Microsaccade Direction Research Articles

When do microsaccades follow spatial attention?

Following up on an exchange about the relation between microsaccades and spatial attention (Horowitz, Fencsik, Fine, Yurgenson, & Wolfe, 2007; Horowitz, Fine, Fencsik, Yurgenson, & Wolfe, 2007; Laubrock, Engbert, Rolfs, & Kliegl, 2007), we examine the effects of selection criteria and response modality. We show that for Posner cuing with saccadic responses, microsaccades go with attention in at least 75% of cases (almost 90% if probability matching is assumed) when they are first (or only) microsaccades in the cue-target interval and when they occur between 200 and 400 msec after the cue. The relation between spatial attention and the direction of microsaccades drops to chance level for unselected microsaccades collected during manual-response conditions. Analyses of data from four cross-modal cuing experiments demonstrate an above-chance, intermediate link for visual cues, but no systematic relation for auditory cues. Thus, the link between spatial attention and direction of microsaccades depends on the experimental condition and time of occurrence, but it can be very strong.

Microsaccade directions are not correlated with cued locations in a spatial working memory task

Microsaccade directions are not correlated with cued locations in a spatial working memory task

Fixational eye movements predict the perceived direction of ambiguous apparent motion

Neuronal activity in area LIP is correlated with the perceived direction of ambiguous apparent motion (Z. M. Williams, J. C. Elfar, E. N. Eskandar, L. J. Toth, & J. A. Assad, 2003). Here we show that a similar correlation exists for small eye movements made during fixation. A moving dot grid with superimposed fixation point was presented through an aperture. In a motion discrimination task, unambiguous motion was compared with ambiguous motion obtained by shifting the grid by half of the dot distance. In three experiments we show that (a) microsaccadic inhibition, i.e., a drop in microsaccade frequency precedes reports of perceptual flips, (b) microsaccadic inhibition does not accompany simple response changes, and (c) the direction of microsaccades occurring before motion onset biases the subsequent perception of ambiguous motion. We conclude that microsaccades provide a signal on which perceptual judgments rely in the absence of objective disambiguating stimulus information.

Microsaccades and Attention: Does a Weak Correlation Make an Index?

In the recent literature on microsaccades and attention, two questions have been conflated. There is a broad question of whether microsaccades are related to attention, and there is a more specific question about whether microsaccades serve as an index of attention. We are happy to agree that microsaccades are related to attention. However, the claim that ‘‘microsaccades are an index of covert attention’’ (in the title of Laubrock, Engbert, Rolfs, & Kliegl, 2007, this issue) depends on a strong correlation. Were this claim true, a researcher might be able to conduct a study relying entirely upon microsaccade direction as a measure of attentional deployment. Although we would be delighted to be able to conduct such an experiment, both our data and those of Laubrock et al. suggest that microsaccades cannot be used as a reliable marker of covert attention. Laubrock et al. make two arguments. First, reviewing our experiment (Horowitz, Fine, Fencsik, Yurgenson, & Wolfe, 2007, this issue), they criticize our selection of trials on which the directions of the cue and microsaccade disagreed, arguing that this selection would inevitably lead to a negative microsaccade-congruency (MC) effect. Second, they report new data demonstrating a correlation betweenMC and reaction time (RT). We deal with these points in reverse order. Laubrock et al. have demonstrated a statistically reliable relationship between MC and RT. An incongruent microsaccade was associated with a 6-ms slowing of RT. This is a weak effect, an order of magnitude smaller than the 81-ms effect associated with an invalid cue. This effect may be statistically significant, but it suggests that microsaccade direction provides very little useful information about the spatial distribution of attention. Also, Laubrock et al. argue, under a seemingly reasonable set of assumptions, that in our study, trials on which the microsaccade direction diverged from the cue direction were dominated by trials on which the microsaccade did not follow attention, even if microsaccades usually did follow attention. The argument is as follows. Assume that observers direct attention toward the cue with probability w, and that the microsaccade reflects the direction of attention with probability x. Let v denote cue validity. There are two kinds of trials on which the cue is valid but the cue direction and microsaccade direction disagree: (a) valid trials (v) on which attention does not follow the cue (1 w) and the microsaccade follows attention (x) and (b) valid trials (v) on which attention follows the cue (w) but themicrosaccade does not reflect attention (1 x). The proportion of trials of the first type is given by p15 v(1 w)x, and the proportion of trials of the second type is given by p25 vw(1 x). Laubrock et al. note that if w5 v and x5 .75 (i.e., the microsaccade is almost as good an index of attention as the cue), the predictions would be qualitatively consistent with our results. However, this scenario is not quantitatively consistent with our results. Although w and x are not directly observable, one can observe the proportion of trials on which the cue direction and microsaccade direction disagree, p5 p1 1 p2 5 v(w1 x 2wx). Because v is known (arrow cues were 80% valid, so v 5 .80), any observed p is compatible with a line through wx space. Figure 1 plots the wx curves that could produce the observed ps in the manual-detection condition of our experiment for all 3 observers (data from the other two conditions lead to similar conclusions). The diamond represents the hypothetical point on which Laubrock et al. base their argument (w 5 .80, x 5 .75); this point is clearly not consistent with the data. In fact, if we assume that observers frequently shifted attention in the direction of the cue (i.e., w .60), then the probability that the microsaccade followed attention must have been less than .55 (note that if x 5 .50, then the direction of the microsaccade is independent of attention). If observers were at least probability matching (i.e., w .80), then x would have been less than .52. Thus, the predictive power of microsaccades is, for practical purposes, negligible. Address correspondence to Todd S. Horowitz, Visual Attention Laboratory, Brigham and Women’s Hospital, 64 Sidney St., Suite 170, Cambridge, MA 02139, e-mail: toddh@search.bwh.harvard.edu. PSYCHOLOGICAL SCIENCE

Fixational Eye Movements Are Not an Index of Covert Attention

The debate about the nature of fixational eye movements has revived recently with the claim that microsaccades reflect the direction of attentional shifts. A number of studies have shown an association between the direction of attentional cues and the direction of microsaccades. We sought to determine whether microsaccades in attentional tasks are causally related to behavior. Is reaction time (RT) faster when microsaccades point toward the target than when they point in the opposite direction? We used a dual-Purkinje-image eyetracker to measure gaze position while 3 observers (2 of the authors, 1 naive observer) performed an attentional cuing task under three different response conditions: saccadic localization, manual localization, and manual detection. Critical trials were those on which microsaccades moved away from the cue. On these trials, RTs were slower when microsaccades were oriented toward the target than when they were oriented away from the target. We obtained similar results for direction of drift. Cues, not fixational eye movements, predicted behavior.

Microsaccade directions do not predict directionality of illusory brightness changes of overlapping transparent surfaces

Tse (2005) recently introduced a new class of illusory brightness changes where shifts of attention lead to shifts in perceived brightness across overlapping, transparent figures, under conditions of visual fixation. In the absence of endogenous attentional shifts, illusory brightness changes appear to shift from figure to figure spontaneously, much as occurs in other multistable phenomena. The goal of the present research is to determine whether fixational microsaccades are correlated with perceived brightness changes. It has recently been demonstrated that microsaccades can reveal the direction of covert attentional shifts either toward (Engbert, R. & Kliegl, R. (2003). Microsaccades uncover the orientation of covert attention. Vision Research, 43, 1035–1045; Hafed, Z. M. & Clark, J. J. (2002). Microsaccades as an overt measure of covert attention shifts. Vision Research, 42(22), 2533–2545) or away from (Rolfs, M., Engbert, R., & Kliegl, R. (2004). Microsaccade orientation supports attentional enhancement opposite a peripheral cue: commentary on Tse, Sheinberg, and Logothetis (2003). Psychological Science, 15(10), 705–707) a peripheral cue under certain circumstances. Others (Horwitz, G. D. & Albright, T. D. (2003). Short-latency fixational saccades induced by luminance increments. Journal of Neurophysiology, 90(2), 1333–1339; Tse, P. U., Sheinberg, D. L., & Logothetis, N. K. (2002). Fixational eye movements are not affected by abrupt onsets that capture attention. Vision Research, 42, 1663–1669; Tse, P. U., Sheinberg, D. L., & Logothetis, N. K. (2004). The distribution of microsaccade directions need not reveal the location of attention. Psychological Science, 15(10), 708–710) found no change in the distribution of microsaccade directions as a function of where attention is allocated, although changes in the rate of microsaccades were observed in all of these studies in response to the onset of attentional reallocation. It is therefore possible that the distribution of microsaccade directions will change as a function of which figure is perceived to darken, or that changes in this distribution predict which figure will subsequently darken. We find no correlation between this distribution and which figure undergoes the effect, and therefore conclude that microsaccade directionality is not influenced by and does not influence which figure undergoes the effect. Moreover, the directions of microsaccades that occur immediately prior to a perceptual switch are not correlated with the perceived position of the figure that undergoes the effect. However, we do find that the rate of microsaccades decreases upon a perceptual switch, signifying an attentional shift coincident with the perceptual shift. We conclude that microsaccade directionality does not determine, predict, or cause which figure will subsequently be perceived to undergo an illusory brightness change.

Are you ready? I can tell by looking at your microsaccades

The direction of microsaccades has been shown to be biased by the allocation of spatial attention. Here, we investigated whether the cognitive processes involved in preparing to respond to an upcoming target can modulate the microsaccadic response. Specifically, we found that optimal manual response preparation, reflected by faster response times, was associated with a reduction in the absolute frequency of microsaccades. The present results are consistent with previous studies suggesting a relationship between oculomotor activity and different sorts of motor responses. Our findings, however, surprisingly demonstrate that the effect of preparation and stimulus expectation extends to an automatic and unconscious oculomotor activity such as microsaccade execution.

Microsaccades Counteract Visual Fading during Fixation

Our eyes move continually, even while we fixate our gaze on an object. If fixational eye movements are counteracted, our perception of stationary objects fades completely, due to neural adaptation. Some studies have suggested that fixational microsaccades refresh retinal images, thereby preventing adaptation and fading. However, other studies disagree, and so the role of microsaccades remains unclear. Here, we correlate visibility during fixation to the occurrence of microsaccades. We asked subjects to indicate when Troxler fading of a peripheral target occurs, while simultaneously recording their eye movements with high precision. We found that before a fading period, the probability, rate, and magnitude of microsaccades decreased. Before transitions toward visibility, the probability, rate, and magnitude of microsaccades increased. These results reveal a direct link between suppression of microsaccades and fading and suggest a causal relationship between microsaccade production and target visibility during fixation.

Microsaccade dynamics during covert attention

We compared effects of covert spatial-attention shifts induced with exogenous or endogenous cues on microsaccade rate and direction. Separate and dissociated effects were obtained in rate and direction measures. Display changes caused microsaccade rate inhibition, followed by sustained rate enhancement. Effects on microsaccade direction were differentially tied to cue class (exogenous vs. endogenous) and type (neutral vs. directional). For endogenous cues, direction effects were weak and occurred late. Exogenous cues caused a fast direction bias towards the cue (i.e., early automatic triggering of saccade programs), followed by a shift in the opposite direction (i.e, controlled inhibition of cue-directed saccades, leading to a ‘leakage’ of microsaccades in the opposite direction).

The Distribution of Microsaccade Directions Need Not Reveal the Location of Attention: Reply to Rolfs, Engbert, and Kliegl

The Distribution of Microsaccade Directions Need Not Reveal the Location of Attention: Reply to Rolfs, Engbert, and Kliegl

Binocular quantification and characterization of microsaccades.

The significance of microsaccades in the visual process has been discussed for more than 50 years. However, only a few studies have measured microsaccades binocularly, and detailed quantification and characterization of these small movements are needed in order to further understand their nature. The amplitude, velocity, acceleration and direction of microsaccades were quantified binocularly in 10 normal test persons during a 40-s fixation task, using an infrared recording technique. All microsaccades for all test persons were performed simultaneously and individually with an almost identical amplitude in the right and left eye (a range of 0.003-0.042 deg between right and left eye mean values). The mean microsaccadic amplitude for the test persons was within a range of 0.223-1.079 deg. The directional difference between simultaneously-performed right and left eye microsaccades was less than 22.5 deg for 84.8% of the saccades, indicating that the majority of microsaccades are conjugated. Three different fixation patterns were identified and characterized: (1) a classic interplay between easily identified drifts and medium-sized microsaccades (mean amplitude range 0.328-0.413 deg); (2) long intersaccadic intervals (4-5 s) with almost absent drifts, followed by three or four large microsaccades (mean amplitude range 0.755-1.079 deg); and (3) low-amplitude drift movements interrupted by low-amplitude microsaccades (mean amplitude range 0.231-0.265 deg). Microsaccades are involuntary, predominantly conjugated, simultaneously performed, and of almost identical amplitude in the right and left eye, suggesting a central control mechanism for microsaccades at subcortical level.