still missing, like a model taking most of these studies into account. This is needed, especially since i feel a lot of recoupment between the categories in this paper. For example, the player skills are a pre-requisite for challenge building, which is very close to achieve a good concentration, and this all is necessary for immersion maybe…

Validation of the criteria:

Two similar games (Warcraft 3 and Lord of Everquest, two RTS games of 2002-03) were evaluated by two experts (nothing is said about them). W3 is a huge success and was highly evaluated by professional reviewers, Lords of EQ is the contrary. The two games were assessed on the 8 criteria of the GameFlow model through a 5 points scale. The complete scales and justifications are most of the paper.

In the end Warcraft 3 outperformed Lords of Everquest (96% to 48%). Since this reflect the sales and professional reviews, the authors claim this validates their criteria…

Well, I’m far from convinced on this… The literature review was good on the game design issues. However, this result certainly is not enough to validate this scale! Maybe it measures something close to what experts look at when reviewing a game, but with only 2 experts here and games that were so different in success… I don’t think this scale is fine enough to differentiate numerous games. Moreover, it would take so much time to apply it, only to get the information a gamer would give in a split second…

Moreover we know nothing about the flow of the players… These scales were compared to ratings, not to user experience. Who said the only quality of a game was to induce flow?

The literature review was interesting but I would like to see a model of how these eight dimensions co-define the flow in games. Then maybe we can find ways to experimentaly verify each part of the model. The outcome in terms of “pleasure of the player” needs to be assessed, flow would definitely be a dimension (but how to measure it?), but what would the other dimensions be?

Allright, maybe some of you remember that last year I followed the courses for a second master (in cognitive and experimental psychology). Maybe you also remember my disapointed post in february when, after a huge week of work, I realised it was too late to have the master thesis accepted for this exam session.

So, now is the next session, I organized myself a little bit better (I just read the diploma’s regulation before). So I can announce that the final work I delivered to my jury is also available online as a pdf. It’s not very long (50p) so maybe people will read it .

But the funny story is that I also present this experiment next week in an international conference of the earli, in Leuven. I didn’t do it voluntarily since I initially wanted to finish my master thesis back in march. However I am very interested in the critics I will receive from both audiences. What if I get a very bad review in Leuven and a very good grade for my master? What if it is the contrary? (that’s even more funny).

Now it made me make a parallel with a well known art piece: the fountain of Marcel Duchamp. This work and all the ready-made (found art) movement questionned what an art piece is by elevating to this rank the most common objects (like an urinal for Duschamp). It is the power of the artist to decide what is art and what is not. In a way, art is what is exposed in museums or institutions. Like for the old notion that “intelligence is what intelligence tests measure” (attributed to a colleague of Binet), art is what artists do/what art institutions show.

So what about a scientific work? In a way, since the work was accepted as a conference paper it is scientific work! The jury for my master thesis can only accept it, else they would enter in opposition with the scientific board of the conference…

Ok, let’s stay scientific anyway: 1) the two works are from the same experience but are not the same papers; 2) the jury is there to help me improve the work, not to accept/refuse it; 3) these two evaluations have very different objectives, to compare them like that is a little bit like convincing someone using a powerpoint presentation…

Still, I like the strange coincidence.

Else, I hope I can blog again soon, two weeks of conferences ahead and a master thesis presentation (3 speeches in 2 weeks, it’s damn new for me).

[Edit: nothing funny happened, both the conference and my dissertation’s presentations went well…]

While reading the paper I just blogged about (Mandryk et al., 2006). An experiment idea came to my head, I write it here in order to start refining it.
The thought came from their second experiment when comparing physiological data and subjective scale from participants playing a game alone against a computer or against a friend, in the same room. Results implied higher engagement and fun when playing with a friend than with a computer.

Collaborative settings are supposed to add something. In learning the proximal zone, artefacts and mutual modelling support and define what is collaboration. the outcome is a better learing when collaborating, partly due to side effects of maintaining a shared representation all along the task.

Now, in this experiment, playing against a friend creates much more engagement.

Efficiency of collaborative learning settings can be due to improved engagement in the task, more cognitive ressources would be allowed (I need a reference showing that more engagement leads to better cognitive efficiency, this must not be too hard). So maybe a shared representation and all these things are side effect of improved engagement. On the contrary we could say that the increased engagement is due to these supplementary tasks due to collaboration (or simply the presence of someone else, when playing the game against the other it is not collaborative, nevertheless a model of the other is still needed).

So the question is maybe like who was first between the hen and the egg:

1. Collaborative/competitive conditions first necessitate several supplementary processings (sharing a representation in collaborative, create a mutual model in both, etc.). Doing this plus succeeding at the task involves more cognitive processes, but the participant reacts by engaging more in the activity, we observe higher physiological responses.
2. Collaborative/competitive conditions engage players intrinsically, various reasons can be invoked such as social comparison, self (other-)esteem, etc. This leads to higher engagement we can call this ego-involvement… So the task becomes more important for the individual and he will develop tools and strategies to achieve it better, such tools can be a shared representation or a mutual model.

I derive something from point 1, related to the Cleap SNF project finishing these days. in fact 1 was partly one of the funding hypothesis of the project. The alternate hypothesis, coming from multimedia learning theories also stated that participants in collaborative learning conditions would need to process several cognitive tasks in order to achieve collaboration (shared representation…). But, relying on cognitive load theories, it predicted poorer understanding since several cognitive ressources necessary to a good comprehension would be used for collaboration and thus unusable (as found by Schnotz, Böckheler & Grzondziel, 1999). But, let’s forget cognitive load for now.

Could we explain Collaborative load and work effects through improved personnal involvement only? Competitive tasks seem more easy to integrate in hypothesis 2 since less “accessory” tasks are needed (mutual modelling is still necessary but a shared representation is less necessary (in fact it is necessary with symbolic games but not in video games, mmh interesting to compare maybe), you don’t need the other to know what you think, you just need to agree on the outcomes of each other’s actions).

Experiment :

What about building an experiment with 2 factors: opponent (computer or human), and knowledge of opponent’s identity (true, false). In order to have self-esteem involved we need to tell the participants that they will meet their opponent afterward, during the debriefing (we can also ask participants to come with a friend and separate them during play, so they know there is someone else, and it is their friend).
Participants would play a competitive game (we also could have another experiment with a collaborative game). They would play 4 rounds, alone in a computer room. Every round the condition would change, twice he would play against another participant once knowing it, once thinking it is a computer, and twice he would play against a computer, once knowing it, once thinking it is a human. Well, let’s make it 8 or 16 rounds in order to have more intra-measures (juste double or quadruple each condition). Balance the order and have another participant in another room playing the same conditions.

Physiological and subjective evaluations would be recorded in order to have informations about engagement, fun and challenge during the tasks.

For the game we can have a quake like (timed rounds), a sports game can be more neutral or a more simple casual game like tetris or another round-based challenge. The kind of game could even be a condition, but in another experiment…

Hypothesis would be that participants thinking that they play against an human opponent will involve themselves in the task more than participants thinking that they play against a computer. The fact that participants really play against an human or a computer will have no effect on the results. We wait no interaction but it is important to prove that only the belief is important.

Ok… well… this experiment is funny but know tell me in what it is related to my bit of theory up there??? damn it, I lost my track or I missed the point… Ok, please do not run this experiment without me!

Reference: mandryk, R. L., Inkpen, K. M., & Calvert, T. W. (2006). Using psychophysiological techniques to measure user experience with entertainment technologies. Behavior and information technology, 25(2), 141-158.

Some theory:

• When looking at games the ease of use is less important thant the enjoyement of the game. This research is a shift from usability analysis to user experience analysis.
• Time and accuracy are traditional measures in cognitive psychology/HCI, but are not really relevant here.
• A successful game environment is something that creates an engaging experience, fosters interactions and is fun, how to measure this?
• Subjective reporting (questionnaires/interviews): convenient and efficient but definitely subjective and not conductive for complex patterns.
• Think-aloud techniques: hard to use here because of the disturbance on the task (play), the experience will not be the same. Retrospective think-aloud techniques can be useful (playback the game and ask to comment), but these techniques qualifies the experience rather than providing concrete, quantitative data. Moreover you don’t measure the task but the reflection on the task!
• Video (gesture, posture and verbal coding): rich but long an rigorus process, on a methodological point of view the analysis is very complex. Concerning analysis time, ratios between 5:1 and 100:1 are reported (duh!).
• Physiological measures: used in emotion research could be very useful and never been employed to assess states like engagement and fun. The article is aimed at clearing the path for this paradigm.

Experiences:
The experiences were designed to verify two research hypothesis:

1. Conjecture A: Physiological measures can be used to objectively measure a player’s experience with enterteinment technology
2. Conjecture B: Normalised physiological measures of experience with entertainment technology will correspond to eubjective reports.

p.143

Collected measures:

• Galvanic skin response (GSR): skin conductivity, related to sweat, linear correlate to arousal (Lang 1995), reflects both emotional response and cognitive activity (Boucsein 1992).
• Electrocardiography (EKG), from which heart rate (HR) and several indicators were derived. reflecting emotional activity, these indicators help differenciate positive and negative emotions. Stress and mental effort are also linked with these indicators (Rowe et al, 1998).
• Electromyography of the jaw (EMG), muscle activity. At this emplacement, it is a very good indicator of tension (jaw clenching). On other muscles it can be used to differenciate positive and negative emotions (brow/cheek). The problem is that if the subject talks or smiles, the indicator measures the wrong thing.
• Respiration amplitude and rate (RespAmp, RespRate), linked with emotional arousal. surprise can cause respiration to stop, negative emotions to cause irregularities, it is also linked with cardiac values.
• They could not collect blood volume pulse (BVP) because it was too sensible to movement (try to play without moving, especially your finger).

Experiment 1:

7 young male participants played 4 times 5 minutes of NHL 2003, once at every difficulty level (beginner, easy, medium, difficult). The order was balanced. Measures were the psychophisiological ones presented above but also several questionnaires, in particular scales about boredom, challenge, frustration and fun.

The results for the 4 scales were significant only concerning challenge (higher when the difficulty is higher). Only the beginner level was perceived as less challenging than the medium and difficult levels. A manova crossing difficulty level and self-identified expertise led to no significant results on the subjective scales.

the level of expertise had a positive effect on the mean respiration rate (expert players breathed more per second than novice and medium players). Both factors showed an interaction on heart rate : expert players had their heart beats stable among difficulties, but novice and medium players showed differences. No other effects of physiological measures were shown.

Individual correlations between physiological and subjective measures often showed results. However, these correlations were rarely consistent across individuals.

Issues:

• Subject enjoyed playing in all conditions, even if the difficulty didn’t match their experience. Experienced players challenged themselves in easier conditions by trying to beat their friends’ scores or score fancyer goals.
• Game play variability: physiological measures are continuous and across time, subjective scales are discrete and measured once. In human factors research (only experts…), physiological measures are averaged for a task, here this technique is maybe not the best (no results in this exeperiment). Achieving fun in games is very variable from person to person, valuable information is lost through averaging.
• High rsting baseline: for some participants resting rates were sometimes higher than play rates (interpreted as anticipation and nervousness). Ways to get an effective baseline and lowering artefacts are discussed.
• Order and interview: The order of presentation may have impacted the result (the latin square design does not allow to use order as a factor here). Subjective measures were collected through interview, this could have an effect on measures also.

Personal note:

In the discussion of experiment 1, the authors state that “without consistent subjective results, we cannot expect consistent physiological results (p150). I don’t understant this, why should it be so? Subjective scales results are cognitive interpretations, after appraisal, after having finished the play, in response to an experimenter. They can be exagerated but also minimized by the participant. I would not be astonished to observe physiological differences even if the participants confess nothing different. I mean, aren’t we doing this kind of reserach because humans can’t precisely assess themselves? I just need to find one study showing physiological effects and no subjective ones. Of course if you get this result you can always say that you don’t assess the same dimension (your subjective scale is rotten)… duh!

Experiment 2:

In order to improve objective physiological evaluation methodology, the authors decided to begin with play conditions engaging players really differently. Ten young males familiar with a game controller (enjoying playing games) took part in two conditions: play against a friend (present in the same room), play against a computer (designed as intra-subject). The difficulty setting was set to match the level of the player. The first participant rested (5min), played against the computer (5min), the second came, both rested (5min), played together, first went away while second rested, and second played. During rest phases, participants were listening to relaxing music through headphones (could not speak together). Concerning scales, questionnaires were used instead of interviews. they also had to decide in retrospect which condition was more enjoyable, fun, exciting and more challenging, and also which condition they would choose to play in.

• Participants who lost against the computer rated the game as more boring than those who won or tied the computer. But when playing against a friend, no differences were seen in the boredom rating (they were never bored, and not the contrary).
• Playing against a computer is significantly more boring, and less engaging, exciting and fun than playing against a friend.
• Same results on retrospective scales, all participants chosed to play against a friend. However half said it was more challenfing to play against a computer (they explainded that their friend was not so good, the other half felt that the computer was too predictable).
• No systematic effect of the game aoutcome (win, loss, tie) was found on physiological data.
• GSR (sweat) was higher when playing against a friend than against a computer (60% of the variance explained by the factor). Examined individually, the increase was significant for 9/10 participants.
• EMG (jaw clenching) was also higher when playing against a friend than a computer (75% of variance explained). But these results take into account facial expressions which would occur more with a player than a computer (really? I need to check that out, I would say yes since the friend was in the same room, but what if players were separated?).
• No differences were observed in HR, RespAmp and RespRate, but they expected this (based on physiological theories which I don’t know anything about).
• Authors also used the continuous physiological data. Response when playing against the friend were very different than when playing against the computer. There is an amazing graph p154 (figure 7) where we see GSR activity when a goal is scored : the response is flat against a computer but a peak rises when the goal is scored against a friend! Also when there is a fight between players (it’s a hockey game), the GSR rises during the fight with a friend, not with a computer.
• Participants who felt that playing against the computer was more challenging had higher EMG than the other half (regardless of play condition).
• The authors normalized all the data using a long formula concerning the physiological data. Then these measures could be correlated across individuals. Normalized GSR was correlated with normalized subjective fun! It was also negatively correlated with frustration (normalized). Several other normalized physiological measures are linked with subjective scales. The best was EMG, correlated with boredom, challenge and negatively with ease (but again the jaw moves not only when your are frustrated).

The authors claim these results (especially experiment 2) support their main hypothesis (conjectures). Physiological responses correspond to how the players preceive the play environment.
These results do not explain if the differences were due to a higher tonic level (the baseline measure, general level of activity) or several positive phasic changes (single responses to stimuli).
Physiological measures provide continuous and high-resolution informations over time and in context. This is great (but so much more work).
Lots of elements can be investigated using such data, the efficient elements concerning video game play are to be underlined. Moreover, the mathematical analysis approach can produce more replicable results than a qualitative annotation of traces.
Other physiological indicators also need to be integrated.
The third kind of data missing in this work is source of objective data (like game events) that are needed to clearely investigate the other data.

Personal note: I would add it is at least true when conditions are really different (solo-duo). Moreover, I don’t like this idea that physiological data is there to confirm subjective scales, these are different things, The few I know about emotions, arousal, appraisal and component models tells me these things are related but different. Next week I’ll report about van Reekum, Johnstone, Etter, Wehrle & Scherer (2004) to get a better idea. These guys used a very simple video game and physiological measures but with an approach issued from psychology of emotions.

This research is very interesting to me since it uses a commercial and complex video game in a factorial experiment. I always have problems in designing experiments for my thesis, the need to control everything in a factorial design drives me to very simple games, games so simplistic I don’t beleive they reflect what games are today… This kind of research gives me hope to integrate “real” games in my thesis, someday, somewhere…

Physiological data are very interesting for me today, they also lack in the multimedia learning community (everyone is rushing for eye trackers but discussions are always around cognitive load and effort… physiological data is the key). But I need to better understand their use in emotion research.

Nicolas had a pretty good insight in his last post. Such moments of peacefullness are precious and need to be underlined.

Now I can take a break knowing I won’t miss anything serious.

I wrote this some time ago as a part of a project. While re-writing the project I throw this away. So I blog it away instead…

Games are fun

In his “Theory of fun for game design”, Raph Koster (2005), tries to define what fun is in a game. The main idea is what he calls a “pattern”, a complex internal representation of how to beat the game. The fun is to constantly challenge this pattern from different angles and difficulties in order to elaborate it more and more. To play “super Mario bros”, the pattern would include factual data (hit the blocks, take the mushroom and jump on the head of the ducks), but also visual checks and motor responses. For Koster, the work of a game designer is to challenge the player’s pattern with care: if it gets too hard too quickly, the player will leave the game frustrated or anxious, but if the player can not challenge his pattern constantly due to a lack of difficulty he will leave the game bored…

Although Koster’s work is not academic and is oriented to game designers, his description of the phenomenon reflects a part of what is very interesting in games for educational researchers. Through the book, the link with (Johnson-Laird, 1983) mental models theory comes in mind very quickly. What a player creates while playing the game is definitely a mental model. At the same time, a mental model is exactly what learning is all about. This link explains the urge to create educational games in the eighties (Willis, Hovey, & Hovey, 1987). However, games based on behaviourist approaches came out and took over the market. The so called edutainment titles alternate educational challenges and reward/punishments reinforcements (a reward would often be to be allowed to play a game). Egenfeldt-Nielsen (2005) argues this genre pushed most of the other educational games out of the market and in the nineties the educational games loss their appeal. Edutainment titles are now only targeted at young and pre-school children. But edutainment is not using the the good element of games, in fact, edutainment are not games, merely a control to access games if the homework are done well… The potential of games for education exists (Egenfeldt-Nielsen, 2005; Frété, 2002; Prensky, 2001), but we still need to better understand what makes games so intense and so overwhelming to compose educational tools in the same way.

Education needs fun

Motivation and how to increase interest of learners in the content or the educational material is a central interest for education researcher. Intrinsic and extrinsic motivation theories (Deci & Ryan, 1985; Kohn, 1996) lead the debate but do not seem really appropriate to describe where the fun of games comes from and how to replicate it easily.

Kuhl (1986) defines three aspects deeply involved in educational psychology: cognitive, motivational and emotional processes. According to him, cognitive processes are about acquisition and representation of knowledge. Motivational processes refer to the learner’s goals and his ways of obtaining them. Emotional processes would be the evaluation of objects and facts in order to accept or reject them. These aspects can help us to lead research in educational psychology but hardly give a way to understand games, fun and how to use them in education.

Csikszentmihalyi flow theory, also called “optimal experience” refers to “the holistic experience that people feel when they act with total involvement” (Csikszentmihalyi, 1975, p36). Observed from play and artistic activities, this state of mind is completely overwhelming. The concentration on the task is so intense that self-consciousness disappears and the sense of time becomes distorted. The experience itself is gratifying to the point that people will want to do it for it’s own sake. Close to intrinsic motivation the interest here is in the task itself and not in the outcome, the task is rewarding every second spent on it, not only through a final gratification. With this theory we are really closer to what we believe is happening in games.

Other researches (Novak, Hoffmann, & Yung, 1997) refined the definition to describe the core experience, its close correlates such as playfulness, its antecedents (skill, challenge, etc.), and its consequences. However, the usual application of the flow theories in education focuses on the quality of design.

…

Reference: Greenwald, A. G. (1982). Ego-task analysis: an integration of research on ego-involvement and self-awareness. In A. H. Hastorf & A. M. Isen (Eds.), Cognitive social psychology (pp. 109-147). New York: Elsevier.

I’m only summarizing the beginning of the chapter. So parts about task-analysis and self-awareness are missing, maybe another day.

Some definitions:

• Allport (1943) enumerated eight senses of ego : knower (experiencing agent), object of knowledge (the body as self), primitive selfishness (antithesis of altruism), dominance-drive (self-esteem, enhancement), passive organisation of mental process (the Freudian concept), fighter for ends (dynamic locus of life goals), behavioral system (defines behavior depending on the environment), subjective organisation of culture (residue of socialization, enculturation). Also look at Allport (1955) for an updated taxonomy.
• The point of Allport (1943) is that when someone is “ego-involved” in a task he acts and reacts differently than when in a neutral/impersonal routine. The eight dimensions (previous point) or most of them are implicated in the action.

Ego-involvement is, as the phrase implies, a condition of total participation of the self-as knower, as organizer, as observer, as status seeker, and as socialized being. (p. 459)

• The effects of ego-involvement is quite varied according to Allport: distortions of judgment, superior memory retention,  necessary for optimum intelligent performance (what does it mean? problem-solving?). The author also asks for learning reforms and revision of motivational concepts.
• Sherif & Cantril (1947) based their work on the definition of ego as a constellation of attitudes (ego-attitudes), everything that define status and/or role. This theory is social-based (one’s defined by what he represents for others). This is created by group activities and social situations. Ego-striving is what the individual do to get the place he wants (the personal part) among a group, and to secure this place.
• Iverson & Reuder (1956) tried to emply ego-involvement as an experimental variable. They stated that ego-involvement appears to be manifest when self-identity (personnal important feelings or status) is threatened. We see ego-involvement in situations where individuals have a need to enhance or mainain self-esteem (possibility of interference/deprivation).

Ego-involvement: three different meanings

1. Concern about evaluation by others: dominant in the interrupted-task research. the task provide a measure of a valued skill, ego-involvement is a motivational state : evaluation apprehension (Rosenberg, 1969). The issue is other-esteem : impression magagement.
2. Concern about self-evaluation: self-esteem threats and achievement orientation define better this aspect than evaluation apprehension. The issue is self-esteem : self-image management.
3. Personal importance, commitment: More on the side of attitude change, audiences are more resistant to change on important or controversial topics (look at Sherif’s social psychology work). this aspect of ego-involvement is implicated in the resistance to change of important beliefs : value management.

Ego-involvement can occur when an otherwise unimportant task becomes relevant, subordinated to an ego task (which can be of the three flavours we’ve just seen).

Research review and paradigms:

1. Ego-involvement is interpretated as a threat to self esteem (interrupted-task paradigm).
2. Ego-involvement attitudes tend to resist change (ego as a protective/defensive agency of personnality)
3. Allport (1943) defined ego-involvement as “a condition of total participation of the self” (p. 459). (Did Csikszentmihalyi refered to this paper somewhere?).

• Interrupted task paradigm: Based on Zeigarnik (1927, 1938) results showing that interrupted tasks were often better recalled than completed ones (from manual tasks to mental problems). The paradigm is to let the subject perform some of the tasks to completion (half) and interrupt him on the others when he is deeply engaged. After that participants had to recall the tasks (unexpectedly). interupted tasks were recalled around two times more than uninterrrupted ones (called the Zeigarnik effect). The explanation is that an engaged task is charged with motivational tension that would persist until the end of the task. Interrupting the task would not completely free the tension, bringing better recalls (and tendency to try to finish the task afterward; Ovsiankina, 1928).
• Personal factors in Zeigarnik effect: Ambitious subjects (much interrested in succeeding at the task) recalled more interrupted tasks. Subjects interpreting the interruption as a failure (believing they had limited time) recalled less of these interrupted tasks (than the other interrupted ones). The effect was also stronger on children than adults. Adults that “let themselves go” (?) recalled also more interrupted tasks  than others.
• Ego-involving instructions: The idea to induce more involvement through instructions. Participants were told to interpret interruptions as failures and that the experiment provided an index of intelligence. In such cases, the effect is inversed and more completed than interrupted tasks were recalled. But this result had exceptions and depends on several personal factors.
• Task variation: some tried to define other tasks than the long single-subject of Zeigarnik. They used rearranging scrambled short phrases. However its harder to compare on the recall test, finished sentences could be easier to remember. Moreover a certain level of isolation (difference) is necessary between the items in order to build different tension systems (else there is only one task, one tension system).
• Ego-strength: Studies tried to define this psychological dimension. Weak ego subjects need to defend self-esteem even when the objective situation is not threatening, but in skill instruction (aka ego-involving) situations they are overwhelmed by the failure and remembers more of them (uncompleted). High-ego subjects would defend self-esteem only in stressful situations by recalling more completed tasks (Alper 1952). Trials to define the dimension linked it in different ways with dimensions like pessimism, need for recognition, ego-ideal (?) but also neuroticism and extra-introversion (Eriksen 1954, Eysenck).
• Experimental involvement and subject roles: Zeigarnik (1938) identified three types of subjects, those who wanted to perform in order to please the experimenter, those by competition with others and those interested in the task itself, tried to solve the problem in the way it was asked. in Weber and Cook’s (1972) subject roles theory, they are “Good subject”, “apprehensive subject” and the “faithful subject” (in the same order). They are seen by Alper (1946) as “ego-involved” or “task-involved”, doing the task because it threatens their self esteem or doing it because they are asked to.
• The task-involvement/ego-involvement differenciation partially corresponds to extrinsic/intrisic motivation. In both cases the performance is motivated by an outcome which can be performance (other’s approval) or characteristics of the task. Links are also done with cognitive dissonance and social comparison theory (Festinger, 1954).

Technorati Tags: ego psychology motivation ego-involvement

I’m just back from a talk of Guido Gendolla at the swiss center for affective sciences, just some hundred meters from the lab. The talk was really inspiring for me, first because I have very few notions on affective sciences research and second because of the relevance of the studies for my actual interests.

Some notes:

• Research based on Brehm & Self (1989) motivational intensity theory. Which was adaptated by Gendolla (2004). I should quickly review at least the first one.
• For Greenwald (1982), the “self” becomes involved in an activity in three conditions: 1. Social evaluation (I want to please the crowd), 2. self-evaluation (intrinsic motivation), 3. Personal values/interests/goals (is it a jocker category??)
• We know that when the difficulty increases, the energy involved in the task increases also (that’s effort). The main questions behind the researches is to understand the importance of the level of motivation when we are self-involved in this relationship between difficulty and effort.
• There were some nice theoretical graphics about a linear relationship between difficulty and effort. Added to this, the motivation potential was the maximum of effort someone would invest in the task. If the effort he would have to invest depending on the difficulty presentied is higher than this threshold, the participant would simply give up and invest no effort. So effort would depend on the difficulty but is limited to a personal motivation potential which the authors claim is mostly defined by self involvement.
• Another nice issue was when the difficulty is unfixed (when the participant can work and answer at its own pace). In this condition, the effort can vary hugely between participants (but is limited to a maximum defined by motivational potential).
• I liked the pizza example : you have 20 bucks (that’s your motivational potential), you want a pizza, if the price is 10.- (difficulty), you’ll pay 10.- (effort). If it’s 2.- you’ll pay 2.-. but if it’s 25.- you’ll just give up. If the price is unfixed, you’ll pay something depending on you (I’ll not develop more) but the maximum will always be 20 bucks.
• The big question is how to measure effort in one moment. Their answer is physiological mesasures (systolic and diastolic blood pressure, heart rate and skin conductance reactivity) : more effort inducing a higher reactivity (or maybe higher reactivity being interpretated as more effort).
• Experiments manipulate the ego involvement (importance of success to the task for the participant) and difficulty. The tasks are letter detection or memory tasks, easy to make more difficult though presentation times and number of items).
• I found a paper of Gendolla & Richter (2005) about one of the experiments presented, might be usefull.
• After that, they added mood in the equation, moods are long-term disconnected feelings (not like emotions which are discrete identifiables and with a clear cause). In some studies (unpublished I think), they manipulated mood or took students in different moods (from depression scales). Initial mood defines the motivational level (defines the behavior related judgement which defines the effort mobilisation).
• Participants in positive moods act as if they would perceive the task as easier (than the ones in negative moods). They invest less effort but also try to resolve harder tasks (they meet their motivational threshold later). I don’t know what it changes in terms of success at the task (more effort means better results or not? I don’t remember what the guy said).

Relevance for me :

This talk gave me ideas for the both sides of my research activities. First about multimedia learning and second about understanding games.

Actual multimedia learning research is like jammed by the cognitive load theory (Sweller, Van Merrienboer & Paas, 1998) (sorry I will not give details here). This theory can be used to explain almost any result. This is not a bad theory but researchers have some problems getting out of it. In particular as many material effects are not significants the research needs to dig into complexity, to understand how hard it was for the participant. The trend (chevreuil) is to look at participants exploration of the instructional materials, eye tracking displays are selling like rabbits.

However I’ve seen no multimedia learning research using such physiological measures to assess continuous (and objective) effort indicators. We need to put learners in one of these machines with captors and compare different multimedia displays. Why not try to replicate Mayer’s (2001) findings and principles? First these principles really need to be replicated and second we would have accurate effort values to consider in the equation.

But will this work? Learning tasks are far longer, continuous and more complex than the ones in psychophysiological studies. Would we find something through these measurments? I’m affraid the global effects would drown due to the complexity and number of cognitive processes involved in time.

About video games I would really be interested to measure effort with the same kind of methodology in different tasks. Here are some experiments ideas:

• Prove that being in a virtual environment does not improve participant’s capacities (it’s an argument heard so many times: use 3d and everything shines). Ask participants to do simple cognitive tasks (memory, response times etc) in a 3d environment, on a computer (simple multimedia) and on paper. And show no difference in terms of efficiency, effort, nor time.
• Study the effect of games on player’s cognitive abilities: Have participants do tasks presented as a game or not (that’s very hard to design in order to be comparable). then could we say that being in play is a positive mood? though these measures and other, could we see comparable effects as what Gendolla found ? Can we explain the (self) involvement in games through Brehm & Self (1989) theory ? And can we verify this experimentally ?
• Another idea involves game design theories: we often see graphics about carefull design of difficulty (not hard enough, the player is bored, too hard he gives up). I would be interested to explain this through Brehm & Self (1989) theory, using difficulty, effort and motivation. Then verify it through empirical research: we can modulate the difficulty curve of a simple game and make the habitual measures.
• Some more long term qualitative analysis of someone playing would also be helpfull. I would be interested in the development of motivation and invested effort over time, in parallell with the improvement of game-related competences. The level of analysis could be a quest, a fight, a level…

Ok, now I need to write a project…

I read this yesteray (writen years ago but just published). Far too provocative to be heard but I felt like the guy had something right. The video games world is made of a lot of people doing a lot of things around a central pillar (with a lot of expectations). But none of them have common visions or goals.

I then have been pointed to this more exhaustive post about the death of computer games: C. Crawford, “Computer Games Are Dead”, Interactive Entertainment Design, Volume 9, http://www.erasmatrazz.com/library.html, 1996.

The point is better made, it’s 10 years old but still inspiring. You’ll find good thoughts about the overwhelming market-driven approach that killed innovation:

I’ll not decry this power shift as an evil; it just happened. Marketing people aren’t bad or stupid or crass, and designers don’t hold the keys to goodness and light. But the shift from a creator-driven atmosphere to a market-driven atmosphere worked a profound change on the organism, transforming it from a future-looking creature to a past-looking creature. At its heart, the creator-driven approach concentrates on the future, on what might be. The creator’s whole point and purpose is to move beyond the existing limits and explore new areas / to change. This emphasis on change is at the core of what we mean by “life”.

Market-driven approach is for products that don’t evolve (I like the example comparing laundry detergent and genetic engeneering).

other points discussed are :

• the inverted gold rush (games always need more money and it is a proof of quality),
• the closed distribution system (The industry copies what worked, you can’t publish something different),
• the death of creativity (IT evolves faster than anything but no one surfs on this),
• loss of community (growth alone don’t explain the disapearance of the family felling of the CGDC),
• absence (disapearance) of a moral attitude in the industry (unwritten ways of doing things, professional morality juste don’t exist, is it because of the gold rush?)

And some other forces are there, multimedia, the internet,

Let’s start with multimedia. What is most striking to me about multimedia is the fact that it isn’t gaming. That is, multimedia is just another term for interactive entertainment, but there’s a clear connotation of differentiation from gaming. We may not know what multimedia really is, but we do know that it isn’t gaming. Yes, computer games use CD-ROMs and sound boards and full motion video, just like multimedia products, but we still know that computer games are distinct from multimedia. This distinction implies divergence, and divergence means that multimedia won’t save computer gaming. I think that multimedia represents a society-wide rejection of computer games. After all, if everybody thought that computer games represent the path to the future, then what need would there be for an alternative path utilizing the same means? The rapid growth of multimedia represents a broad desire for something other than computer games, something different. Therefore, the progress of multimedia represents not the salvation of computer games, but its bane.

Well, I would have liked to know in what those media differ exactly, in what can we compare them or not. That would be useful for my PhD…

I’ll skip some part about projections with the internet culture (The article is 10 years old and the response to forecastings is always “something else”).

I am not suggesting that computer games will drop off the face of the earth. Indeed, they will surely persist with the same durability demonstrated by, say, model railroading, amateur photography, and woodworking. But this generation has dropped the torch in its scramble for quick gain, and has lost its shot at creating a living medium with a bright future. Instead, we have created a hobby, a good and fine thing, to be sure, but nothing approaching the potential that we optimistically contemplated back in the early 80s.

Yep, we are far from a media, a language like the one cinema created: artistic pieces are far from using games as a media. Some of them are looking towards multimedia and interactivity, but no games (show me some if i’m wrong). Well, this is not completely true, some web/casual games have the power to carry a strong (political) message (about darfur genocides, iraq bombings).

The article: Dickey, M. D. (2005). Engaging by design: how engagement strategies in popular computer and video games can inform instructional design. Educational technologies research & development 53(2), 67-83.

The purpose:

• In video games, participants are really engaged in gameplay, can we understand this involvement and descripe what creates it in order to design more engaging instructional materials?
• Compare the way video games engage players with instructional models of engagement.

Very quickly:

Designing games or engaging instructional materials are very close activities. They share several aspects and the later could really take inspiration from the former. Several aspects of game design are described and parallels are made with instructional design theories.
The main points:

• Research on educational games focused on how to improve existant curiculum and materials. But lack of research into design aspects and what made enternainement games so exciting (p.68).
• Review about studies showing a potential of the methods we see in video games :

• Bruckman (1993, 1997) on community/RP in MOO.
• Rieber (1996) on intrinsic motivation, imitation, feedback, challenges provided by games.
• Prensky (2001) on requested engagement, discovery, trial/error, problem solving but also consistency with contemporary youth and necessary actual skills.

• Bowman (1982) on motivational support in Pac-man and how to extend this to classrooms : extrinsic support (visual and aural feedback) and accomplishments are not everything. Linked with the flow theory.
• Provenzo (1991), based on Malone (1981), few negative consequences in games. Also linked to motivation : challenge, fantasy and curiosity to be incorporated in educational environments.

• In short: clear goals and tasks, reinforcing feedback and increasing challenge. But the games observed (pac man and super mario 2) are simplistic as compared to what gamers play today.

Research from a constructivist perspective focuses on the relationship between the player-learner and the environment, and the social aspects of the design, whereas research from a cognitive perspective focuses on the internal aspects of motivation and schema as fostered by design.

• Player positioning: representation evoluted in games: out of shot, isometric view, POV, first person (become part of the environment) (Riddle 2002). Parallell with learner positioning (in different theoretical perspectives): behaviorism (external positioning) to construction of knowledge (become part of the content) in project or problem-based pedagogies. The engagement seems to follow this centering on the player/learner. Winn (1993) argues that virtual environments can be a bridge between experiential learning and information representation, immersion would mean better understanding through interaction (Personal note: I disagree with this, computer interaction is still a metaphor, we still have a strong level of representation, 3d is no miracle and will never be).
• Narrative: If narrative is part of the gameplay is a debate (Aarseth 2001, Juul 2001). narratives are traditionnaly linear but games present a new nonlinear approach. Different manners are described. Then fantasy, roles and case studies are used in education (but for specific domains). Integrating narrative in instructional design supports multiple aspects (reflection, comprehension, evaluation, exemplification…) (Eisner 1988; Conle, 2003). (virtual) space as narrative places and player’s interactions allow to experience a story and to achieve their own conclusions. What about designing educational spaces instead of educational materials?
• Interactive design: Settings is a main elements in game design in order to engage the player. The game settings are defined in terms of physical, temporal, environmental, emotional and ethical dimensions (defined by Rolling & Adams 2003). The careful blending of these dimension defines (in part) the “efficiency” of a game. From an instructional perspective, settings and scenarios are a base in problem or project-based methods, often as a context in order to embed the learning activity and the environment. However these settings are never as fully defined as in games.
• Interactive design: Roles and characters are also a main game design element and are essential to engaged play. The building of the player character and the NPC’s support both the immersion and the narrative. Techniques are used in order to design empathic and compelling characters (models for the player) (Freeman, 2003). the use of roles, role reversals and the emergence of anonymous identities lead to a reduction of inhibitions. Scaffoholding agents might gain something with this (personal note: agents… urk!).
• Interactive design: Actions, feedback and affordances. As most gameplay is goal oriented, working with hooks (action, resource, tactical and strategic or time hooks; Howland, 2002). These are elements that grab the player’s attention and activities, elements of design to augment engagement (personal note: meaning dependency and impossibility to get out). These are sub-activities, micro management, resources informations, etc. So choices are in the center, providing things to do, tools and awareness, games let the player do their best. Educational simulations share elements of interactive design but also results (reflection and analysis). Methods to describe an instructional material could be borrowed from video games.
• Engaged learning theories theorically describe key components of how to achieve this state. However few models or recipes are in the literature. Game design elements clearly have something to give. However, possible activities and environments are numerous so it is premature to generate operational guidelines (Personal note: good, there is still work to do). Exploratory questions and guidance would be more clever.

The main idea:

• I want to link my research on animated pictures learning with my PhD on video games.
• Right now I’m interested in motivational and emotional aspects of video games
• Some reading on multimedia learning theory makes me see a link: from a certain point of view researchers on multimedia learning tried to boost motivation in their materials, with no success (learning was depreciated). Why? Why is it different? Does it mean that we can’t learn from games?

The post: 

In Huib Tabbers’ Dissertation (.doc), page 10:

First of all, information that is irrelevant to learning but only meant to make the multimedia instructions more fun is redundant. These information elements are often added to spice up the learning materials and keep the students motivated. However, these “seductive details” as Mayer calls them, seem to do more harm than good to learning, as the results of several studies have shown. Adding text and pictures or video clips that illustrate the subject of the learning material to make it more interesting (Harp & Mayer, 1997, 1998; Mayer, Heiser & Lonn, 2001, experiments 1, 3 and 4), or adding entertaining background music and sounds (Moreno & Mayer, 2000) all resulted in a decrease in transfer performance

Well, these studies are investigating redundancy and possibly split-attention effects but the way Tabbers put them here, it’s about things that we put in materials in order to keep the learner’s attention. In multimedia learning, those things are known to be bad for comprehension/learning. Their motivational potential does not work (I still need to investigate this potential and what they meant in those precise studies).
Maybe that’s my missing link between multimedia learning and games!

Right now i’m interested in what makes the motivation in games so huge. I try to see video games as very high level interactive animated pictures (to simplify). Clearly this motivational aspect is present in games. What is different so we can’t put them in more rude multimedia presentations? Well it’s not cristal clear right now, these things are so different, we’re on different aspects and i do not want to “fall” in a description of what is a game element or not.

Cryptic post, I hope I can understand it next week !

The main ideas:

• “Augmented-reality” video games are starting to be developped, based on board games ideas
• The board games community is an inspiration for video games designer
• Great and elegant game design ideas, come from this community, due to the market and the small budgets. Does less money implies better ideas?
• How could we look at similarities and differences between board games and video games ?

The post:
Kotaku has a video straight from E3 presenting a new collectible cards game, eye of destruction. Based on sony’s Eye toy technology it should be relatively easy to play at (well, if you have a playstation, an eyetoy and if you buy those collectible cards…). In fact I mean “more accessible for the upper-average occidental home”. if you liked it, look also at this video of gizmondo.

Anyway I wanted to report again this interest of game developers for tangible interfaces. But I’m more interested in the mix between console games and classical table games. Video game designers are now looking into the classical, tangible, board, table (…) culture to get new ideas. It is a good idea because the costs to develop such games are so infinitely lower that you can find more passionate and original designers, game desing ideas are also more clever.

As a matter of fact the “table” gamers are often harder to contempt (smaller market, more dedicated players), and you simply can’t put a lot of development on the visuals (well, some graphic design but that is not so important). Game mechanics are more important than in a video game because the player is conscious of them. They also have to stay simple and elegant because the player has to know them all by heart ! (I already reported about this pregnance of rules in board games). The lack of money makes a lot of little boardgame companies to disapear after one or two great products, sales are often low (like for original video game developpers in facts).

A broad study comparing video games and board games would be interesting to find parallels and differences. What elements in board games are transferable to video games, what can we keep and what can we improve? Is it a good idea to make the rules and micro-processes disapear behind the computer (D&D online lets you see the dice rolls to keep the feel of table top rpg)? A lots of processes (especially social ones) are lost through the “digitalization” process, which ones? how? why? Is a game such as “eye of destruction” a video game?? you don’t need a computer to play it (just use the cards), the console only shows nice 3D animations and maybe runs some hit points related rules, it is an augmented board game…

This is related:
A board game based on cards already tried to spread into the third dimension (like eye of destruction), but through using lead miniatures (to be painted by the player), it is the french game confrontation. I couldn’t help thinking of that game while looking at the video.

Interactive tables are hot and games are getting excited…

Last week my research lab spent 1.5 days in the country. It was a great oportunity to present what we’ve all really been doing and what we would like to do now. Everyone presented his work and I must say It’s good to know our colleagues’ deep interests. Interests emerged and recognised themselves. It’s just sad that we only had time to present our works and talk about them. Another half a day in workshops or projects building would have been very profitable, but that’s for next year.
Anyway, I had a hard 30 minutes in talking about 4 things (get the ppt, its in french) :

1. The cleap project : it’s now done with 4 experiments and 10 publications. I reviewed some aspects on the split-interaction hypothesis. Last stats, papers and archiving are still pending…
2. My PhD about video games : I quickly presented the rejected SNF project. Then I talked about the project i’m writing right now : more on the emotional perspective on flow and motivation. More about it on this blog as soon as I finish to write it! Till then you can look at the 2 slides concerning it.
3. My actual job: managing 10 tutors in an online course for students of the geneva university. That was only informative…
4. One of those projects I would like to do when I have some time for it (which very probably means never). That’s about this idea of a movie with totally free viewpoint, in a 3d engine. I already elaborated on the concept somewhere on this blog.

About the randominaire itself you can find shared pictures here. But don’t miss Swiss music tryout, the place itself, mustache taming, team building exercices or serious games (?).

I just let dust accumulate on this blog but don’t worry, my project’s refusal didn’t kill me. In fact I have had to find other ways of subsistance and this led to lots on things at once. No that much but all different and hard to cluster. However I’m blogging today to launch a new use for an everyday word, here is the story:

As my girlfriend is in contemporary art I often wander in arty places, galleries and stuff. And for some time now I, see a kind of hip for deers (and she too)… When I started noticing that deers and other big forest animals were sometimes in art pieces, I also started noticing them everywhere. Of course, me wanting to see them helped a lot and I do not have any quantitative analysis to prove that… but who cares!

Of course I don’t have precise things in hand but a short search through flickr got me some nice pictures from art pieces: a deer with walking braces, a strange deer head, this one ate a toaster. But pictures are hard to find because those deers in art pieces are usually in the background, like hinted or referenced. Try to find them for yourselves!

Deers in art pieces are present for some time now, In fact they start to appear in T-shirts, I start to beleive I’m not the only one noticing this trend and designers get on the deer, I mean on the hip…
So from now on I will use the word “chevreuil” (french for deer) in order to refer to something which is hip, fashion, “on the wave”, trendy or so on. I guess that like with my past favourite expression (”rocknroll”) I’ll start to open the meaning to every aspect/concept close to “interesting”.

So animated pictures are not so chevreuil right now, my actual chevreuil is on emotions in games with a deery experimental psychology approach.
Well, I guess it sounds better in french. Just find another word you englishers (but avoid “stag”, I just discovered the alternate meaning of this word in english…).