1 Entrainment of Gamma and Theta oscillations:
Can Attentional Effects of Visual Flicker Predict Affective Responses?
Bob Bramson 6050174
Scientific Internship Research Master
Supervisor: R.H. Phaf
Brain & Cognition Program, Psychology department University of Amsterdam
10-10-14
2 Abstract
Neural oscillations in the Theta and Gamma band have been linked to different attention and affective states. Evidence from recent evolutionary simulations (Heerebout & Phaf, 2010) suggests that Theta oscillations code for rigid attention and negative affect, whereas Gamma oscillations are involved in fluent attention and positive affect. In this study we tested these hypotheses by measuring reaction times to moving stimuli, and affective responses on neutral stimuli while eliciting neural oscillations in the Gamma and Theta band through visual flicker. We found attentional cuing effects of both Theta and Gamma oscillations but little to no evidence of affective priming by visual flicker. We argue however, that the lack of affective priming is attributable to our failure to succeed in entraining neural activity to the flicker frequency. Suggestions to enhance the affective priming by visual flickers, and to remove interference by other inadvertent priming factors, are discussed.
3 Introduction
Neural oscillations serendipitously emerged in the evolutionary simulations of Heerebout and Phaf (2010a). A near-doubling of fitness due to this emergence suggested that these
oscillations served an important function. Many different functions have indeed been postulated for neural oscillations; among others, change detection (Donner & Siegel, 2011) and attention (Bauer, Cheadle, Parton, Muller & Usher, 2009). It remains to be seen, however, whether the results of these computer simulations on artificial neural networks can be
extrapolated to biological neural networks. Some progress in this respect has been made by Heerebout, Tap, Rotteveel, and Phaf, (2013), who extended the attentional effects of Gamma oscillations (Bauer et al., 2009) to the experimental demonstration of affective influences that were also predicted by the evolutionary simulations (Heerebout & Phaf, 2010b). The other part of these predictions that lower-frequency, presumably Theta, oscillations code for
negative affect and cause attentional inflexibility has however, not yet been corroborated. The induction of these oscillations by visual flicker may, however, suffer from a number of
practical problems. It is for instance not clear whether the visual flicker is capable of
entraining neural oscillations in all cases. In this study the occurrence of attentional effects is taken to signal successful entrainment, so that also the strongest affective influences should be found in these trials.
In the evolutionary simulations by Heerebout and Phaf (2010a), an agent roamed an artificial environment, looking for food and evading predators. Selection of the fittest agents for reproduction led to the emergence of both organized behavior and an architectural
organization in the initially random networks that controlled the behavior of the agent. In the simulations, different types of oscillations also developed, that corresponded with different types of behavior by the agent. When the agent was looking for food, high-frequency, presumably gamma oscillations (50Hz) emerged and led to a near doubling of fitness
4 (Heerebout & Phaf, 2010a). A further analysis of the, apparently adaptive function of these oscillations revealed that these oscillations allowed attention to shift rapidly from one place to another. When gathering food, rapid attentional switches are probably adaptive, because the agent can cover more ground when attention can quickly shift away from places without food. When the agent encountered a predator, however, rapid attention shifting may no longer be adaptive, even harmful. You do not want to be distracted by food when you are on the run from a predator. “It is better to miss dinner than to be dinner.” In these situations attention became more fixed and lower frequency (presumably Theta, 4-8Hz) oscillations occurred.
The distinction between rapid and slow attention shifting as an effect of Gamma and Theta oscillations can also be extended to basic affect. Gamma oscillations may act as a neural code for positive affect (Heerebout et al., 2013). This might be the case because these oscillations are linked to potentially fitness-increasing stimuli, whereas Theta oscillations may code for fitness-decreasing situations and therefore for negative affect (Johnston, 2003).
The unexpected findings with the evolutionary simulations could just be fanciful hypotheses arising from abstract computer programs that lack sufficient biological
plausibility. However, recent studies found evidence for, at least part of, the theory proposed by Heerebout and Phaf (2010). After inducing Gamma oscillations in their participants through visual flicker, Bauer et al. 2009 found shorter reaction times on a change detection task, indicating increased attention. Heerebout et al. (2013) moreover, found that visual flicker in the Gamma band not only increased attention switching, but also resulted in a more positive evaluation and more approach behavior to neutral faces. These studies support the theory that Gamma codes for positive affect and attention switching. However, so far, no one has investigated the involvement of Theta oscillations in rigid attention and negative affect.
One way to investigate whether Gamma oscillations code for positive, and Theta oscillations for negative, affect is to generate these oscillations using visual flicker in the same
5 frequency band (e.g. Bauer et al., 2009; Heerebout et al., 2013 who used this procedure to influence attention). The effects of oscillations, induced by visual flicker, on attention are more straightforward than the influencing of affect. In the former case, only neurons in the visual cortices have to entrain to the flicker frequency. To influence affect, however, entrainment needs to propagate to higher cortical areas; for instance to the amygdala (e.g., LeDoux, 1996). The success of this study might depend on the successful long-range propagation of neuronal activity throughout the cortex.
As mentioned before, potential attentional or affective influences of visual flicker rely on the entrainment of neural activity to the flicker frequency (Williams, Mechler, Gordon, Shapley & Hawken, 2004). Whether or not neurons entrain to a specific frequency of the flicker can depends on a number of different stimulus characteristics. For instance, it seems more difficult to induce entrainment with a liquid crystal display (LCD) or light emitting diode (LED) computer screen than with cathode ray tube (CRT) screens (Neutel, 2011). In CRT screens electron beams excite the phosphor at the backside of the screen which subsequently lights up. Because the phosphor decays rapidly within 3 to 4 ms (Bridgeman, 1998) high peek intensities are required to keep the average luminance at the same level as in LCD and LED screens. It is likely that the higher peak luminance supports more entrainment; somewhat in the same manner as higher contrast stimuli do (Williams et al., 2004).
The position on which the visual flicker is presented on the retina may also be of vital importance. Where the center of the retina (fovea) contains more cones; sensitive to color and bright light, the periphery contains only rods and is therefore more involved in the detection of dimly lit objects. Since the cones in the fovea provide about 90% of the visual input to the brain (Masland, 2001) entrainment is probably highest when offering high contrast or color stimuli in the center of the visual field (see also Krepel, 2013). In the experiments of Bauer et al. (2009) and Heerebout et al., (2013) visual flicker was presented using suboptimal (low
6 contrast) Gabor patches in the periphery of the visual field. However as mentioned before, this is not necessarily the optimal procedure. In this study we therefore presented visual flicker in the center of a CRT monitor, using high contrast stimuli, to investigate the following question: Do Gamma oscillations elicit positive-and do Theta oscillations elicit negative affect?
To investigate this research question we selected trials with evidence for attentional effects and determined the affective priming by Gamma and Theta flickers in these trials. We induced oscillations by presenting high contrast masks in the center of the screen that
flickered in the Gamma (50 Hz) and Theta (6.25 Hz) band. The participants performed an attentional selection task and an affective priming task. In the attention task they had to indicate, as fast and accurately as possible, to which position a mask had moved. In the affective priming task, the participants rated Chinese characters; either positively or
negatively. We capitalized on the correspondence of attentional and affective influences of flicker, which should only occur with entrainment, by selecting the fastest and slowest trials in the attention task. We tried to maximize affective influences by selecting the fastest 15 trials in the Gamma condition and the 15 slowest trials from the Theta condition. These were compared with the 15 fastest and slowest trials from the control condition. We eliminated the Simon effect (e.g., Krepel, 2013) by using right or left buttons for responding to stimuli that move up or down, respectively. Furthermore, gravity makes downward movements more familiar, and more positive, than upward movements (cf. Phaf & Rotteveel, 2009; Spalek & Hammad, 2004). The affective priming effects of the flickers are therefore examined
separately for the two positions. This leads to the following expectations (see also Figure 1): - Responses on the attention task will be faster in the Gamma condition than in the
7 - Responses on the attention task will be slower in the Theta condition than in the
control and Gamma condition.
- The mean affective response on the Chinese ideograph will be more positive in trials that move downwards than when the trial moves upwards.
- The trials with the fastest responses on the attention task will have a more positive evaluation of the Chinese characters in the Gamma condition than similarly selected trials in the no-flicker condition.
- The trials with the slowest responses on the attention task will have a more negative evaluation of the Chinese characters in the Theta condition than similarly selected trials in the no-flicker condition.
8 2. Method
2.1 Participants
Fifty psychology students (mean age 21.48 yr, sd 1.83 yr) from the University of Amsterdam participated in the experiment, after signing informed consent, in exchange for course credit or €10,-. All participants had normal or corrected to normal vision.
Participants were selected for right handedness (minimum score of 8 on the Van Strien handedness scale; van Strien, 1988) and were screened for epilepsy, also in first and second degree family members, because the visual flicker may trigger seizures. The Van Strien handedness scale and epilepsy screening form are included in the Appendix.
Participants familiar with migraine headaches were also advised not to participate.
Participants who had knowledge of Chinese or related languages could not participate in this experiment, because the target stimuli had to be meaningless to the participants.
Before analyzing the data we determined the number of errors that each participant made on indicating the direction movement of the masks. Since this attention shift task was not very hard, we excluded participants that made too many errors (> SD from the average error rate), as these were presumably not sufficiently motivated. Participants that failed to give a reaction on a substantial number of Chinese ideographs were also removed.
2.2 Design
The experiment had a 3x2 within-participants factorial design. Theta, Gamma or no flicker served as the first independent variable. The second independent variable was the shift in position of the mask, up or down.
Dependent variables were reaction times on the correct responses on the moving mask and proportion of positively evaluated Chinese ideographs.
9 We removed outliers (>2 SD from the mean) per participant per condition. After this the means were recalculated and in every condition, the 15 fastest and 15 slowest trials were selected for further analyses. The second dependent variable was percentage of positively rated Chinese ideographs per participant per condition. This percentage was also calculated for the 15 fastest and slowest trials in the Gamma-Theta-and the no-flicker condition.
To counterbalance effects of left and right button-press, we instructed half of the subjects to push the right button if the stimulus moved upwards and the left button if the stimulus moved downwards. The other half of the participants received opposite instructions. Response-button positions for the affective priming task were fixed. Participants always pushed the right button for positive and the left button for negative with the right-hand index finger. This constitutes the most compatible response positioning for right-handed persons (see Casasanto, 2003). Also, balancing these response buttons would not remedy any design problem, but would only increase error variance.
Effects were analyzed by calculating confidence intervals (<2 SD deviation from the mean) and Cohen’s d (Cohen, 1988) for effect sizes. We calculated Cohen’s d using the following formula: M1-M2 / SDp, where M1 is the mean of control condition and M2 the mean
of the experimental condition. SDp is the pooled standard deviation, which is calculated by
SD1+SD2 / 2.
2.2 Materials and apparatus
Stimuli were presented on a 15 inch LG Flatron 795FT CRT monitor with a refresh rate of 100 Hz. Resolution was set at 1034x768 pixels. The TL lights were switched off because of their potential flicker, but instead the room was lit indirectly by a LED lamp. Participants were seated at a distance of 60 cm from the screen and delivered responses by pressing one of two buttons (left and right) that were placed on a table directly in front of them.
10 Flicker stimuli were created by alternating 50 mid-luminance and 50 low-luminance masks on a grey background (RGB 127,127,127). The masks consisted of 22 pictures of scrambled faces taken from the stimuli used by Heerebout, Todovoric, Smedinga, and Phaf (2013). They were matched on contrast and luminance. The masks were processed in Photoshop by decreasing the RGB values by 100 in the low-luminance stimuli and maintaining the original value in the mid-luminance masks. This resulted in a mean
luminance of 0.28 cd/m² and 9.69 cd/m² for the low and mid luminance masks, respectively. Targets in the affective priming task consisted of 80 neutral and for the participants
meaningless Chinese characters printed in black on a grey background with a mean luminance of 50.12 cd/m². All stimuli were projected on a plane of 200x300 pixels, which made every stimulus the same size; 6x9 cm2. Visual angles for all the stimuli ranged from -4.29° to 4.29°
for the centrally presented stimuli, 0.95° to 9.46° for the stimuli presented on the top, and -9.46° to -0.95° for the stimuli presented at the bottom of the screen. Examples of the masks and Chinese Ideographs can be found in the Appendix.
Each trial started with a fixation cross that was presented centrally for 500-1000 ms. Following the fixation cross, a mask was presented in the center of the screen for 2000 ms. Depending on the condition this mask either flickered at 50 Hz (Gamma), 6.25 Hz (Theta), or 0 Hz (continuous presentation, control condition). After 2000 ms the stimuli moved either up or down the Y-axis by 200 pixels and continued to flicker in the same frequency for another 2000 ms, regardless of whether the participant made a response. This was followed by a Chinese ideograph that disappeared upon response; with a maximum presentation duration of 1500 ms. Including the inter trial jitter of 1000-2000 ms, this made each trial last between 5.5 and 7.5 seconds. Stimuli were selected randomly for each trial, with the exception that the Chinese ideographs could only be used once every condition. Figure 3 shows an example of a flicker trial from the Gamma or Theta condition.
11 Figure 3: An example of a flicker trial in which the mask is moved upwards.
2.4 Procedure
Participants were told that they would participate in an experiment investigating the influence of attention on the classification of otherwise meaningless characters. They were instructed to focus on a fixation cross in the middle of the screen and they were told that they would see flickering or non-flickering stimuli that would move either up or down. Their objective was to react as fast and accurately as possible to the movement of the stimulus. After the reaction on the attention task, subjects had to rate the Chinese character as either positive or negative, by pressing the right or left button.
The trials were divided in one practice block, containing 15 practice trials with
feedback, and three blocks each containing 80 trials, without feedback. Feedback consisted of the text “good job” presented in green after a correct response. If participants made an error in
12 the practice block, the words “you can do better” were presented in red. This practice block was presented only once, regardless of performance. The regular blocks lasted for
approximately 10 minutes each, and were separated by a grey screen with a text indicating it was time for a break. Subjects could resume the experiment at their own pace by pressing one of the buttons.
After the experiment an open exit interview followed. In this exit interview the experimenter enquired for their general impressions and asked whether the participants had used any strategies for responding to the Chinese ideographs. They were also asked how they experienced the flickers, and what influence they thought they had on them. At the end, participants could also pose any questions they had regarding the experiment.
3. Results 3.1 Participants
Six out of the initial fifty participants were excluded from the analyses. Two were removed because of too many errors (> 2 SD) in determining the position shifts of the masks. Three participants were excluded because they missed too many responses (> 2 SD) to the Chinese ideographs. Finally, two participants were excluded from the analyses because they performed more slowly than average (> 2 SD) on every condition. The group of 44 remaining subjects consisted of 14 males and 29 females, age 18-25 (mean 21.45 yr., SD 1.83 yr.).
Three of the participants commented on the experiment. They indicated that they saw sad faces in the masks and thought that they were primed by these faces. The experiencing of emotional faces, and faces in general, is a possible confound in our experiment and is
13 3.2 Attention task
The average RTs and SDs are reported in Table 1 and visualized in Figure 3. Attention-shifting was faster with Theta priming than with Gamma priming which, in turn, was faster than in the no-flicker conditions. Calculating effect sizes gave large effects for Theta (d = 1.42) and Gamma (d = 1.19) versus the control condition. The effects of Theta and Gamma versus control were both statistically significant with t (84) = 6.98, p < 0.05 and t (84) = -5.75, p < 0.05 respectively. This difference between the control condition and the flicker conditions is remarkably large and shows that the flicker acts as an attentional cue to the position in both flicker conditions
The expected faster shifting after Gamma than after Theta flicker was not found. However, the reverse effect occurred with d = 0.29 for Gamma versus Theta. However, this difference not statistically significant t (84) = 1.37, p = 0.173.
Table 1: Reaction time averages and SD in ms of the different conditions.
Condition Gamma Theta Control
Reaction time 534 (77) 512 (75) 641 (107)
Table 2 presents the average RTs and SDs of the trials when split over up and down position. Differences between RTs are negligible for different positions, with the largest effect d = 0.15 for the control up versus control down conditions. There is therefore no evidence that attentional shifts in the downward direction are favored above upward shifts.
14 Table 2: RTS and SDs in ms split out over position.
Gamma Theta Control
Up 534 (78) 514 (73) 638 (106)
Down 534 (74) 510 (77) 654 (106)
When selecting the fastest and slowest trials the differences between flicker conditions remain relatively similar (see Table 3), which indicates that overall we may not have
succeeded in inducing neural entrainment to the visual flickers.
Table 3: RTs and SDs in ms for fastest, slowest and all trials.
Condition/Position Fast Trials All Trials Slow Trials
The differences between the trials with the fastest RTs yield effect sizes of; d = 1.17 and d = 1.37 for Theta up and down versus control up and down and d = 0.97 and d = 1.25 for Gamma up and down versus control up and down. Theta and Gamma do not differ, with d = 0 and d = 0.14.
The slowest trials show approximately the same effects between conditions. This yields the following effect sizes: d = 1.34 and d = 1.64 for Theta up and down versus control
Theta Up 429 (59) 514 (73) 612 (95) Theta Down 420 (63) 510 (77) 611 (93) Gamma Up 441 (66) 534 (78) 641 (94) Gamma Down 429 (63) 534 (74) 629 (99) Control Up 518 (94) 638 (107) 761 (128) Control Down 525 (90) 654 (106) 796 (134)
15 up and down; d = 1.07 and d = 1.44 for Gamma versus control, and d = 0.31 and d = 0.19 for Gamma versus Theta.
Figure 3. Reaction time averages and SD
3.3 Affective Priming
We calculated the percentage positively rated Chinese ideographs divided over condition and position. These averages and their SD’s are visualized in Table 5 and Figures 6 and 7.
Overall, there was little evidence for affective priming by Gamma and Theta flickers in the experiment. Only the direction of attentional shifts produced a convincing affective priming effect. 0 100 200 300 400 500 600 700 800 Rea ction times in ms Condition
Mean reaction times per condition
Gamma Control Theta
16 Table 5: Percentage positively rated ideographs split over condition/position.
Condition/Position Fast Trials All Trials Slow Trials Theta Up 53,5 (20.3) 51,9 (18.1) 52,6 (19.2) Theta Down 61,1 (20.2) 60,0 (18.2) 58,3 (17.6) Gamma Up 48,2 (20.3) 50,1 (20.0) 51,9 (23.5) Gamma Down 59,7 (21.9) 60,1 (20.5) 62,9 (21.5) Control Up 52,7 (20.2) 52,3 (18.0) 52,7 (21.6) Control Down 61,1 (20.6) 61,0 (18.2) 60,8 (22.1)
In every condition, the ideographs were rated more positively after a downward movement than after an upward movement (see Table 6). This resulted in a significant
medium effect of position: d = 0.48, t (256) = -3.84, p < 0.01. This effect remains stable if we select the fastest or slowest trials and thereby confirms our expectation of a positive effect of downward movement, perhaps due to familiarity.
Table 6: Percentage positive and SD for up and down trials.
Condition Up Down
Percentage Positive 51.4 (18.6) 60.4 (18.9)
The overall differences between conditions are less pronounced. The effect size of Theta versus control is d = -0.02 for up trials and d = -0.06 for down trials. Gamma versus control yields effect sizes of d = -0.07 for up, and d = -0.04 for down trials. Theta and Gamma conditions also do not differ substantially, the small effect that is seen is in an opposite direction to our expectations, d = -0.09 for up and d = 0 for down. None of these effects is significant: for all comparisons t (42) < 0.6, p > 0.5. When considering the effects
17 based on the CI’s and d (Cohen, 1988), we conclude that there is no effect of flicker over all trials.
The effects of the flicker differed slightly when selecting the fastest trials. Effects of Theta vs. control are d = 0.04 and d = 0 for down. Comparing Gamma and control trials results in d = -0.22 and d = -0.07 for up and down trials, respectively. There seems to be a small effect here, however it is opposite of our expectation that selecting the fastest Gamma trials results in the most positively rated ideographs. Theta and Gamma differ slightly with d = 0.25 and d = -0.07, again in opposite direction of our expectations. These effects are also not significant: all t < 1.2 , p > 0.25.
Selecting the slowest trials in each condition results in the following effects; d = 0 and d = -0.12 for Theta and control up and down trials; d = -0.04 and d = 0.1 for Gamma versus control; and d = 0.03 and d = -0.24 for Theta versus Gamma. The effects between Theta and control and Theta and Gamma (down) seem to be in line with our hypothesis that Theta causes more negative mood, however, since these effects are very small, non-significant (all t < 1.1 , p > 0.23) and the results are not stable between up and down trials they should be interpreted with caution.
Overall, the effects of flicker, if any, are marginal and provide little to no evidence of affective priming by flicker. We conclude that we probably did not succeed in entraining neuronal oscillations to flicker frequencies.
18 Figure 6. Percentage and SD of positively Figure 7. Percentage and SD of positively
rated ideographs for the upward trials. rated ideographs for the downward trials.
3.4 Exploratory analysis
We reasoned that the cuing function of the flickers, as seen in the decreased RTs, might be caused by a more fluent attention shift and therefore also increase positive affect. In this case RTs should correlate negatively with affective rating. To test this, we calculated the correlation coefficient Pearson’s r for RTs versus percentage positive.
Overall we found no reason to assume that attention shifts after flicker cues were more fluent than after no cuing and lead to more positive affect (biggest r was -0.23).
4. Discussion
In this study we found that Theta and Gamma flicker acted as an attentional cue to the position of the attentional shift. The additional fluency of attention shifts did not result in more positive evaluations than without a flicker cue. Contrary to our initial expectations, the attention effects were larger in the Theta than in the Gamma condition. Selecting the fastest trials in the Gamma condition resulted in more negative responses than in the control
condition, whereas we expected more positive affect. In the Theta condition the fastest trials yielded the same affective responses as the trials without a flicker. In the slowest trials however, there was some indication in favor of our hypothesis of Gamma coding for positive and Theta for negative affect. Responses were most positive in the Gamma condition and
19 more negative in the Theta condition. However, these results need to be interpreted with extreme caution because the effects were small and not consistent over different positions.
Furthermore, we found a large difference in affective rating between downward and upward movement of the masks. Downward movement was overall rated more positively than upward movement. This difference might be due to the familiarity of downward movement (i.e., due to gravity). That this type of familiarity can elicit more positive affect than non-familiar movements has previously been shown by Chokron and De Agostini (2000), who found that left to right readers (French) showed an affective preference to pictures that
showed rightward directionality; right to left readers (Israeli) however, showed a preference to stimuli with a leftward direction. Differences in affective responses between familiar and unfamiliar movement might occur because familiar movement results in less cognitive
conflict and is therefore processed more fluently, leading to positive affect (Phaf & Rotteveel, 2009).
Despite the attentional cuing by the flicker, the attentional effects in our study differed from the results from the studies of Heerebout et al., (2013) and Bauer et al., (2009) in that they did not seem to arise from neural oscillations. Whereas in the study of Heerebout et al., (2013) the attention effects were accompanied by changes in affect, this affective reaction was mostly absent in our results. Furthermore, if neuronal activity entrained to flicker frequency, you would expect differential attentional effects of the flicker in the fast, slow and all trial selections. When selecting the fastest trials, the difference in evaluations between Gamma and no-flicker should be larger than over all trials. When selecting the slowest trials, the
difference in evaluations between Theta and no-flicker should be larger than over all trials. In view of the largely absent differences in flicker effects between fast, slow and all trial
20 place in this experiment. Therefore the question is now how to improve the chances of
obtaining entrainment and affective influences of flickers?
There are several possible reasons as to why the results of the affective priming task differ from our expectations. Firstly, high-contrast flickers in the center of the retina may result in less entrainment than low-contrast flickers in the periphery as in Bauer et al. (2009). Secondly, the facial features in the flickering masks may inherently induce conflict and affect (cf. Phaf & Rotteveel, 2012) that may have obscured the affective influences of the flickers. Finally, given the big difference in luminance between the mask and the ideographs the affective priming stimuli might have caused a transition flash (Van Diepen, Born, Souto, Gauch & Kerzel, 2010).
As for the first explanation; we decided on the properties and placement of our stimuli based on knowledge of the retinal organization, as mentioned in the introduction. However, we now reason that the high-contrast flicker in our experiment possibly only saturates a small part of the dynamic range of the properties that the cones, in the center of the visual field, are sensitive to. The low-contrast flicker in the experiments of Bauer et al., (2009) and Heerebout et al., (2013) probably saturates a much bigger portion of the luminance range of the rods. In future experiments we advise to offer low contrast stimuli in the periphery of the visual field.
The fact that our masks contain features of faces did not seem to be a problem at first. However, in a recent study at our department, Kovacs (2013) found that suboptimal affective priming effects disappeared when masking the affective stimuli using a scrambled face mask (the same masks as we used in this experiment). Suboptimal affective priming re-appeared when an empty screen, instead of a mask, followed the prime (Kovacs, 2014). It is possible that the masks we used in our experiment somehow overrule the affect elicited by the flicker; perhaps because scrambled faces create some kind of conflict (Phaf & Rotteveel, 2012), much in the same manner as a color-word conflict in the classical Stroop task has been shown to
21 elicit affect (Dreisbach & Fisher, 2012). We advise to use non-conflicting stimuli, for
example Gabor patches, to avoid overruling possible affective priming affects.
Another factor that may have reduced the affective reaction on the affective priming task is the occurrence of a transition flash. The effect of transition flash was first proposed by Van Diepen et al. (2010), who found that transition between 50Hz and 100Hz flicker caused a ‘flash-like impression’ in their subjects. The sudden increase of luminance between the mask and Chinese ideograph might have caused such a flash, possibly overruling a smaller flicker effect on the affective priming task. To avoid this transition flash in future experiments, we propose to match stimuli luminance beforehand.
In light of these limitations, we propose a follow-up experiment in which our research question remains unchanged: Do Gamma oscillations elicit positive- and do Theta oscillations elicit negative affect?
In this follow-up experiment, participants will focus their eyes on a fixation cross and will be instructed to react as fast as possible if a stimulus appears somewhere on the screen. These stimuli will be presented on a random location, located on an invisible ring around the center of the screen (much like Bauer et al., 2009) and consist of a non-conflicting image (e.g. a neutral face). The stimuli are preceded by a Gabor patch on the same location, either
flickering in Gamma or Theta frequency, or not flickering at all. After response, the target stimulus is replaced by a neutral stimulus - with the same luminance - and this stimulus is rated using the same forced choice paradigm that we used. This design will theoretically maximize effects of flicker (low contrast in the periphery), show no unanticipated conflict due to conflicting stimuli, and will eliminate transition flash because luminance is kept stable. If this procedure is adhered to, chances are that affective priming can be predicted by selecting on attention effect.
22 5. References
Bauer, F., Cheadle, S. W., Parton, A., Muller, H. J., & Usher, M. (2009). Gamma flicker triggers attentional selection without awareness. Proceedings of the National Academy of Sciences, 106, 1666-1671.
Bridgeman, B. (1998). Durations of stimuli displayed on video display terminals: (n-1)/f + Persistence. Psychological Science, 9, 232-233.
Canolty, R. T., & Knight, R. T. (2010). The functional role of cross-frequency coupling. Trends in cognitive sciences, 14, 506-515.
Cohen, M. X., Axmacher, N., Lenartz, D., Elger, C. E., Sturm, V., & Schlaepfer, T. E. (2009). Good vibrations: cross-frequency coupling in the human nucleus accumbens during reward processing. Journal of Cognitive Neuroscience, 21, 875-889.
Cohen, W. M., & Levinthal, D. A. (1989). Innovation and learning: the two faces of R & D. The economic journal, 569-596.
Donner, T. H., & Siegel, M. (2011). A framework for local cortical oscillation patterns. Trends in Cognitive Sciences. 15, 191-199.
Fingelkurts, A. A., Fingelkurts, A. A., Rytsälä, H., Suominen, K., Isometsä, E., & Kähkönen, S. (2006). Composition of brain oscillations in ongoing EEG during major depression disorder. Neuroscience research, 56, 133-144.
Fries, P. (2005). A mechanism for cognitive dynamics: Neuronal communication through neuronal coherence. Trends in Cognitive Neuroscience, 9. 474-480.
Heerebout, B. T., & Phaf, R. H. (2010). Good vibrations switch attention: An affective
function for network oscillations in evolutionary simulations. Cognitive, Affective, and Behavioural Neuroscience, 10, 217-229
Heerebout, B. T., Todovorić, A., Smedinga, H. E. & Phaf R. H. (2013). Affective
197-23 211.
Heerebout, B. T., Tap, Y.A.E., Rotteveel, M., & Phaf, R.H. (2013). Gamma flicker elicits positive affect without awareness. Consciousness and Cognition, 22, 281-289, Johnston, V. (2003). The origin and function of pleasure. Cognition and Emotion, 17,
167-179
Linkenkaer-Hansen, K., Monto, S., Rytsälä, H., Suominen, K., Isometsä, E., & Kähkönen, S. (2005). Breakdown of long-range temporal correlations in theta oscillations in patients with major depressive disorder. The Journal of Neuroscience, 25, 10131-10137. LeDoux, J. E. (1996). The emotional brain. New York: Simon & Schuster.
Masland, R. H. (2001). The fundamental plan of the retina. Nature Neuroscience, 4, 877-886.
Mathewson, K. E., Gratton, G., Fabiani, M., Beck, D. M. & Ro, T. (2009). To See or Not to See: Prestimulus Phase Predicts Visual Awareness. The Journal of Neuroscience, 29, 2725-2732.
Van Diepen, R. M., Born, S., Souto, D., Gauch, A. & Kerzel, D. (2010). Visual Flicker in the Gamma-Band Range Does Not Draw Attention. Journal of Neurophysiology, 103, 1606-1613
Van Strien, J. W. (1992). Classificatie van links-en rechtshandige proefpersonen. Nederlands Tijdschrift voor de Psychologie en haar Grensgebieden. 47, 88-92
Williams, P. E., Mechler, F., Gordon, J., Shapley, R., & Hawken, M. (2004). Entrainment to video display in primary visual cortex of macaque and humans. Journal of
24 Appendix
Van Strien Handedness questionnaire:
Handvoorkeursvragenlijst “Intuitieve Beslissingen”
Met welke hand teken je? Links Rechts Beide
Met welke hand poets je je tanden? Links Rechts Beide
Welke hand gebruik je om een fles te openen? Links Rechts Beide Welke hand gebruik je om een bal te gooien? Links Rechts Beide
Met welke hand hou je een hamer vast? Links Rechts Beide
Met welke hand hou je een tennis racket vast? Links Rechts Beide Met welke hand hou je een mes vast tijdens het snijden? Links Rechts Beide
Met welke hand roer je met een lepel? Links Rechts Beide
Welke hand gebruik je om iets uit te gummen? Links Rechts Beide Met welke hand steek je een lucifer aan? Links Rechts Beide
Epilepsy questionnaire:
Vragenlijst voor het onderzoek “Intuitieve Beslissingen”
25 2. Heeft u migraine (bv. met eenzijdige hoofdpijn, gezichtsuitval en andere visuele
verschijnselen)?
3. Heeft u een andere neurologische aandoening? 4. Heeft u ooit een beroerte gehad?
5. Heeft u hersenletsel, als gevolg van een hersenoperatie, een zware verwonding aan het hoofd of een ziekte?
6. Heeft u ooit een zware hersenschudding2 gehad?
7. Heeft u ooit een aandoening aan de hersenen gehad, zoals hersenvliesontsteking?
2 zware hersenschudding: dermate ernstig dat de persoon 24 uur ter observatie in het
ziekenhuis heeft moeten blijven; waarschijnlijk optreden van hersenkneuzing. In geval van twijfel proefpersoon weigeren.
Stimuli example, original and contrast mask:
Stimuli example, Chinese Ideograph:
