Bachelorproject Brein & Cognitie
Universiteit van Amsterdam 27-05-2016
Naam: Hidde Kamst
Studentnummer: 10359737 Begeleider: Jasper Winkel Aantal woorden: 5008
The effects of realism of effort, realism of reward and individual
differences in extraversion on effort-based decision making: a study
in virtual reality.
Abstract
The goal of this study was to examine if realism of effort, realism of reward and individual differences in extraversion influence presence and the amount of expended physical effort by humans. Fifty-one participants completed an effort-based decision making task in virtual reality in which both the realism of effort and the realism of reward were manipulated compared to a baseline condition. Participants also completed the Dutch NEO-FFI extraversion scale and an adjusted version of the Igroup Presence Questionnaire. An effect of both realism of effort and realism of reward on presence was found. No effect of extraversion was found. Also, no effects of realism of effort, realism of reward and extraversion on effort expenditure were found. Many subjects were excluded due to ceiling effects, which was a major limitation. The results seem promising for the use of VR in
research. Much more studies have to be done to overcome limitations however.
Introduction
Cooking a fancy dinner might include going to multiple grocery stores and spending a lot of time in the kitchen, while making a pizza only includes going to the supermarket across the street and placing it in the oven. Decisions like these consist of weighting costs and benefits. To get a certain reward one needs to expend a certain amount of effort. This type of decision making is called effort-based decision making (EBDM).
If the reward between two actions is equal, then people tend to choose the action that will minimize the amount of physical or cognitive effort (Kool, McGuire, Rosen, & Botvinick, 2010). As the previous example illustrates, however, rewards between different actions are rarely equal and choosing high effort actions often results in a higher reward. Making the optimal choice based on effort and reward is not always straightforward because it can depend on many factors. To study the mechanisms underlying these choices, EBDM tasks have been developed for both human and other animals. In EBDM tasks, a choice is made between high effort/high reward (HH) and low effort/low reward (LL) options. The rewards mostly consist of food in animal experiments and money in human
experiments. The effort can be of physical and cognitive nature. An example of a physical EBDM task is the Effort-Expenditure for Rewards Task (“EEfRT”) (Treadway, Bossaller, Shelton, & Zald, 2012). In this task participants have the choice to make many or few button presses. Participants earn more money by making many presses.
In the field of neuroscience, several studies investigated neural mechanisms underlying EBDM. Monkeys and rats, for example, are able to make optimal decisions based on effort and reward, while rats with lesions in the anterior cingulate cortex (ACC) and medial frontal cortex (MFC) can’t (Walton, Kennerley, Bannerman, Phillips, & Rushworth, 2006). In a review, Kurniawan, Guitart-Masip, and Dolan (2011) present evidence of fMRI activity in humans in the striatum (Croxson, Walton, O’Reilly, Behrens, & Rushworth, (2009); Kurniawan et al., 2010) and ACC (Walton, Devlin, & Rushworth, (2004)) during EBDM tasks. They also present evidence of a link between dopamine and EBDM by discussing several experiments where depletion of dopamine led to less willingness to expend effort in rats (Denk et al., 2005; Phillips, Walton, & Jhou, (2007)). In the clinical field EBDM has been linked to anhedonia, intrinsic motivation and negative symptoms (Treadway et al., 2012; Horan et al., 2015). EBDM has also been suggested to be an adequate measure of motivation in patients with a major depressive disorder. Combined, these studies shed light on the way we make decisions and may have implications for the clinical field.
Previous EBDM experiments concerning humans have two major drawbacks, namely time differences and the realism of effort in EBDM tasks. In many of the EBDM tasks, like the EEfRT, HH choices are more time consuming than LL choices. Humans are biased toward less time consuming choices (Bogacz et al., 2006), which possibly makes them more likely to make a LL choice. Time differences also create a difference in the possible reward earned per minute. Because HH choices are more time consuming, the reward differences between HH and LL choices get relatively smaller. EBDM tasks that control for the amount of time a choice costs should therefore be developed. Another drawback is the realism of effort in EBDM tasks. EBDM tasks, up to now, consist of
conventional psychological tests. The realism of conventional psychological tests is generally low, and while they give experimenters high experimental control, they often lack ecological validity which lowers generalisability (Araujo, Davids, & Passos, 2007).
A possible way to overcome the low ecological validity of EBDM tasks is the use of virtual reality (VR). VR combines ecological validity and experimental control by offering highly realistic environments that can be controlled (Bohil, Alicea, & Biocca, 2011). Two important factors of VR are immersion and presence. ‘Immersion refers to the objective level of sensory fidelity a system provides, presence refers to a user’s subjective psychological response to a system’ (Bowman & McMahan, 2007). Presence is the sense of being in the virtual environment instead of the physical environment and is the result of immersion. Both immersion and presence are generally high in VR,
which can have an impact on the results of an experiment. As Bohil et al. (2011) state: ‘By providing realistic stimulation to multiple sensory channels at once, VR engages the sensorimotor system more fully than simple stimuli used in most psychological research, increasing the potential to elicit
realistic psychological and behavioural responses.’ VR could therefore be a more realistic alternative to conventional psychological tests in EBDM tasks, since it offers the possibility to frame a decision in a realistic manner. By making the effort more realistic during an EBDM task, the effort might be perceived as being heavier, which would lead to less effort expenditure. By making the reward more realistic during an EBDM task, the reward might be perceived as being more attractive, which would lead to more effort expenditure.
Another interesting subject concerning EBDM is individual differences in personality. As noted above, several correlations have been found between personality differences and EBDM. Extraversion might be of particular interest. It has been shown, for example, that extraverts invest more effort in a task (Beauducel, Brocke, & Leue, 2006). Extraversion might therefore be an
influencing factor in EBDM. In a fMRI study, extraverts exhibited more activity in their reward system (figure 1), including the nucleus accumbens, during a decision making task (Cohen, Young, Baek, Kessler, & Ranganath, 2005). These findings suggest that extraverts are more sensitive to rewards and might therefore be more willing to expend effort to get it. In rats, dopamine depletion of the nucleus accumbens leads to a reallocation of the rats away from high effort towards low effort behaviour (Salamone, Correa, Farrar, & Mingote, 2007). More activity in the nucleus accumbens might therefore lead to more high effort behaviour in an EBDM task. Extraversion has also been shown to positively relate to immersive tendency (Weibel, Wissmath, & Mast, 2010). Immersive tendency can be described as the capability of an individual to be immersed or involved in a virtual environment (Witmer & Singer, 1998), and is used to measure if someone easily experiences presence. These results suggest that extraverts might more easily experience presence in VR than introverts, and might therefore be more sensitive to the effects of VR. Framing a decision in a realistic manner might therefore have a larger effect on extraverts.
Figure 1. Cohen et al. (2005) found a correlation between extraversion and activation in the left- and
right orbitofrontal cortex and in the nucleus accumbens.
The goal of the current study was to examine if realism of effort, realism of reward and individual differences in extraversion influence presence and the amount of expended physical effort by humans. There were three conditions in the experiment; a baseline condition, a realistic effort (RE) condition and a realistic reward (RR) condition. In the RE condition the realism of effort was adjusted compared to the baseline condition. In the RR condition the realism of reward was adjusted compared to the baseline condition. All other factors were kept constant. A positive effect of realism of effort and realism of reward on presence was expected. Extraversion was also expected to have a positive effect on presence. Realism of effort was expected to have a negative effect on the amount of expended effort, and realism of reward was expected to have a positive effect on effort
expenditure. Finally, extraversion was expected to have a positive effect on the amount of expended effort. This effect was expected to be smaller in the RE condition and larger in the RR condition, compared to the baseline condition (figure 2).
Figure 2. A visualisation of the hypotheses concerning extraversion and effort expenditure. More
effort expenditure is expected in the RR condition, and less in the RE condition, compared to the baseline condition. Effort expenditure is expected to be higher in extraverts. This difference is expected to be larger in the RR condition and smaller in the RE condition.
Methods
Participants
Fifty-one persons participated in the experiment. Of the participants, 37% were female, 80% percent were students and the average age was 23 (SD = 3.22) years. The participants signed an informed consent that was approved by the ethical commission of the University of Amsterdam. People with carrousel sickness or another first langue than Dutch could not participate in the experiment. This exclusion criteria were used to prevent nausea and language effects. Participants were not paid for their attendance, but could win payment of the average amount of coins earned. One participant was awarded this money.
Measurement materials
Extraversion. The Dutch version of the NEO Five-Factor Inventory (NEO-FFI) extraversion
scale (McCrae & Costa, 2004; Hoekstra, Fruyt, & Ormel, 2007) was used to assess the participants’ degree of extraversion. The Dutch version was chosen to avoid effects of language. The NEO-FFI extraversion scale contains 12 items. The items are scored on a five-point Likert scale, ranging from 1 (‘I strongly disagree’) to 5 (‘I strongly agree’). An example item from the NEO-FFI is ‘I like talking to people’. RR Baseline RE Ef fo rt e xp en di tu re Condition Extraversion Low Extraversion High
Presence. The Igroup Presence Inventory (IPQ) (Schubert, Friedmann, & Regenbrecht, 2001)
was used to assess the participants’ sense of presence during each condition. The questions were translated to Dutch and modified to fit the experiment. The modified version contains 13 items for each condition used in the experiment. Each item is scored on a seven-point Likert scale, ranging from 1 (‘not at all true’) to 7 (‘completely true’). An example item from the IPQ is ‘I had the feeling that the virtual world was real during...’, for each of the three conditions. As a manipulation check, two questions were added to the IPQ. These questions were ‘I had the feeling I had to work hard during..’ and ‘I found the reward attracting during.. ’, for each of the three conditions. The questions were scored on the same scale as the IPQ. These questions were a manipulation check for perceived effort and reward attractiveness, respectively.
Effort expenditure. To measure effort expenditure, a video game was developed(see the appendix for a more detailed description of the development process) containing the experiment. The program was designed using Unreal Engine 4, which is a suite of game development tools that is commonly used to make video games and has a plugin for VR. The Oculus Rift DK2 was used as VR headset. The experiment (figure 3) consists of powering a mine cart over a track by making pumping motions with a bicycle pump. At the beginning of each trial a choice is made between a HH route or a LL route. Colour-coding in presentation of the tracks gives information about the amount of effort a route requires; green sections of the track require no pumping input, orange sections of the track require medium effort and red sections of the track require high effort pumping. The different route options and coin rewards are displayed on two different computer screens, one on the left side of the virtual room and one on the right.
In the baseline condition the coins are displayed abstractly on these screens as stacked orange bars. After choosing a route with a mouse click, a third display in the middle of the room shows a power bar and the progress within the chosen track. Participants are only able to track their progress on this screen, as no visible cart moves. In the RR condition, the environment and
procedure are the same as in the baseline condition. The rewards are, however, realistically represented as stacks of golden coins on the left and right side of a desk in front of the participant and no longer on the computer screens. In the RE condition the reward is the same as in the baseline condition, but the environment is different. After selecting a route, a large door opens and the cart drives into a natural environment. Some sections of the outside tracks are overgrown with either grass (medium effort) or shrubs (large effort). For a video demonstration, see the following link https://www.youtube.com/watch?v=B87NGwa5jlU.
Figure 3. A visualization of the experiment. Figure 4A shows the baseline condition, where the
subjects sit behind a table. The reward is represented by orange bars and the effort is represented by two different tracks. The chosen track will be displayed on the screen in the middle. Figure 4B shows the RR condition, where the reward is represented by golden coins that stand on the table and go into the box when earned. Figure 4C shows the RE condition, where the subjects drive through an environment and the effort is represented by grass and shrubs.
To mimic the motion made by the cart, a bicycle pump was chosen as input device. Operating a bicycle pump resembles the motion one would make by operating a handcar. Furthermore, the air resistance felt while operating the bicycle pump makes its usage an effortful activity. In order to make this bicycle pump an appropriate input device for a computer, a strip of aluminium was attached to the handle of the pump. This strip of aluminium covers the entire length of the pump. Since this strip is only attached to the handle, it goes up and down, along with the pumping motion. Over the aluminium strip, a computer mouse (Logitech G300) was fixed. Consequently, when moving the bicycle pump-handle up and down, the aluminium strip moves similarly along the fixed computer
mouse. In this manner, the computer mouse can register the motions of the pump.
To keep the pump in its place whilst being used, it was fixed to a MDF board. A chair was placed on the board, in front of the pump. In this manner the weight of the participant sitting in the chair keeps the board (and the pump fixed to the board) in its place. In this way, potential differences in tilting of the pump and participant-to-pump distance between participants are minimized.
To assess effort expenditure, the Point of indifference (POI) for each subject was determined per condition. To determine the POI value for each condition, the average values of the reward modifier of the last four trials were calculated. The reward modifier is used to calculated the reward for both tracks and is updated before each choice. For each HH choice the difference between track rewards decreases and for each LL choice the difference between track rewards increases. For each trial the reward for both tracks is calculated by subtracting the total effort values from both tracks (one for green, two for orange and four for red) and multiplying this value with the reward modifier. The outcome is then added to ten for the high effort track and subtracted from ten for the low effort track.
The value of the reward modifier ranges between 0, which means no difference between rewards, and 1.25, which is the maximum possible reward (20 coins) divided by the maximum possible difference between tracks (16). Each condition starts with a reward modifier value of 0.625, which is the maximum reward modifier value divided by two. For each HH choice a value is
subtracted from the reward modifier value and for each LL choice a value is added to the reward modifier. The added or subtracted values grow for each consecutive choice of the same effort type. This value is 0.02 for the first, 0.05 for the second, 0.1 for the third and 0.2 for the fourth or a higher consecutive choice of the same type of effort. For each switch in choice type this value drops back to 0.02. The minimum value of the reward modifier has been set to 0 to avoid that the LL choice has a higher reward than the HE choice. The maximum value of the reward modifier has been set to 1.25 to avoid scores that are higher than 20 coins.
Assigning POI values is demonstrated to be a reliable method for measuring individual differences in subjective effort (Westbrook, Kester, & Braver, 2013). For each choice the difference between rewards gets either smaller or larger. The POI is reached when the subject no longer expresses a preference for the HH or LL option. At this point the subject is satisfied with the reward gained per effort and will interchangeably choose between HH and LL choices. Figure 4 shows the reward modifier sores of three participants in one of the conditions. As can be seen, the reward modifier scores of these subjects increase or decrease up to a point where they stabilize. This stable point represents a participants’ POI.
Figure 4. reward modifier scores for three subjects in one of the conditions. The trial number is
shown on the x-axis and the reward modifier score is shown on the y-axis. The dotted line represents the point from where the POI was measured.
Procedure
There were two different testing rooms, one for the experiment and one for the
questionnaires. An informed consent was signed at the start. In the experiment the participants completed the baseline condition, RE condition and RR condition in counterbalanced order. Each condition consisted of 13 trials. Before the start of the experiment, participants were given instructions and completed at least one test trial for each condition to make sure they understood the amount of effort the different track colours represented, knew how to operate the cart and knew where to look for effort, reward and progress information. Participants could earn ten eurocents for each virtual coin they earned. Two subjects were tested simultaneously, one starting with the experiment and the other with the questionnaires. Participants therefore completed the NEO-FFI either before or after the experiment. The IPQ was always completed after the experiment, since these questions are about the experiment.
Data analyses
The first hypothesis is that that realism of effort and realism of reward have a positive effect on presence. To test this, a one-way repeated measures ANOVA will be performed comparing the IPQ scores of the three conditions. The second hypothesis is that extraversion has a positive effect on
0 0,2 0,4 0,6 0,8 1 1,2 1 2 3 4 5 6 7 8 9 10 11 12 13 Re w ard mo di fie r s co re Trial 17 baseline 7 baseline 27 RE
presence. To test this, extraversion will be added to the one-way repeated measures ANOVA as covariate. The third hypothesis is that realism of effort has a negative effect on effort expenditure, and that realism of reward has a positive effect on effort expenditure. The fourth hypothesis is that extraversion has a positive effect on effort expenditure, and that this effect is smaller in the RE condition and larger in the RR condition. To test these hypotheses, a factorial mixed ANOVA will be performed. Condition will be the within subject variable of the analysis. The extraversion scores will be split in two groups using a median split and will be used as between subject variable in the analysis. The POI scores will be the dependent variable.
Results
Fifty-one subjects participated in the experiment. Five of the subjects did not fill in the IPQ. These subjects were excluded from the analyses concerning presence. Two of the subjects did not completely finish the NEO-FFI extraversion scale and were excluded from the analyses concerning extraversion. Due to ceiling and floor effects in the POI scores, 17 subjects were excluded from the analyses concerning effort expenditure. Ceiling effects include subjects having a minimal or near minimal POI in more than one of the three conditions. Floor effects include subjects having a maximal or near maximal POI score in more than one of the three conditions. These criteria for ceiling and floor effects were maintained because when a near maximal or near minimal POI score is reached in more than one condition, nothing can be said about the difference between all three conditions. For each condition the POI score and the IPQ score were measured. The mean scores and standard deviations are shown in table 1, including the scores in original counterbalanced order. For each subject the NEO-FFI score on extraversion was measured. All analyses have been done using SPSS.
Table 1
The mean POI scores and IPQ scores per condition and the mean POI scores in original counterbalanced order. The standard deviations are shown in parentheses.
POI IPQ Mean (SD) Mean (SD) Condition .46 (.29) .48 (.38) .53 (.35) 38.00 (4.54) 38.72 (4.91) 41.07 (4.60) - Baseline - RR - RE Block .47 (.32) .45 (.32) .54 (.37) - 1 - 2 - 3
For the manipulation checks for perceived effort and attractiveness of reward, two one-way repeated measures ANOVA were performed comparing the scores on these questions. For the manipulation check of perceived effort, the assumption of sphericity had been violated, X2(2) = 8.20
p = .017, therefore Greenhouse-Geisser corrected tests are reported. No main effect of condition was
found, F(1.59, 44.37) = .35, p = .658. A main effect of condition was found for the manipulation check of reward attractiveness, F(2, 56) = 7.49, p = .001. Contrasts reveal a significant difference between the RR condition and the baseline condition, F(1, 28) = 14.49, p = .001, but not between the RR and the RE condition, F(1, 28) = 3.74, p = .063.
To test if presence was higher in the RE and RR condition compared to the baseline condition, a one-way repeated measures ANOVA was performed. The IPQ scores of the three conditions were compared against each other. The assumption of sphericity had been violated, X2(2) = 22.95 p < .001, therefore Greenhouse-Geisser corrected tests are reported. The results show a significant effect of condition, F(1.42, 63.99) = 14.89, p < .001. Contrasts reveal a significant effect of both the RE condition, F(1, 45) = 20.37, p < .001 and the RR condition, F(1, 45) = 4.11, p = .049 compared to the baseline condition (figure 5). To test the hypothesis that extraversion has a positive effect on presence, a one-way repeated measures ANOVA containing the NEO-FFI scores one
effect of extraversion, F(1, 42) = 1.25, p = .270, and no interaction effect between extraversion and presence in the different conditions, F(1, 42) = 1.15, p = .289.
Figure 5. The mean IPQ scores for the three conditions. The error bars represent the within-subject
normalized standard errors.
To test for the effect of time, a one-way repeated measures ANOVA was performed. The POI scores in the original counterbalanced order of the first, second and third block were compared. The assumption of sphericity had been violated, X2(2) = 6.43, p = .040, therefore Greenhouse-Geisser corrected tests are reported. The results show no significant effect of time on the POI scores, F(1.68, 51.98) = 1.52, p = .230. Finally, a factorial mixed ANOVA was performed to test the hypothesis that realism of effort has a negative effect on effort expenditure and that realism of reward has a positive effect on effort expenditure. The hypothesis that the POI scores in extraverts are higher in the baseline condition, this difference is larger in the RR condition and smaller in the RE condition was also tested. A median-split was done on the extraversion scores, splitting them in a high and low extraversion group. The scores from the median (40) were put in the low group (N = 17) and the other scores were put in the high group (N = 15). The POI scores of the high and low extraversion group of the three conditions were compared against each other. There was no main effect of condition, F(2, 60) = .83, p = 0.441. There also was no main effect of extraversion, F(1, 30) = .33, p = 0.568. Lastly, there was no significant interaction effect between extraversion and condition, F(2, 60) = .22, p = .800. Figure 6 shows the different POI means for high and low extraverts.
35 36 37 38 39 40 41 42 43 RR Baseline RE
Mean IPQ score
Figure 6. The mean POI scores of the high and low extraversion group for the three conditions. The
error bars represent the within-subject normalized standard errors.
Discussion
The current study examined if realism of effort, realism of reward and individual differences in extraversion influence presence and effort expenditure in EBDM tasks, using VR. Effects of realism of effort and realism of reward on presence were found, supporting the hypothesis that realism of effort and realism of reward have a positive effect on presence. No effect of extraversion on
presence was found, however, contradicting the hypothesis that extraversion has a positive effect on presence. No effect of realism of reward on effort expenditure and no interaction effect between realism of effort and extraversion were found, contradicting the hypothesis that realistic effort is perceived as heavier in general and even more so by extraverts. Also, no support was found for the hypothesis that realism of reward has a positive effect on effort expenditure and even more so in extraverts. Lastly, no effect of extraversion on effort expenditure was found, contradicting the hypothesis that extraversion has a positive effect on effort expenditure.
The main hypothesis of the current study, which stated that realism of effort has a negative effect on effort expenditure and that realism of reward has a positive effect on effort expenditure, was not supported. The manipulation for realism of reward, nevertheless, seems to have succeeded. The participants rated the reward attractiveness the highest in the RR condition. Also, there was an effect of realism of reward on presence. Possibly, attractiveness of reward alone in EBDM is not
0,2 0,3 0,4 0,5 0,6 0,7 RR Baseline RE
Mean POI scores
Extraversion Low Extraversion High
enough, and an incentive is also needed. To test this, future research should be done examining the relation between realism of reward, reward incentive and effort expenditure. The manipulation of realism of effort also seems to have succeeded, since there was an effect of realism of effort on presence. Effort was, however, not perceived as being heavier in the RE condition. This finding, together with the finding that realism of effort has no effect on effort expenditure, contradicts the idea that realism of effort has an influence on EBDM tasks.
The findings concerning extraversion are not in line with previous studies. Weibel et al. (2010), for example, found that extraversion positively relates to immersive tendency, which is a measure if someone easily experiences presence. In the current study no relation between presence and extraversion was found. Immersive tendency hasn’t been measured directly however, and while immersive tendency and presence seem closely related they don’t represent the same construct. Future research should therefore examine the relationship between immersive tendency, presence and extraversion. Cohen et al. (2005) found that extraverts exhibit more activity in their reward system than introverts, suggesting that extraverts are more sensitive to reward. In the current study however, no relation was found between extraversion, realism of reward and effort expenditure. Lastly, Beauducel et al. (2006) found a positive effect of extraversion on effort expenditure in an experimental task. In the current study, no effect of extraversion was found.
A possible reason for the absence of any effects of extraversion is the high mean score on extraversion, which was 40.43. In their review, McCrae and Costa (2004), find the means on extraversion of 30.58 and 28.50 in two samples. This is a difference of 9.85 and 11.97 respectively, which is a rather large difference on a questionnaire with scores ranging from 0 to 60. There are several possible reasons for the high scores on extraversion. First, the participants might almost all have been high extraverts, which seems unlikely. It might also be due to reliability problems of the NEO-FFI extraversion scale. The extraversion scale of the NEO-FFI, however, has proven to be the most internally consistent scale of the questionnaire (McCrae & Costa, 2004). The Dutch version has an average Cronbach’s alpha value of .77 (Hoekstra et al., 2007). Whatever the reason for the high extraversion scores is, it presents a possible confound in the analyses concerning extraversion. More research on extraversion and EBDM should therefore be done, perhaps using multiple questionnaires on extraversion.
Another limitation of the experiment was the high amount of ceiling effects in the experiment. Fifteen of the participants had to be excluded due to ceiling effects, which is 29.4 percent of the data. These ceiling effects might have obscured effects. There are several possible explanations for the high percentage of ceiling effects. First, the effort from the input device might have been too low. The ceiling effects were due to participants consequently making HH choices. By increasing the effort, these ceiling effects should be reduced. Increasing the effort might, however,
also lead to more floor effects, which also leads to exclusion. A second explanation is social desirable behaviour in participants. A high percentage of the participants were friends of the experimenters. These participants might have felt pressured to make HH choices, even though it was explicitly stated that they were free to make their own choice. Future research should therefore not use participants that are known to the experimenters. The small amount of trials in each condition might also have caused the ceiling effects. Due to the way the reward modifier was set, it was quite easy to reach the maximal or minimal value by making a couple of consecutive choices of the same type. Once the maximal value is reached and a person decides to switch, however, it takes quite some time to get a new reward modifier value, which might take too long in 13 trials. A possible solution is the use of two conditions with 20 trials in further research.
In summary, no effects of realism of effort, realism of reward and extraversion on effort expenditure in EBDM tasks were found. Realism of effort and realism of reward did improve presence in VR, showing that realism of effort and realism of reward did have an effect. With the swift technological advancement of this era, the use of VR seems promising for science. By
overcoming the limitations of the current VR research, which is still in its infancy, VR can improve the ecological validity of the EBDM paradigm, along with other fields of science. Nevertheless, a large quantity of research will have to be done to overcome these limitations.
Literature
Araujo, D., Davids, K., & Passos, P. (2007). Ecological validity, representative design, and correspondence between experimental task constraints and behavioral setting: Comment on
Rogers, Kadar, and Costall (2005). Ecological Psychology, 19(1), 69-78.
Beauducel, A., Brocke, B., & Leue, A. (2006). Energetical bases of extraversion: Effort, arousal, EEG, and performance. International Journal of Psychophysiology, 62(2), 212-223
Bogacz, R., Brown, E., Moehlis, J., Holmes, P., & Cohen, J. D. (2006). The physics of optimal decision making: A formal analysis of models of performance in two-alternative forced-choice tasks.
Psychological Review, 113, 700–765Bohil, C. J., Alicea, B., & Biocca, F. A. (2011). Virtual reality in neuroscience research and therapy. Nature reviews neuroscience, 12(12), 752-762.
Bowman, D. A., & McMahan, R. P. (2007). Virtual reality: how much immersion is enough?. Computer, 40(7), 36-43.
Cohen, M. X., Young, J., Baek, J. M., Kessler, C., & Ranganath, C. (2005). Individual differences in extraversion and dopamine genetics predict neural reward responses. Cognitive Brain
Research, 25(3), 851-861.
Croxson, P. L., Walton, M. E., O'Reilly, J. X., Behrens, T. E., & Rushworth, M. F. (2009). Effort-based cost–benefit valuation and the human brain. The Journal of Neuroscience, 29(14), 4531-4541.
Denk, F., Walton, M. E., Jennings, K. A., Sharp, T., Rushworth, M. F. S., & Bannerman, D. M. (2005). Differential involvement of serotonin and dopamine systems in cost-benefit decisions about delay or effort. Psychopharmacology,179(3), 587-596.
Hoekstra, H.A., Fruyt, F. de & Ormel, J. (2007). NEO-PI-R en NEO-FFI persoonlijkheidsvragenlijsten:
Handleiding. Amsterdam: Hogrefe Uitgevers B.V.
Horan, W. P., Reddy, L. F., Barch, D. M., Buchanan, R. W., Dunayevich, E., Gold, J. M., ... & Green, M. F. (2015). Effort-based decision-making paradigms for clinical trials in schizophrenia: part 2— external validity and correlates. Schizophrenia bulletin, sbv090.
Kool, W., McGuire, J. T., Rosen, Z. B., & Botvinick, M. M. (2010). Decision making and the avoidance of cognitive demand. Journal of Experimental Psychology: General, 139(4), 665.
Kurniawan, I. T., Seymour, B., Talmi, D., Yoshida, W., Chater, N., & Dolan, R. J. (2010). Choosing to make an effort: the role of striatum in signaling physical effort of a chosen action. Journal of
Neurophysiology, 104(1), 313-321.
Kurniawan, I. T., Guitart-Masip, M., & Dolan, R. J. (2011). Dopamine and effort-based decision making. Frontiers in neuroscience, 5.
McCrae, R. R., & Costa, P. T. (2004). A contemplated revision of the NEO Five-Factor Inventory. Personality and individual differences, 36(3), 587-596.
Phillips, P. E., Walton, M. E., & Jhou, T. C. (2007). Calculating utility: preclinical evidence for cost– benefit analysis by mesolimbic dopamine. Psychopharmacology, 191(3), 483-495.
Salamone, J. D., Correa, M., Farrar, A., & Mingote, S. M. (2007). Effort-related functions of nucleus accumbens dopamine and associated forebrain circuits. Psychopharmacology, 191(3), 461-482.
Schubert, T., Friedmann, F., & Regenbrecht, H. (2001). The experience of presence: Factor analytic insights. Presence, 10(3), 266-281.
Treadway, M. T., Bossaller, N. A., Shelton, R. C., & Zald, D. H. (2012). Effort-based decision-making in major depressive disorder: a translational model of motivational anhedonia. Journal of
abnormal psychology, 121(3), 553.
Walton, M. E., Devlin, J. T., & Rushworth, M. F. (2004). Interactions between decision making and performance monitoring within prefrontal cortex. Nature neuroscience, 7(11), 1259-1265.
Walton, M. E., Kennerley, S. W., Bannerman, D. M., Phillips, P. E. M., & Rushworth, M. F. (2006). Weighing up the benefits of work: behavioral and neural analyses of effort-related decision making. Neural networks, 19(8), 1302-1314.
Weibel, D., Wissmath, B., & Mast, F. W. (2010). Immersion in mediated environments: the role of personality traits. Cyberpsychology, Behavior, and Social Networking, 13(3), 251-256.
Westbrook, A., Kester, D., & Braver, T. S. (2013). What is the subjective cost of cognitive effort? Load, trait, and aging effects revealed by economic preference. PLoS One, 8(7), e68210.
Witmer, B. G., & Singer, M. J. (1998). Measuring presence in virtual environments: A presence questionnaire. Presence: Teleoperators and virtual environments, 7(3), 225-240.
Appendix
For the experiment, a previous version of the experiment was updated. Several things were adjusted and some new things were implemented. The effort expenditure in the previous
experiment, coming from a joystick, was too low. We therefore decided to change the input device to a bicycle pump. We also introduced the reward condition, to test if VR causes enjoyment and therefore leads to more effort expenditure. In the previous experiment only one condition was in VR, which could be a source of bias. In the current experiment all the conditions are in VR. We also changed the scoring range of the game. In the previous version the scores ranged from 0 to 2000, which we set to range from 0 to 20. Because of this change in scores we also needed to change the reward modifier. Besides the range of the scores, the visualisation of the scores was also changed from a number into a stack of coins. Lastly, the graphics of the game were changed.
was written. UE4 is a game developers programme that uses a visual programming language called blueprint. We first needed to learn how to use blueprint. After that we spent a week and some extra days in the experimenting room to implement all the changes. I personally changed the scoring range, the reward modifier and worked on the standard settings of the experiment.
Reflectieverslag
FeedbackUit de feedback op mijn inleiding en data-analyse kwam vooral naar voren dat ik af en toe kernzaken onderbelicht laat en onnodige informatie juist aanhaal, en dat ik vaak deelzinnen op een rare manier aan elkaar koppel. Daarnaast waren er nog een paar kleine puntjes zoals het onduidelijk uitleggen van bepaalde concepten en één van de hypothesen, en het foutief interpreteren van eerder onderzoek. Ik was het grotendeels eens met deze feedback. Ik ben begonnen met aanpassen van kleine puntjes (concepten duidelijker uitleggen etc.), heb vervolgens wat minder relevante informatie weggehaald (misplaatste maatschappelijke relevatie en wat eerder onderzoek) en ben vervolgens mijn hypothesen beter gaan onderbouwen. Het beter onderbouwen van mijn hypothesen over extraversie ging redelijk makkelijk, maar ik merkte dat ik meer moeite had met het
onderbouwen van de hypothese dat realisme van effort een negatieve invloed heeft op effort
expenditure en dat realisme van reward een positieve invloed heeft op effort expenditure. Want
hoewel deze hypothese voor mij heel logisch klinkt, kon ik geen wetenschappelijk onderzoek vinden die dit onderbouwt. Uiteindelijk heb ik het na overleg met Jasper op een framing effect gegooid. Ik had het echter liever onderbouwt met wetenschappelijk onderzoek.
Uit de feedback op mijn eerste versie kwam vooral naar voren dat het taalgebruik af en toe te informeel was en er nog een aantal dingen aan mijn grafieken verbeterd konden worden.
Daarnaast waren er nog een aantal punten als de opening van de inleiding, het onhandig formuleren van zinnen en vermelden naar oorspronkelijke bronnen. Ik heb me eerst gefocust op het taalgebruik, en dat aangepast waar ik kon. Vervolgens heb ik de kleine punten verbeterd. Hierbij merkte ik dat ik het punt oorspronkelijke bronnen nog wel lastig vond. Hoe ik het nu heb gedaan is dat ik eerst de review (Kurnaiwan et al., 2011) benoem en vervolgens achter bepaalde bevindingen de
oorspronkelijke bron heb gezet. Dit is wel een beetje dubbelop en ik weet niet of dit een gangbare manier is om dat te doen in de wetenschappelijke wereld. Tot slot heb ik geprobeerd alle grafieken die niet helemaal helder waren te verhelderen. Het enige waar dat niet gelukt is, is bij het omdraaien van de as van figuur 6 omdat de x-as dan raar ging doen.
Verloop project
Ik vond dat het project goed begon. De inleidende presentaties van zowel Manon als Jasper waren helder. Daarnaast vond ik de mini-presentaties die we in het begin deden (over belangrijke aspecten van ons onderzoek) een goede manier om in het onderwerp te komen. Zelf zat ik toen met wat stress van een uitgelopen deadline van mijn minor, waardoor ik het lastiger vond om in het onderwerp te komen. Hierna was er een periode dat er niet veel gebeurde en er veel onduidelijkheid was over wat men moest doen. Ik denk dat dit mede komt doordat drie weken (naar mijn mening) te lang is om naar literatuur te zoeken. Daardoor stelt iedereen het uit tot het laatste moment en gebeurt er de eerste twee weken vrijwel niets. Het lijkt me dan ook een goed idee om deze periode te verkorten naar twee weken.
Vervolgens kwam het ontwikkelen van het experiment en het schrijven van de inleiding en data-analyse. Dit verliep hectisch en was verreweg het zwaarste onderdeel van het project. Het verliep hectisch omdat we met het tech-groepje hadden besloten om de opzet van het experiment aan te passen zonder dat goed te communiceren met de anderen. Dit leidde tot tegenstand en veel
discussie waardoor we pas laat konden beginnen met het implementeren, wat al een hele kluif was. Het aanpassen van een bestaand experiment in een onbekende programmeertaal was denk ik net iets te hoog gegrepen. Zonder Kas was het in ieder geval niet gelukt. Het was de zwaarste periode omdat het aanpassen van het experiment het grootste deel was van ons project en ik persoonlijk de inleiding het lastigste deel van een verslag vind. Eén of twee extra weken voor deze periode zouden fijn zijn geweest.
Het verzamelen van de data vond ik erg goed gaan. Het is natuurlijk wel een leerproces waarbij dingen in het begin mis gaan, maar dat hoort er denk ik bij. Het laatste schrijfdeel vond ik ook goed gaan. Ik vond het wel lastig dat ik steeds lang op feedback moest wachten maar het is
uiteindelijk allemaal gelukt en ik ben tevreden met mijn verslag. Sterke en zwakke punten verslag
Sterke punten: Opening van inleiding en overgang naar maatschappelijke en wetenschappelijke relevantie, onderbouwing hypothesen over extraversie, gebruik van plaatjes en grafieken om dingen te verhelderen, resultaten en discussie.
Zwakke punten: Onderbouwing hypothese over invloed van realisme van effort en reward, Mijn hypothesen komen misschien niet heel duidelijk naar voren uit mijn inleiding, de methode is
misschien wat bij elkaar geraapt en komt waarschijnlijk niet helemaal overeen qua schrijfstijl, en het taalgebruik is soms waarschijnlijk wat te informeel en bevat hier en daar grammaticale fouten. Ethische aspecten
Een belangrijk ethisch aspect was het gebruik van virtual reality. Er zijn mensen die daar misselijk of duizelig van worden of zelfs een epileptische aanval krijgen. Om dit zoveel mogelijk te voorkomen zijn hier bijpassende exclusiecriteria voor gebruikt. Ook is tijdens het onderzoek duidelijk aangegeven dat een proefpersoon kon stoppen als die misselijk werd. Daarnaast moesten mensen best wel hard werken voor het onderzoek. Ook hiervoor geldt het dat mensen altijd konden stoppen als zij het werk te zwaar vonden. Tevens was er geen budget om de elke proefpersonen uit te betalen. Daarom is ervoor gekozen om een loting te doen en één proefpersoon een geldbedrag te geven. Eerstejaars psychologiestudenten konden twee proefpersoon punten verdienen met het onderzoek. Tot slot moesten proefpersonen gevoelige informatie invullen bij de vragenlijsten, waaronder dingen over drugsgebruik. De vragenlijsten waren echter anoniem dus met deze gevoelige informatie zijn we goed omgegaan.
