Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2014 Aug 18.
Published in final edited form as: Vision Res. 2012 Jan 5;55:47–51. doi: 10.1016/j.visres.2011.12.009

Motion adaptation does not depend on attention to the adaptor

Michael J Morgan 1
PMCID: PMC4135072  EMSID: EMS59838  PMID: 22245710

Abstract

Prolonged inspection of moving stimuli causes stationary stimuli to appear moving in the opposite direction to the adapting stimulus (the Waterfall effect). It has been claimed that distracting the viewer’s attention from the adapting stimulus by a secondary task reduces the strength of adaptation. However, the method used to show the effect of distraction (the duration of the aftereffect) is potentially susceptible to bias. The experiments reported here show no effect in genuinely naïve subjects, or in experienced observers using a variety of cancellation procedures to measure the effect.

Keywords: Motion, Adaptation, Attention, 2AFC, Bias

1. Introduction

Prolonged inspection of a moving image such as a waterfall causes subsequently-viewed stationary stimuli to appear as if they are moving in the opposite direction (Addams, 1834; Mather, Verstraten, & Anstis, 1998). Evidence for this motion after-effect is also seen in the reduced sensitivity of neurons to their preferred direction of motion after adaptation (Barlow & Hill, 1963), and in neurones in primary visual cortex of primates (Kohn & Movshon, 2003). It is therefore interesting that the strength of adaptation is apparently reduced if the observer’s attention is distracted away from the adapting stimulus by a competing task (Chaudhuri, 1990; Rees, Frith, & Lavie, 1997; Rezec, Krekelberg, & Dobkins, 2004; Taya et al., 2009). The great merit of the adaptation paradigm over others such as the ‘dual task’ method, is that in the former, the distracting task does not compete for high-level processes such as memory or response selection in the test phase.

Unfortunately, the evidence for the distraction effect is not as strong as it might be. The great majority of experiments on the effects of attention on adaptation have used the duration measure of the after-effect, a measure known to be highly susceptible to experimenter/subject bias (Sinha, 1952). It is difficult to decide when a stimulus appear to stop moving, particularly when the observer knows that it is actually stationary. The observer has to adopt some criterion, and this criterion could easily be altered by an unconscious wish to give the experimenter the desired result (Rosenthal & Rubin, 1978). Unsurprisingly, then, the literature is not unanimous. In the very first experiment on the topic, Wohlgemuth (1911) found no effect of a distracting task in central vision on the after-effect. In other cases where negative results have been reported (Georgiades & Harris, 2002; Rees, Frith, & Lavie, 2001), authors have ascribed the failure to a difference in conditions, rather than entertaining the possibility that the positive results are Type I errors. The Rees et al. study (Rees, Frith, & Lavie, 1997) used only four subjects (out of 6 in the accompanying fMRI study) and it not stated whether or not they knew the predictions of the experimenters.

Nishida and Ashida (2000) investigated the effect of distraction on both the monocular and interocular MAE, using both duration and counterphase-grating nulling methods (see Methods in the present paper for a description of this technique). In contrast to Chaudhuri (1990), they found no clear effect of distraction on the MAE duration, a discrepancy they did not explain. They did find an effect with the nulling method on the interocular MAE but not on the monocular MAE. Thus they found an effect in only one of four conditions. We return to the statistical issues raised by such findings in discussion.

The only study so far to have used a direct measure of the loss of sensitivity to the adapted direction of motion, using 2AFC to measure the complete contrast discrimination function before and after adaptation under two conditions of attentional load, failed to find any effect of load during adaptation (Morgan, 2011).

The effects of attentional tracking on adaptation have been used to support the claim that distraction affects adaptation. Indeed, experimental results (Alais & Blake, 1999; Lankheet & Verstraten, 1995; Raphael, Dillenburger, & Morgan, 2010) have convincingly shown that tracking one component of a transparent-motion display produces a motion aftereffect opposite to the direction of the attended component. However, this is conceptually different from the distraction effect and need not depend on the same mechanism. In fact, using an adapting stimulus with balanced expansion and contraction, Raphael, Dillenburger, and Morgan (2010) reported an effect of attentional tracking, but no effect of distraction when using only one adapting component.

The evidence for an effect of distraction on adaptation is therefore less than compelling. For this reason, and because the under-lying theoretical issue is important, it would be desirable to have confirmation of the effects of attentional distraction on adaptation using a variety of procedures. The present paper reports an investigation using the duration measure in six genuinely unbiased subjects, and a second duration experiment in six different subjects given instructions similar to those in Rees et al., who warned their subjects that if they failed to obey the exhortation not to attend to the adapter, they might get an ‘unpleasant’ aftereffect. We also report two further experiments using a cancellation paradigm to measure the after-effect.

2. General methods

Stimuli were computed with MATLAB and displayed by a Cambridge Research System VSG 2/3 graphics card on a Sony monitor (resolution 640 pixels width by 479 pixels height; pixel size 1.03 arcmin, mean luminance 37.5 cd/m2). Viewing distance was 2 m. In Experiments 1–3 the adapting stimulus consisted of 100 white dots/frame (each dot 4 × 4 pixels; luminance 60 cd/m2), randomly placed in a circle of diameter 10°, moving outwards from the centre of the screen at a velocity of 2°/s. The dots had limited lifetime (Morgan & Ward, 1980) and were randomly replaced by a dot in a new position with a probability 0.05/frame. When a dot reached the edge of the circle, it was replaced by a dot in a random position within the circle. The attentional task was based on a recent paper showing a greater BOLD response to a peripheral stimulus under low vs. high load (Schwartz et al., 2005). Coloured ‘T’ like stimuli were presented at a rate of 2/s at fixation within a mean-luminance ellipse of dimensions 1.1 × 0.73° at the centre of the adapting stimulus. The low-load task was to spot an infrequent red stimulus, independently of orientation, and to press the ‘enter’ key on the computer keypad. The high-load task was to spot either of two conjunctions, e.g. green-upright and blue-inverted.

2.1. Experiments 1 and 2

2.1.1. Methods

The expanding dots stopped moving after 60 s of adaptation and the subject was instructed to press a key when they appeared to stop. Four separate trials were run, two in the Low Load and two in the High Load condition, the order being counterbalanced over subjects. The first experiment used six optometry students who were unaware of the predictions of the ‘attentional load’ hypothesis. Each subject was tested for four trials (ABBA or BAAB design in different subjects) by another student as part of a research project, who was told as was indeed the case, that the idea was to repeat previous work as a preliminary to further experimentation to determine whether load was affecting response gain or the semi-saturation constant of the adaptor. In a second experiment with six new subjects the instructions were based on Rees et al., who warned their subjects that if they attended to the adapting stimulus to the adapter, they might get an ‘unpleasant’ aftereffect. Otherwise the procedure was the same as in Experiment 1.

3. Results and discussion

Load manipulation at fixation in Experiment 1 was effective; reaction times were significantly longer in the high-(vs. low-) load conditions (1356 ms vs. 847 ms; p < .001) and were comparable to those reported in an earlier study (Morgan, 2011). The difference was also significant in each of the subjects analysed separately (p < .01). The RT ratio between the two conditions (1.6) was greater than that (1.17) recently reported by Bahrami, Lavie, and Rees (2007) and given by them as evidence that the tasks involved different perceptual loads. However, results for the after-effect duration (Fig. 1a) showed no obvious effect of the central task.

Fig. 1.

Fig. 1

Open in a new tab

The figure shows results of two experiments measuring the duration of the motion after-effect (vertical axis) under two conditions of attentional load (white bars: high load, grey bars, low load) during adaptation. The top panel shows results of Experiment 1 and the bottom panel shows Experiment 2. The error bars show the range of the data (n = 2). The subjects (label on horizontal axis) were optometry students unaware of the experiment or the theory behind it. In Experiment 2 instructions were deliberately manipulated to reduce the size of the after-effect. Results showed no effect of the attentional load.

In Experiment 2 subjects were instructed not to attend to the adaptor in order to avoid getting an ‘unpleasant’ after-effect. Once again, reaction times were significantly longer in the high- (vs. low-) load conditions (1256 ms vs. 823 ms; p < .001) and were comparable to those reported in an earlier study (Morgan, 2011). The main effect of the instruction seems to have been to significantly reduce the duration of the after-effect and increase its variability but again there was no effect of the central task. A 4-factor ANOVA (subject × trial × load × experiment) with subjects nested within Experiment showed a highly significant difference between subjects (F(10,47) = 12.02; p = .0001) and between Experiments 1 and 2 (F(1,47) = 21.2; p = .0008) but no significant main effect of load (p > 0.2) or interactions. The most likely explanation of the difference between Experiments is the very low durations of the MAE in three subjects of Experiment 2, possibly due to the instructions, but because of the inevitably nested design, the effect could be due to individual differences, rather than experimental procedure.

It may be asked whether the present design was sufficiently powerful to reject the null hypothesis, given the size of effect and variance reported by Rees et al. The present experiments used 12 observers in total vs. only 4 in Rees et al., but only two trials/condition vs. 7 in Rees et al. There was a reason for using more subjects and fewer trials. The use of large numbers of trials in a small number of subjects in this context is arguably an example of pseudo-replication (Hurlbert, 1984), since a subject who has a cognitive bias against the null hypothesis on one trial is likely to have it present on all. Subjects may also try to make their responses consistent, in which case adding more observations than one per condition gives little additional information. This is why it was considered important to use a relatively large number of subjects in the present study rather than a large number of trials in few subjects. Nevertheless, we addressed the issue of power by considering the four subjects used by Rees et al. as random samples from the population. Using the individual means and standard deviations provided by Rees et al., it can be shown that the probability of the high-load mean score being greater than the low-load mean score given just two observations per condition would have been (for each of the four observers in turn) 0.966, 0.904, 0.849, 0.978. Thus two observations per observer is in principle quite sufficient to show a difference between conditions. The reason for this is that the variability within observers is very low compared to that between observers, reinforcing the need to take account of the possibility of pseudo-replication when interpreting the results.

To further quantify the power, the present experiment was simulated 10,000 times, using a random selection of the four observers in Rees et al. on each occasion to make up 6 or 12 simulated observers. The probability of obtaining a significant result at the 0.05 one-tailed level (as used by Rees et al.) by the same ANOVA used in analysing the actual data was 0.897 when using six observers and 0.999 using 12. It may be concluded that we would have been very unlucky indeed to reject the null hypothesis with the number of 12 independent observers used in the two experiments combined.

3.1. Experiment 3

3.1.1. Methods

The adapting stimuli were the same as in Experiments 1 and 2 and the High Load task was also the same. However, the ‘low load’ task was replaced by a ‘no load’ procedure in which the subjects were not required to respond to a target and were encouraged instead to attend to the adaptor. There was an initial period of 60 s adaptation to expanding motion, followed by a series of five trials in each of which a 30 s adaptation period was followed by a 30 s test. The test consisted of the same dot pattern as the adapter with the stimuli moving initially moving inwards (contracting) at a velocity of 0.5°/s. The observer was provided with two toggle switches, which they pressed as frequently as possible to indicate the perceived direction of motion (inwards vs. outwards). Each press on the ‘inwards’ switch decreased the amplitude of the inwards movement by a small amount. If the amplitude of inwards motion became negative, the stimuli moved outwards. Presses on the ‘outwards switch had the opposite effect. The effect of this procedure was to home in on the motion null point. To discourage tracking of individual dots the probability of dot replacement was increased to 0.2/frame. Each test was followed immediately by the next adaptation period.

3.1.2. Results

The effectiveness of the nulling procedure is seen from the fact (Fig. 2) that all curves eventually converged on the true motion null (zero velocity). However, the same curves show that observers initially over-compensated by making the dots actually expand, presumably because of the contracting after-effect. There was no obvious effect of the central task.

Fig. 2.

Fig. 2

Open in a new tab

The figure shows a time series (horizontal axis) of results of Experiment 3 in which six practiced psychophysical observers tried to null the motion after-effect of an expanding dot pattern by pressing on toggle switches. The whole series covers 30 s. Pressing one switch made the stimulus contract (negative values on vertical axis) and pressing the other made in expand (positive values). The stimulus started out physically contracting. All observers over-compensated the amount of expanding motion before returning gradually to baseline. The average results (bottom right panel) shows that there was no overall effect of attentional load during adaptation.

3.2. Experiment 4

3.2.1. Method

In Experiment 4 (counterphase flicker), the adapting stimulus consisted of a 45° oriented, drifting 2.05 cyc/° sinusoidal grating of temporal frequency 7.5 Hz windowed by a stationary Gaussian envelope (s = 2~33°, presented initially for 60 min and subsequently for 5 s before each test stimulus. Its Michelson contrast unless otherwise stated was 0.075. Contrast was controlled by a look-up table with 15 bits resolution. To ensure a linear relation between DAC voltage and luminance, the display was calibrated with the Cambridge Research Systems OPTICAL. The three DAC’s were individually calibrated. The test consisted of a 300 ms presentation in which alternating frames of the same grating moved in the same direction as the adapter and in the opposite direction. The contrast of the two components could be independently manipulated. On each trial the observer pressed one of two buttons to indicate whether the stimulus appeared to move left or right. They also had a third button to press (Garcia-Perez, 2010) if they were genuinely uncertain of the direction, in other words, if they saw the stimulus as stationary. It was stressed that this ‘stationary’ button should only be used exceptionally. ‘Leftwards’ responses increased the relative strength of the ‘Rightwards’ component and vice versa, the summed contrast of the two components remained constant. Two independent staircases were randomly interleaved starting with high relative contrasts of the Rightwards and Leftwards component respectively (see Morgan, Chubb, and Solomon (2006) for details). The perceptual load conditions were the same as in Experiment 3, i.e. high load vs. no load.

3.2.2. Results

Results of five practiced psychophysical observers were analysed (Fig. 3) as psychometric functions of response probability against relative component contrasts and showed the expected effect of a shift in the 50% point towards higher contrasts of the adapted component. The probability of choosing the ‘stationary’ button peaked near to the point where the probability of reporting the two directions was equal. There was no evidence for an effect of the attentional load.

Fig. 3.

Fig. 3

Open in a new tab

The figure shows results of Experiment 4 in which five practiced psychophysical observers pressed buttons to indicate whether they saw a test stimulus moving in the previously adapted direction (triangles), in the previously unadapted direction (circles) or not moving at all (squares). The test stimulus was the sum of two gratings, one moving in the previously adapted direction and the other in the opposite (unadapted) direction. The proportion of the total contrast energy in the two directions (horizontal axis) was varied over trials by a staircase procedure and all the subject’s responses subsequently collated to make the psychometric functions in the figure. Adaptation was carried out under two conditions of attentional load of a distracting task during adaptation. For further explanation see the text. The results shown are means over the five subjects. All subjects showed the same pattern.

4. General discussion

In summary, we have been unable to find an effect of attentional load on any of three measures of adaptation. There may be some as-yet unspecified methodological detail that is necessary to get the effect. However, in view of the presence already of negative results in the literature (Georgiades & Harris, 2002; Nishida & Ashida, 2000; Rees, Frith, & Lavie, 2001; Wohlgemuth, 1911) the possibility should be considered that the positive reported results are Type I statistical errors; and that the prevalence of positive results reflects the ‘File Drawer Effect’ (Rosenthal, 1979), or in other words, the putative reluctance of Journals to report negative results.

As noted in the Section 1, negative results have typically been explained as due to differences in conditions rather than as true failures to replicate. This raises interesting statistical issues. Georgiades and Harris (2002), for example, found an effect of distraction at one temporal frequency of adaptor but not at another. Rather than treating this as a 1–1 drawn contest between the null hypothesis and the alternative, they explain the results as due to an interaction between distraction and temporal frequency of adaptor. This argument would be convincing if backed up by an Analysis of Variance, but it was not. Similarly, in the study by Nishida and Ashida (2000) described in Section 1, an effect of distraction was found in only one of four conditions. It was absent in the very case (duration measure of the monocular MAE) previously reported to have an effect. Nishida and Ashida suggest that distraction only has an effect on the later stages of the hierarchy involved in motion processing. As with Georgiades and Harris, however, no Analysis of Variance was given to demonstrate the presence of an interaction. The possibility thus has to be considered that the occasional positive results reported in the literature arose by chance.

The other possibility is that subjects in some experiments may have been unintentionally biased in favour of the hypothesis that attentional load affects adaptation. Most of these studies have used the duration measure of the after-effect, a measure known to be highly susceptible to subject bias (Sinha, 1952). If previous reports of effects of attentional distraction on adaptation have indeed depended on experimenters’ unintentionally biasing their subjects, the question arises how such biases should be guarded against in future experiments. In traditional psychophysics, a distinction was drawn between ‘Class A’ observations, which are relatively immune to bias and ‘Class B’ where the observer has to adapt some form of criterion, and which are thus susceptible to bias (Brindley, 1960). Class A observations determine conditions under which the observer is unable to distinguish between two or more stimuli and are thus able to measure thresholds for discrimination. An example of a Class A procedure is the two alternative choice (2AFC) method in which the observer decides which of two successively or simultaneously presented stimuli has the higher intensity. The important point stressed by Brindley (1960) is that if an observer can in fact distinguish between two stimuli, this cannot be due to a subjective bias. There must be some mechanism in the brain able to respond differently to the two stimuli. An example of a Class B procedure, on the other hand, would be one in which the observer has to say whether a green stimulus is moving faster than a red stimulus. By adjusting the relative speeds of the two stimuli we can find a point at which they appear to the observer to move at the same speed, but this point reflects a decision by the subject, which could well be biased by expectations and instructions.

All the Methods used in the experiments reported here and in most previous studies of the problem are Class B. It is possible, however, to test the effects of attentional distraction by a Class A procedure. It is known that motion adaptation selectively reduces contrast sensitivity to stimuli moving in the same direction as the adaptor (Sekuler & Ganz, 1963), and this effect can be measured by the 2AFC procedure. In fact, in a recent study using 2AFC to measure the complete contrast discrimination function before and after adaptation under two conditions of attentional load, no effect of load was found (Morgan, 2011). If Class A observations can be carried out they should be preferred to Class B because of their relative immunity to bias. If they are not possible, for example when the point of the experiment is measure the effect of context upon the appearance of a stimulus, then precautions are needed to avoid influencing the subjects in the direction of the expected result. Rather than casually describing subjects as ‘naïve’ it would be useful to have details of what they did and did not know and exactly how they were instructed (Morgan et al., 2011). Where possible the experimenter as well as the subject should also be naïve.

Acknowledgment

Supported by the Max-Planck Society.

References

  1. Addams R. An account of a peculiar optical phenomenon seen after having looked at a moving body etc. London and Edinburgh Philosophical Magazine and Journal of Science (3rd series) 1834;5:373–374. [Google Scholar]
  2. Alais D, Blake R. Neural strength of visual attention gauged by motion adaptation. Nature Neuroscience. 1999;2(11):1015–1018. doi: 10.1038/14814. [DOI] [PubMed] [Google Scholar]
  3. Bahrami B, Lavie N, Rees G. Attentional load modulates responses of human primary visual cortex to invisible stimuli. Current Biology. 2007;17(6):509–513. doi: 10.1016/j.cub.2007.01.070. doi:10.1016/j.cub.2007.01.070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barlow HB, Hill RM. Evidence for a physiological explanation of the waterfall phenomenon and figural after-effects. Nature. 1963;200:1345–1347. doi: 10.1038/2001345a0. [DOI] [PubMed] [Google Scholar]
  5. Brindley GS. Physiology of the retina and visual pathway. Edward Arnold; London: 1960. [Google Scholar]
  6. Chaudhuri A. Modulation of the motion aftereffect by selective attention. Nature. 1990;344:60–62. doi: 10.1038/344060a0. [DOI] [PubMed] [Google Scholar]
  7. Garcia-Perez MA. Denoising forced-choice detection data. The British Journal of Mathematical and Statistical Psychology. 2010;63(Pt. 1):75–100. doi: 10.1348/000711009X424057. [DOI] [PubMed] [Google Scholar]
  8. Georgiades M, Harris J. Evidence for spatio-temporal selectivity in attentional modulation of the motion aftereffect. Spatial Vision. 2002;16(1):21–31. doi: 10.1163/15685680260433887. [DOI] [PubMed] [Google Scholar]
  9. Hurlbert SH. Pseudoreplication and the design of ecological field experiments. Ecological Monographs. 1984;54(2):187–211. [Google Scholar]
  10. Kohn A, Movshon J. Neuronal adaptation to visual motion in area MT of the Macaque. Neuron. 2003;39:681–691. doi: 10.1016/s0896-6273(03)00438-0. [DOI] [PubMed] [Google Scholar]
  11. Lankheet MJ, Verstraten FA. Attentional modulation of adaptation to two-component transparent motion. Vision Research. 1995;35(10):1401–1412. doi: 10.1016/0042-6989(95)98720-t. [DOI] [PubMed] [Google Scholar]
  12. Mather G, Verstraten F, Anstis S, editors. The motion aftereffect: A modern perspective. MIT Press; Cambridge, MA: 1998. [DOI] [PubMed] [Google Scholar]
  13. Morgan MJ. Wohlgemuth was right: Distracting attention from the adapting stimulus does not decrease the motion after-effect. Vision Research. 2011;51(20):2169–2175. doi: 10.1016/j.visres.2011.07.018. doi:10.1016/j.visres.2011.07.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Morgan M, Chubb C, Solomon JA. Predicting the motion after-effect from sensitivity loss. Vision Research. 2006;46(15):2412–2420. doi: 10.1016/j.visres.2006.01.019. [DOI] [PubMed] [Google Scholar]
  15. Morgan MJ, Dillenburger B, Raphael S, Solomon JA. Observers can voluntarily shift their psychometric functions without losing sensitivity. Attention, Perception and Psychophysics. 2011 doi: 10.3758/s13414-011-0222-7. doi:10.3758/s13414-011-0222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Morgan M, Ward R. Interocular delay produces depth in subjectively moving noise patterns. Quarterly Journal of Experimental Psychology. 1980;32:387–395. doi: 10.1080/14640748008401833. [DOI] [PubMed] [Google Scholar]
  17. Nishida S, Ashida H. A hierarchical structure of motion system revealed by interocular transfer of flicker motion aftereffects. Vision Research. 2000;40(3):265–278. doi: 10.1016/s0042-6989(99)00176-5. [DOI] [PubMed] [Google Scholar]
  18. Raphael S, Dillenburger B, Morgan MJ. Program no. 172.12/II17 2010 neuroscience meeting planner. Society for Neuroscience (online); San Diego, CA: 2010. The effect of attentional modulation on adaptation to transparent expanding/contracting motion. [Google Scholar]
  19. Rees G, Frith CD, Lavie N. Modulating irrelevant motion perception by varying attentional load in an unrelated task. Science. 1997;278:1619. doi: 10.1126/science.278.5343.1616. [DOI] [PubMed] [Google Scholar]
  20. Rees G, Frith C, Lavie N. Processing of irrelevant visual motion during performance of an auditory attention task. Neuropsychologia. 2001;39(9):937–949. doi: 10.1016/s0028-3932(01)00016-1. [DOI] [PubMed] [Google Scholar]
  21. Rezec A, Krekelberg B, Dobkins KR. Attention enhances adaptability: Evidence from motion adaptation experiments. Vision Research. 2004;44(26):3035–3044. doi: 10.1016/j.visres.2004.07.020. [DOI] [PubMed] [Google Scholar]
  22. Rosenthal R. The file draw problem and tolerance for null results. Psychological Bulletin. 1979;86:638–641. [Google Scholar]
  23. Rosenthal R, Rubin D. Interpersonal expectancy effects: The first 345 studies. The Behavioral and Brain Sciences. 1978;3:377–386. [Google Scholar]
  24. Schwartz S, Vuilleumier P, Hutton C, Maravita A, Dolan RJ, Driver J. Attentional load and sensory competition in human vision: Modulation of fMRI responses by load at fixation during task-irrelevant stimulation in the peripheral visual field. Cerebral Cortex. 2005;15(6):770–786. doi: 10.1093/cercor/bhh178. [DOI] [PubMed] [Google Scholar]
  25. Sekuler R, Ganz L. Aftereffect of seen motion with a stabilized retinal image. Science. 1963;139:419–420. doi: 10.1126/science.139.3553.419. [DOI] [PubMed] [Google Scholar]
  26. Sinha D. A experimental study of a social factor in perception: The influence of an arbitrary group standard. Patna University Journal. 1952;7:16. [Google Scholar]
  27. Taya S, Adams W, Graf E, Lavie N. The fate of task-irrelevant visual motion: Perceptual load versus feature-based attention. Journal of Vision. 2009;9(12):1–10. doi: 10.1167/9.12.12. [DOI] [PubMed] [Google Scholar]
  28. Wohlgemuth A. On the aftereffect of seen movement. British Journal of Psychology, Monograph, Supplement. 1911;1:1–117. [Google Scholar]

RESOURCES