2D, stereoscopic 3D and HMD virtual reality video games

In comparison to the conventional 2D technology, which does not give depth to the objects, 3D and VR technology offer additional experiences for the user. The technologies allow for the perception of spatial depth on the screen. In the 3D technology, this is realized by using stereoscopic displays, which creates spatial depth off the screen with 3D pop-up visualizations, so that some objects within the game appear closer to the player and seem touchable, the usage of 3D-capable screens or the usage of dedicated 3D-capable glasses. 3D technology can be more realistic, immersive and allows e.g. improved eye-hand coordination, as compared to the 2D technology experience that lacks depth perception. One of the first 3D home video game devices was Tomytronic 3D in which 3D was simulated using two LCD panels, developed by a Japanese toy maker in 1982. Other early home devices were developed by e.g. Nintendo in 1995 [28,29].

In comparison to the 3D technology, HMD VR tries to deliver an even stronger feeling of being in the world or in the moment [30]. VR is a computer-simulated reality, where the player immerses himself/herself in a fictive 3D world by using a special head-mounted display (HMD), which is a headset that shows visual effects directly in front of the player’s eyes (e.g. Google Cardboard, Oculus Rift, Samsung Gear VR). In VR, the player can interact with and in the environment and is sheltered from the outside world as compared to Augmented Reality (AR) (e.g. Google Glass), where the visuals can be projected on glasses, too, but the player is not sheltered from the outside world) [9,31–33]. In VR the user can move through and experience the game world while thinking that he/she is truly somewhere else. Hence, a VR video game might offer the possibility to be more realistic than a game displaying a 2D or 3D condition [34–36]. Arguably, the first VR headset with goggles was developed in the Mid-Eighties by VPL research and Jaron Lanier. Nowadays, many different VR types and headsets exist. One of the least expensive open source models, developed for Android smartphones, is Google Cardboard. Google Cardboard can be made by the user him- or herself or is sold at prices between 5 USD and 20 USD [30,37]. A more sophisticated device is the Oculus Rift that is currently sold at about 399 USD. Other well-known technologies include the Microsoft HoloLens, a holographic computer, where the user can engage with the digital content and interact with holograms in the world around him or her [38], the PlayStation VR [39], HTC’s Vive, another content streaming headset [40] or the Samsung Gear VR [30].

However, the quality of 3D technologies [13,41,42] as well as VR technologies [35,36] have been debated critically and researchers have demonstrated a range of negative effects, too, e.g. discomfort, eye fatigue, dizziness, headache, disorientation or motion-sickness when using these technologies. Furthermore, VR users might become socially isolated when using head-mounted displays, as a consequence of locking their eyes and ears into a fictional video game generated world for longer periods of time [36,43,44].

Presence in the 2D, stereoscopic 3D and HMD VR video game

The concept of presence in virtual environments has received a lot of attention during the last decades, especially with the rise of interactive technologies in the 90s, and has been debated from different perspectives (e.g. [31,45,46]). Presence in a virtual environment can be described as the sense of being in a virtually mediated location instead of being in the real location (the place the person is actually located in) [47]. The sense of presence plays an important role in linking perceptions, intentions and actions of an individual in the environment (e.g. [48–52]). High levels of presence in a virtual environment allow the subject to put his / her intentions into action, to monitor the actions and adjust activities if needed. The subject can adapt the own action to the environment [52].

The level of presence seems to be of special interest for the comparison of different technological environments and in the game context. The question is how deeply participants are immersed or “inside” the game [53,54]. Kim and Biocca [55] speak of “departure”, which describes the feeling of not being in the physical environment anymore and “arrival”, which describes the feeling of being within in the mediated environment. Especially in the VR environment, and to a lesser extent in the 3D environment, game players are likely more immersed in the mediated environment (“arrival”) and perceive less of the physical environment (“departure”). If elements of the 3D or VR game play are seemingly touchable as they appear to be around the player, the player probably pays more attention to the 3D effects of the mediated environment, hence the level of immersion and presence should be increased. In the VR video game, wearing the goggles, the player is even sealed off from all visual stimuli around him/her in the physical environment. Hence, there will be an even higher level of “departure” than in 3D. The player will be more immersed and absorbed in the mediated environment, as the stimuli from the mediated environment are practically the only stimuli he/she receives while playing the game, leading to a higher level of “arrival”.

We therefore expect that presence should be higher in the VR than in the 3D than in the 2D video game.

“H1: Presence will be highest in the VR video game, lower in 3D and lowest in the 2D video game.”

Arousal while playing the video game

Arousal can be defined as stimulation, alertness or activation and is a process, which initiates behavior [56]. According to Bolls et al. [57] arousal “indicates the level of activation associated with the emotional response and ranges from very excited or energized at one extreme to very calm or sleepy at the other” ([56]; p. 629). Measuring arousal in different media formats, such as 3D or VR, has become a common practice in research settings (e.g. [58–61]). Video game players will experience arousal, depending on, for instance, how exiting or involving the playing experience is. Levels of arousal that are too low might lead to boredom and reduced attention directed towards the game. In contrast, if the gamer experiences too much arousal, attention can also be diverted. Hence, the level of arousal elicited by a video game is an important variable [62]. According to previously conducted research, games which are played with 3D or VR can cause a higher arousal than games played with a simpler technology, such as 2D (e.g. [58,63–65]). Thus, we expect a higher arousal in the VR condition than in the 3D than in the 2D condition.

“H2: Arousal will be highest in the VR video game, lower in the 3D and lowest in the 2D video game.”

Attitude towards the video game

Attitude towards the game in this research is defined as the overall evaluation of the game played. Attitudes are a composite of feelings and beliefs as well as behavioral intentions toward an object [66]. The components are highly interdependent and influence each other. When playing a computer game, individuals will like the game more or less and evaluate it more or less positively or negatively. Attitude towards the game is an important variable for game producers as it determines to a large extent how the player will react to the object, e.g. how much time the player is willing to devote to the game, whether or not he/she replays the game or recommends it to somebody else. Attitudes toward the game are also important because they impact the brands that are embedded in the computer games [67,68].

As outlined above, 3D and VR offer additional technological features (e.g. higher immersion) that may allow for a better attitude towards the game. On the other hand, negative aspects are also related to the technological enhancements (e.g. higher visual fatigue, dizziness), which are likely to impair attitudes towards the game. Since the literature is contradictory with regard to whether the positive or negative aspects related to the technology enhancement dominate, and since no research has examined the attitude towards the game when playing the game either in a 2D or in a 3D or in a VR condition, hence no empirical evidence is available so far, we investigate the impact of the technology on attitude towards the game and formulate the following research question:

“RQ1: Do attitudes toward the game differ in the 2D, stereoscopic 3D and HMD VR condition?”

Attitude towards the brands placed in the video game

As outlined above, brand placements play an important role in video games. Though previous research has analyzed several factors that influence the recipients’ attitude towards brands placed in computer games, such as game involvement [69], enjoyment and attitude toward the game [70], or brand prominence [71], surprisingly little research has addressed how different delivery modes such as 2D, 3D, or VR might impact the brands placed in video games. Drawing from related research fields, mainly from advertising and product presentation on websites, there is some indication that brands might benefit from 3D as compared to 2D presentation. Li et al. [72] found that ‘flat’ 3D advertisements on websites generate more positive brand attitudes than 2D advertisements. Flat 3D means that the product representation is 2D, but that users can rotate products, animate their functions and features or can zoom in or out for inspection. These findings are also consistent with the study conducted by Choi and Tylor [73], which shows that flat 3D brand representations (by moving, zooming and rotating the object) on websites leads partly to a higher brand attitude than 2D brand representations (static pictures). However, this was only the case for a geometric product (watch) but not for the tested material product (jacket) [73]. Debbabi et al. [74] report similar findings, since the brand attitude was more positive for flat 3D Internet-based advertisements than for 2D ones. According to Lee et al. [75], consumers’ brand attitudes were more responsive and were held with greater confidence for flat 3D visualized products on an interactive website than for 2D products on a website that was static. Kerrebroeck et al. [76] demonstrate that attitude toward the ad, attitude toward the brand and purchase intentions were higher in the case of VR versus 2D. In their experiment participants watched a video either in a 2D condition on a mobile phone or in a VR condition on a HMD based Google Cardboard-type device.

To the best of the authors’ knowledge, no study has addressed this question in the context of 2D, stereoscopic 3D and HMD VR video games. Hence, several studies have shown that an enhanced technology might influence the attitude towards the brand positively. We therefore derive the following hypothesis:

“H3: Attitude toward the placed brands will be more positive in the VR video game compared to the 3D, and it will be least positive in the 2D video game.”

Memory for the placed brands

Recall and recognition are the most common measures to examine the memory of brand placements (e.g. [25,27,77–79]). Recall is the ability of a person to retrieve a brand name correctly from memory without any mention of other brand names or the product class. Recognition is the ability to remember that there exists past exposure to the brand and it is usually measured by using aided memory based techniques, e.g. where brands are listed and the person can choose the brand/s which he/she has recognized [80,81]. One model that may explain effects of technology enhancement on the memory for brand placements in computer games is the limited capacity model of motivated mediated message processing [82]. The model assumes that an individual’s attentional capacity and his/her ability to process information cognitively is limited. The cognitive capacity, which is used to perform a primary task, cannot simultaneously be used to accomplish a secondary task. When playing video games, the game play is the primary task because the player primarily focuses his/her attention on those aspects which are relevant for a successful game play. While the player focuses his/her attention on the primary task, fewer cognitive resources are available for secondary tasks, such as processing embedded advertisements [82–85]. The more attentional capacity is needed to play the video game, the less capacity will be left for processing the information about the placed brands [83,84].

It can be assumed that a 3D and a VR condition require more cognitive resources than a 2D condition. The depth perception in the 3D and the VR environment and the higher complexity of the VR world in general are additional items of information that occupy more cognitive resources. There is also empirical evidence that supports this assumption. Mun et al. [86] demonstrated in cognitive tests that the brain activity of recipients who were exposed to a stereoscopic 3D environment was higher than that of subjects who were exposed to a 2D environment. Furthermore, those who were exposed to the 3D environment needed longer execution times for tasks and paid more attention to the 3D effects than to other areas because of their visual fatigue (exhaustion of the eyes). Yim et al. [87] explored how stereoscopic 3D technology in comparison to a 2D display influences the viewers’ memory of brand names embedded in a soccer game. Results showed that the viewers remembered less brand names in the 3D condition compared to the 2D condition [87]. Other studies also found that subjects have a longer reaction time in 3D conditions than in 2D conditions, since their cognitive load is increased (e.g. [88–90]). Furthermore, in a 3D as well as in a VR environment, the backgrounds are often blurred and the 3D or VR environment can stress the viewers’ eyes, which can also lead to cognitive fatigue and a reduction of attention. Comparing 2D, 3D and 4D movies (3D plus scent), Terlutter et al. [91] found that memory for brand placements suffered in the 3D and 4D condition as compared to the 2D condition, except for one extremely prominent placement that was better memorized in the 3D condition, and they attribute the typically lower memory in 3D and 4D to the greater amount of cognitive resources needed for processing the central movie plot, leaving less resources for processing the brand placements.

Thus, the player has to process more information in a VR and a 3D video game in comparison to a 2D video game and hence more cognitive resources are needed for the game play (the primary task), leaving less cognitive resources for secondary tasks, such as memorizing the brand placements in the video game. This leads us to the following hypothesis:

“H4: Recall and recognition of the brands included in the video game will be lowest in the VR condition, higher in the 3D condition and highest in the 2D condition.”

Control variables

Skepticism towards advertising, general attitude towards video games, prior video game playing experience and video game literacy serve as control variables.

Main study design

In order to address the above proposed hypotheses and research question, a “jump’n’run” video game was designed and developed in a 2D, 3D and VR condition by a professional game designer. The 2D condition was developed with traditional pictures and without depth of the objects. The 3D condition refers to stereoscopic three-dimensional technology, where the objects pop-up with true depth and float off the screen [14]. The VR condition was based on the HMD technology. The game was designed with the software tool Unity 3D. We chose a jump’n’run action game for several reasons: Action video games place a focus on the player’s reflexes as well as on his / her reaction time, the eye-hand coordination is important and players try to accomplish a goal [92]. “Jump’n’run” video games are one of the most common game genres and spatial depth perception is likely to be of relevance for game play. The game consisted of three levels of increasing difficulty. Playing time was between 7 and 10 minutes. In the 2D and 3D condition, the game was played on a large 46-inch, stereoscopic 3D-capable television. Participants who played the game in the 3D condition received special 3D glasses, which allowed the players to experience depth perception. Players of the VR condition wore an Oculus Rift headset. The whole game was accompanied by the same music in all three conditions. Participants played the game by using an Xbox 360 controller for Windows.

In total eight different brands were integrated in the game, each brand appeared in each level (hence each brand appeared three times during game play). The brands belonged to the following product categories: airline, chocolate, bank, energy drink, coffee, fitness club, nachos, and smartphone. All brands were fictitious to avoid confounding effects of previous brand knowledge. The size and presentation type of all brands were kept constant across all three different levels and across all three game conditions. The brands remained on screen until they were collected by the player. One goal that players pursued was to collect as many brands as possible in order to get to the next level. All participants reached the last level. At the end of the game, participants could record their name in a high score list. After playing the game, each participant filled out an electronic questionnaire (see S1 Appendix).

Data was analyzed with SPSS (see S1 Data). Analyses of variance and t-tests were carried out. In order to determine the practical and theoretical relevance of an effect as well as the power of the analyses, effect sizes were estimated for each analysis [93]. Partial eta-squared (η²_p) was used for the effect size measurement for the analysis of variance (ANOVA). It measures the strength of the effect on a continuous field, where η²_p = 0.01 indicates a small effect, η²_p = 0.06 indicates a medium effect and η²_p = 0.14 indicates a strong effect [94–96]. Hedges’ g (g_Hedges) was used to evaluate the effect of group differences of t-tests. Hedges’ g accounts for different group sizes and also allows for smaller group sizes [97–100]. Values of g_Hedges = 0.2 indicate a small effect, g_Hedges = 0.5 a medium effect, and g_Hedges = 0.8 a large effect. Phi (ϕ) was used to measure effect size for the chi-squared test, whereas ϕ = 0.1 indicates a small effect, ϕ = 0.3 indicates a medium effect and ϕ = 0.5 indicates a large effect [100,101].

Pre-test

Before conducting the main study, a first pre-test was carried out to develop and to test the video game in all three conditions (n = 10 students). The aim of the pre-test was to ensure that the game was neither too difficult nor too easy to play, that the brand positions were appropriate and not annoying, and that the questionnaire was comprehensible and not too long. None of the participants had any prior video game experiences in 3D or VR. Three subjects played the 2D video game, four people played the 3D video game and three subjects played the VR video game. Minor adaptions were made in the video game and in the questionnaire based on students’ feedback. All ten students were of the opinion that the programed game was of high quality and fun to play, regardless of the technology they had played. None of the students mentioned any concerns regarding playing the game or filling in the questionnaire.

Participants

237 students from a midsize university in Europe participated in the main study, held in computer labs on the campus. Participants were randomly assigned to one of the three game conditions (2D, 3D, VR). Three participants were excluded from the analysis because of an extremely short answer time and/or inconsistent answer patterns (e.g. flatliners, contradictions), resulting in 234 usable respondents (2D: 79 subjects; 3D: 78; VR: 77). We used a student sample, because most video game players are aged 18–49 and students are in the main target group for a jump’n’run video game like the one at hand [102,103]. Students received extra credit in a course in exchange for their participation; in addition, chocolate bars were given as small incentives as well as the option to win a voucher for a local shopping center. Respondents were between the ages of 18–46, with a mean age of 24.52 years (SD = 4.13). Age distribution did not differ between the three conditions (2D, 3D, VR) (F(2,230) = 2.165; ns, η²_p = .018). Females (n = 134) were slightly overrepresented in the study in comparison to males (n = 100), but gender distribution did not differ between the three conditions (χ² = 1.826_{(df = 2)}, ns, ϕ = .088). 48.3% of the participants played video games at least once per month with an average playing time of 22.29h per month (SD = 33.04). Men (M = 31.26h per month, SD = 40.19) played video games almost three times as often as women (M = 11.34h per month, SD = 15.62) (t = -4.272_(117,205), p<0.01, g_Hedges = .63), but average monthly playing time did not differ between the three conditions (F(2,157) = .041; ns, η²_p = .001).

Measurement of variables

Measurement of the variables was based on existing literature (see S2 Appendix). All interval scaled items had a “no answer” category as an alternative. Demographic data had to be provided (e.g. gender, age, field of study) at the end of the questionnaire. Data were analyzed using SPSS version 22 (IBM Corp, Armonk, NY, USA) (see S1 Data and S2 Data).

Attitude towards the game (α = .864) was measured with six bipolar adjectives, using 7-point semantic differential scales, with the negative adjectives coded 1 and the positive adjectives coded 7. The adjectives were: unappealing-appealing, unpleasant-pleasant, dull-dynamic, unattractive-attractive, not enjoyable-enjoyable, and depressing-refreshing [104]. The mean value of the items was calculated and used for further analyses.

Attitude towards the brand (α_nachos = .916; α_{chocolate_bar} = .932; α_smartphone = .930; α_{energy_drink} = .951): In order to avoid an overly long questionnaire, attitude toward the brand was measured for only four out of the eight brands that were placed in the game. Subjects were asked to evaluate the brands based on the following six items, using a 7-point scale: this is a bad / good product, I feel negative / positive toward the product, the product is awful / nice, the product is unpleasant / pleasant, the product is unattractive / attractive, I disapprove / approve of the product. The questions were adapted from Shamdasani et al. [105]. The mean value of the items was calculated and used for further analyses.

Arousal (α = .846) was measured immediately after playing the video game. Subjects were asked to indicate their perceived level of arousal based on three bipolar adjectives and on a 7-point semantic differential scale, hence, a self-reported arousal measurement was used. The items read excited-calm, stimulated-relaxed and alerted-soothed [56]. The mean value of the items was calculated and used for further analyses.

Presence was assessed by asking about the level of agreement to the following statement: “I was totally absorbed in what I was doing” [106]. The answer scale ranged from 1 (low agreement) to 7 (full agreement). Hence, the current research applied a single item measurement of presence. A single item measurement was chosen for practical reasons to avoid an overlong questionnaire. In addition, the level of absorption in what somebody is doing in an environment is an important indicator for the individuals’ presence in that environment. If individuals’ presence in an environment is high, they are able to adapt their actions to the environment [52] and individuals do not necessarily perceive the state of being in the virtual environment [53].

Brand recall was measured by asking the participants to write down the remembered brands, which they had seen while playing the game.

To measure brand recognition each of the eight brands that had appeared in the game was presented along with two mock brands from the same product category which had not appeared in the game. Participants had the possibility to choose “others”, if they felt they could not remember any of the three brands. The eight brands with their respective mock brands were presented one after the other, and participants could only tick one option. This option was either the correct brand (true memory) or not (i.e. false or no memory). If the participants just guessed, there would be a 25% probability of choosing the brand that appeared in the game for each brand category, which would be same for all conditions. We deliberately chose to measure recognition for each brand one after the other, instead of presenting a list with all appearing brands and the mock brands at once. Our measurement avoids the problem that is related to presenting a long list of all appearing brands and mock brands at once, namely that memory is easily overestimated if participants just start ticking many alternatives (and by doing so have the chance to hit the correct brands, too, even though they do not recognize them). However, our measurement does not allow for a recognition sensitivity test as suggested by e.g. Grier [107]. Brand logos were presented in random order for each participant to avoid order and context effects. 1 was coded if a subject named or ticked the correct brand, otherwise 0 was coded.

For both, recall and recognition, the named or ticked brands were added up to an 8-point sum scale for each participant.

Presence in the 2D, 3D and VR video game

H1 postulates a higher level of presence in the VR condition than in the 3D condition than in the 2D condition. An ANOVA with the three different technologies as independent variable and presence as dependent variable revealed that the three groups differed significantly in their presence (F(2,228) = 5.104, p<0.01, η²_p = .043). Results show that the mean was lowest in the 2D condition (M = 5.28; SD = 1.71), followed by the mean of the 3D condition (M = 5.63; SD = 1.38). The mean was highest in the VR condition (M = 6.03, SD = 1.21). Contrast tests show that significant differences were found between the 2D and VR condition (t = 3.13_(139.19), p < .01, g_Hedges = .499). The differences between the 2D and 3D condition (t = -1.40_(147.24), ns, g_Hedges = .225) as well as between the 3D and VR condition (t = -1.87_(147.95), .063, g_Hedges = .300) were as expected, but were not significant. Hypothesis H1 was partly supported by the data.

Arousal while playing the video game

Regarding the second hypothesis that arousal will be higher in the VR video game than in the 3D than in the 2D video game, an ANOVA with technology (2D, 3D, VR) as independent variable and arousal as dependent variable revealed no significant differences between the three different technological conditions (F(2,228) = .984, ns, η²_p = .009). Players in all three technology groups were moderately aroused (2D: M = 4.18, SD = 1.57; 3D: M = 4.25, SD = 1.60; VR: M = 3.90; SD = 1.63). H2 is rejected by the data.

Attitude towards the video game

In order to answer the research question whether attitudes toward the game differ in the 2D, 3D and VR condition, an ANOVA with technology (2D, 3D, VR) as independent variable and attitude towards the game as dependent variable was carried out. The analysis showed no significant differences between the three technologies (F(2,226) = .365, ns, η²_p = .003). The mean values in all three technology groups were relatively high, indicating that the players liked the video game, regardless of the technology they had played (2D: M = 5.08, SD = 1.09; 3D: M = 5.24, SD = 1.16; VR: M = 5.16; SD = 1.08).

Attitude towards the brands placed in the video game

To address the third hypothesis that attitude toward the placed brands will be higher in the VR video game than in the 3D than in the 2D video game, four consecutive ANOVAs, one for each brand in the four product categories, were performed (independent variable = technology, dependent variable = attitude towards the brand). Results of the ANOVAs revealed that the attitude towards the placed brands did not differ among the three technology groups, for any of the four brands (nachos: F(2,224) = .437, ns, η²_p = .004; chocolate bar: F(2,226) = 1.340, ns, η²_p = .012; smartphone: F(2,225) = 1.186, ns, η²_p = .010; energy drink: F(2,227) = .748, ns, η²_p = .007).

Memory for the placed brands

Hypothesis H4 expected that recall and recognition of the brands included in the video game will be lower in the VR condition than in the 3D condition than in the 2D condition.

Cognitive load

Cognitive load was measured to test our assumption made in the theoretical part that participants playing the video game in the 3D and VR condition need more cognitive capacity for playing the game, as compared to participants in the 2D condition. Cognitive load refers to the total amount of mental effort that is used in the working memory when performing a learning task, it is “the manner in which cognitive resources are focused and used during learning” ([110]; p. 294). To measure cognitive load, participants in all conditions were asked to memorize the same 8-digit number prior to playing the video game [111]. After playing the video game, participants were asked to recall as many digits as possible of the number they had been given. A lower number of digits that can be memorized and recalled result from higher cognitive load while playing the game.

The results showed that the mean of correctly recalled digits was indeed lowest in the VR condition (M = 3.47; S = 1.06). However, other than expected, the mean of the 2D condition (M = 4.00, SD = 1.56) was lower than the mean in the 3D condition (M = 4.42, SD = 1.56). Hence, cognitive load while playing the video game appeared to be highest in the VR condition and lowest in the 3D condition. While an ANOVA with the three conditions as independent variable did not reveal significant differences between the three groups (F (2,50) = 2.50, ns, η²_p = .091), post hoc t-tests between two groups revealed that the mean of the number of correctly recalled digits in in the 3D and VR condition differed significantly (t = -2.236₍₅₎, p = .03, g_Hedges = .696). There were no differences between male and female participants (t = .233₍₅₁₎, ns, g_Hedges = .074).

Brand memory (recall and recognition)

Recall and recognition of brands were measured in the same way as in the main study. The results of the post hoc study confirmed the findings of the main study. With regard to recall, the three technologies differed significantly (Recall: F(2,50) = 5.237, p < .01, η²_p = .181). Recall was lowest in the VR condition (M = .40, SD = .91), followed by the mean in the 3D condition (M = .63, SD = 1.07), and highest in the 2D condition (M = 1.58, SD = 1.30). Significant differences between 2D and 3D (t = 2.609₍₅₀₎, p = .012, g_Hedges = -.798) and 2D and VR (t = -3.049₍₅₀₎, p < .01, g_Hedges = 1.03) were revealed. The results for brand recognition were similar (F(2,50) = 5.479, p < .01, η²_p = .180). Recognition was lowest in the VR condition (M = 3.00, SD = 1.73), followed by the 3D condition (M = 4.52, SD = 1.71), and highest in the 2D condition (M = 4.79, SD = 1.55). Post hoc t-tests revealed significant differences between 3D and VR (t = 2.661₍₅₀₎, p < .01, g_Hedges = .884) and 2D and VR (t = -3.230₍₅₀₎, p < .01, g_Hedges = 1.097). Values for brand recognition were again much higher than values for brand recall.

Dizziness and motion-sickness while playing

Perceived dizziness and motion sickness while playing the game were also measured. Dizziness was measured with the question “Did you feel dizzy when playing the video game”, motion-sickness with the question “Did you become motion sick while playing the video game?”. The questions were adopted from Jones et al. [112] and Merhi et al. [113]. The answer scale ranged from 1 (not at all) to 7 (very much).

Subjects reported significantly different levels of dizziness that they had experienced during game play (F(2,50) = 10.265, p < .01, η²_p = .291). Dizziness was significantly higher while playing the VR game (M = 2.60, SD = 1.55) than while playing the 3D game (M = 1.47, SD = .84; t = 2.536_(20.416), p = .019, g_Hedges = .939) and the 2D game (M = 1.11, SD = .32; t = -3.677_(14.918), p < .01). No significant differences were found between playing the 2D and 3D game (t = 1.788_(22.960), ns, g_Hedges = .566).

Results for motion-sickness were similar (F(2,50) = 12.359, p < .01, η²_p = .331). Subjects reported significantly higher levels of motion-sickness in the VR game (M = 2.40, SD = 1.64) than in the 3D game (M = 1.05, SD = .229; t = 3.160_(14.434), p < .01, g_Hedges = 1.229) and the 2D game (M = 1.05, SD = .229; t = -3.160_(14.434), p < .01, g_Hedges = 1.229). Again, no significant differences were found between 2D and 3D (t = .000_(36.000), ns, g_Hedges = 0).