In the multistable perception domain, many studies have found that directing fixation or spatial attention to different regions can bias the perception of the Necker cube (Toppino, 2003; Kawabata, 1986, 1987; Meng & Tong, 2004; Xu & Franconeri, 2010). The explanation in these studies is that the fixated or attended region would look closer, so the perception is biased. However, no study has examined this assumption directly. On the other hand, in research concerned with attention, few studies have mentioned whether spatial attention can affect depth perception. Some studies have found that spatial attention can influence many properties of stimulus perception, which may relate to depth perception (e.g., Carrasco, 2006; Carrasco, Williams, & Yeshurun, 2002; Gobell & Carrasco, 2005). Thus, it is possible for spatial attention to affect multistable perception through the depth mechanism.
The purpose of this study is to investigate the effect of spatial attention on multistable figure perception and the possible underlying mechanism of this effect.
The diamond stimulus was used to test the hypothesis that spatial attention can bias multistable motion perception by making attended areas look closer. Four experiments are designed to examine this hypothesis and the main results will be briefly reviewed first.
‧
Spatial attention can bias multistable perception
The effect of spatial attention was found to be very robust at normal conditions in Experiments 1a, 1b, 3, and 4, making the percentage of time perceiving coherent motion significantly higher in the attending-to-occluders condition than that in the attending-to-moving-lines condition. This result shows that spatial attention, like the effect of fixation, can bias multistable perception. However, this result still cannot directly support that spatial attention biases multistable motion perception by affecting depth perception. Accordingly, in Experiment 2, this question was verified directly.
Spatial attention can alter perceived depth slightly
Experiment 2 was a depth-judgment task. The same two attention conditions and nine different levels of binocular disparity between the occluders and the moving lines were manipulated. Results show that the mean probability of reporting lines behind occluders in small binocular disparities is significantly higher under the attending-to-occluders condition than that under the attending-to-lines condition. This result is consistent with the hypothesis that attention can affect depth perception, making attended areas look closer. Thus, it is more likely that the effect of spatial attention on multistable motion perception found in Experiments 1a and 1b occurs
‧
through the mechanism of making the attended areas look closer in depth.
Binocular and monocular depth cues can weaken the effect of spatial attention
Experiments 3 and 4 were designed to test whether the effect of spatial attention on biasing multistable motion perception can be blocked with binocular or monocular depth cues that define the depth relationship between occluders and moving lines. If the effect of spatial attention can be affected by the depth cues, it is more possible that attention is biasing multistable motion perception through depth mechanisms. The results of Experiment 3 show that the effect of spatial attention is only significant at binocular disparity 0 (occluders and lines were in the same depth) but not significant at the other three binocular disparity conditions (lines behind the occluders in three different levels). The results of Experiment 4 show that the effect of spatial attention is blocked when the monocular depth cues (by filtering out high spatial frequency information) define the occluders in front of the lines. In summary, these results are all consistent with the prediction of the hypothesis that spatial attention is biasing multistable motion perception through the depth perception mechanism. Further discussions will follow, including the integration of the results of this study and other relative studies as well as possible mechanisms of spatial attention on multistable perception.
‧
Spatial attention versus intention
The relationship between the mechanisms of spatial attention and intention on multistable perception is an interesting issue worth investigating. Although these two mechanisms are all top-down processing under our cognitive control, they seem to act in different ways. The results of Experiment 1a show that the effects of spatial attention and intention on multistable motion perception are additive, implicating they are independent mechanisms. In addition, one of my other experiments, not included in this study, investigated whether participants would allocate their spatial attention on occluders or lines in order to maintain demanded perceptions. Intention was manipulated in a between-participant design, as in Experiment 1a, and the accuracy of continuous probe detection of occluder lightening or line lightening was measured.
The results show that the difference of detecting occluder lightening under the two intention conditions is not significant (F(1,10) = 0.117, p = 0.740). Also, the difference of detecting line lightening under the two intention conditions was not significant (F(1,10) = 0.177, p = 0.683). On the other hand, the percentage of time perceiving coherent motion in “hold coherent” condition (63.29 %) was higher than that in the “hold separate” condition (37.07 %) (F(1,10) = 8.23, p < 0.05, partial η2 = 0.457), which is consistent with the intention control effect found in Experiment 1a.
These results imply that participants did not need to attend to particular areas to exert
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
their intentional control. Instead, intentional control is more likely to be exerted through top-down priming of wanted representation, as Toppino (2003) mentions.
However, in Experiment 1 of the aforementioned Tsal and Kolbet’s study (1985), they found that observers tend to attend to the focal area of the interpretation that they had to maintain. The reason may be the short presentation time of the multistable figure (30 milliseconds). Directing their spatial attention to the focal area would be a more effective and faster way to perceive the demanded perception than just top-down priming of wanted representation.
Spatial attention versus depth cues (top-down vs. bottom-up mechanism)
In the four experiments of this study, the mechanisms of spatial attention and intention on multistable perception belongs to top-down processes, while binocular and monocular depth cues belong to bottom-up factors. The entire theoretical framework of the study is presented at Figure 18. Since all these factors are found having influence on multistable perception in this study, it consists with the hybrid model mentioned previously, which assumes both bottom-up and top-down processes together contribute to the multistable figure perception.
‧
Figure 18. Theoretical framework of the study. According to Palmer (1999), depth
perception is formed from integrating four kinds of depth information: ocular information (including accommodation and convergence), dynamic information (such as motion parallax and optical flow), binocular information (also called stereoscopic information), and monocular depth cues (also called pictorial information). The former two types of information are not discussed in this study. Both bottom-up processing (depth perception) and top-down processing (intention and attention) contribute to depth-reversible figure perception.
An important issue worth investigating is how these top-down and bottom-up factors act together on multistable perception. It seems that spatial attention can
Depth‐reversible
‧
For example, in Experiment 2, spatial attention has influence on perception—making attended areas look closer—only when the manipulated binocular disparities are small.
In Experiment 3, the effect of spatial attention shows only at the binocular-disparity-0 condition, not at other large-binocular-disparity conditions. In Experiment 4, the manipulated monocular depth cues block the effect of spatial attention. These pieces of evidence all imply that spatial attention can exert its affection only when the depth cues of stimulus (such as the depth relationship of the occluders and the lines in the diamond stimulus) are not well-defined. A possible reason is that bottom-up depth cues are more dominant than attention in affecting depth perception. Thus, the effect of spatial attention is easily overridden by bottom-up depth cues.
Mechanisms of spatial attention on multistable perception
The mechanism of spatial attention on multistable perception is complicated.
Some parts of the effect of spatial attention are like the effect of fixation found previously (e.g., Meng & Tong, 2004; Toppino, 2003). For example, the results of Experiment 1a show that the effect of attention is comparatively smaller than the effect of intention. In addition, it has no interaction with the effect of intention, implying that the effect of spatial attention and intention are independent mechanisms in multistable perception. This part of the effect of spatial attention may be involved
‧
in affecting bottom-up processing of monocular depth cues, perhaps by increasing perceived contrast, spatial frequency, and the spatial resolution of attended areas (e.g., Carrasco, Williams, & Yeshurun, 2002; Carrasco, 2006; Gobell & Carrasco, 2005).
Through this, the attended areas look closer and affect multistable perception.
Another piece of evidence supporting this notion is that monocular depth cues can almost block the effect of spatial attention in Experiment 4, implying that attention may act through this way. To summarize briefly, the effect of spatial attention on multistable perception is because it affects its bottom-up processing (depth perception), as shown in Figure 18.
However, it is still unclear whether there are other ways for attention to affect multistable perception through the depth perception mechanism. For example, attention may affect the integration of different sources of depth information, or attention itself may be a source of depth information. The main question comes from the uncertainty of how different sources of depth information are combined into a single coherent representation. For example, it is not clear whether different sources are independent processes, whether they interact, or whether their integration is additive or multiplicative (Palmer, 1999). Further studies are needed to discover a clearer and complete role of spatial attention in the depth perception mechanism.
‧
Limitations and restrictions
One limitation in the manipulation-check task of spatial attention in Experiments 1b, 2, 3, and 4 is that the probe showed in every trial. Thus, it seems possible for participants to give the probe-detecting response in every trial, even though they did not see the probe. However, there are three reasons that can demolish this possibility. First, there is no trial of which the RT to the probe is shorter than 400 ms, and trials of which the RT to the probe are longer than 1500 ms are very rare. It implies that participants responded to the probe due to seeing it. The second reason is that the showing of the probe is so obvious that it is hard to create a false-alarm response or miss it. The third reason is that by the observation of the experimenter, participants did not select the probe-detecting response randomly or repetitively in each trial. Hence, for these reasons, it is unlikely the participants downhill choose the probe-detecting response when not seeing the probe.
Another limitation of this study is that nearly half the data are deleted due to the negative “index of attention” in Experiments 2, 3, and 4. The large abandoned data may affect the whole pattern of results and the interpretations of it. However, it should be noted that it is not easy to sustain spatial attention on four (rather than one) occluders or four moving lines for a period of time, not to mention that participants also have to hold coherent motion perception and report their perception at the same
‧
time in Experiments 3 and 4. It is necessary to use a strict criterion—the index of attention—to rule out the data of which participants did not allocate their attention on demanded areas. Thus, it is more convincing that the retained data were manipulated validly.
In summary, all four experiments in this study support the notion that spatial attention can bias multistable motion perception by affecting depth perception, making attended areas look closer. However, this notion is restricted only to depth-reversible figure perception. The effect of spatial attention on other multistable figure perception may involve mechanisms other than depth perception.
Contribution and future studies
How attention can influence or mediate brain processing of visual stimuli and the possible underlying mechanism driving it is an important issue that interests many researchers. In the multistable perception domain, the results of this study support the assumption that spatial attention can bias depth-reversible figure perception by making attended areas look closer. This assumption had been proposed in many studies of the Necker cube, but it is examined directly in this study. This study is also contributive in understanding how top-down and bottom-up factors act together on multistable perception. In future studies, an issue worth investigating is that whether
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the effect of spatial attention on depth-reversible figure perception is only via the depth mechanism or if there are other ways attention can influence depth-reversible figure perception. In addition, the effect of spatial attention on the multistable figures other than the depth-reversible figure is also an interesting topic. The effect of spatial attention and the underlying mechanisms of them can be compared with depth-reversible figures.
In the depth perception domain, few studies have mentioned whether or how attention affects depth perception, which is one of the most important pieces of information with respect to vision. This study also provides some evidence that spatial attention can affect depth perception by means of the direct, subjective depth-judgment task (Experiment 2) and the indirect multistable-perception task (Experiment 1, 3, and 4). In future studies, the degree to which spatial attention can affect depth perception should be examined more precisely by using stimuli other than multistable figures. Also, it is also important to clarify the relationship between spatial attention and other depth cues and how they are integrated into a final depth perception result in further studies.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
‧
Anton-Erxleben, K., Henrich, C., & Treue, S. (2007). Attention changes perceived size of moving visual patterns. Journal of Vision, 7(11), 1-9.
Bisiach, E., Ricci, R., Lai, E., De Tanti, A., & Inzaghi, M. (1999). Unilateral neglect and disambiguation of the Necker cube. Brain, 122(1), 131-140.
Carrasco, M. (2006). Covert attention increases contrast sensitivity: Psychophysical, neurophysiological, and neuroimaging studies. Progress in Brain Research, 154, 33-70.
Carrasco, M., Fuller, S., & Ling, S. (2008). Transient attention does increase perceived contrast of suprathreshold stimuli: A reply to Prinzmetal, Long, and Leonhardt (2008). Perception & Psychophysics, 70(7), 1151-1164.
Carrasco, M., Ling, S., & Read, S. (2004). Attention alters appearance. Nature Neuroscience, 7(3), 308-313.
Carrasco, M., Williams, P. E., & Yeshurun, Y. (2002). Covert attention increases spatial resolution with or without masks: Support for signal enhancement.
Journal of Vision, 2(6), 467-479.
Cohen, J. (1988). Statistical Power Analysis for the Behavioral Science (2nd ed.):
Hillsdale, NJ: Lawrence Erlbaum Associates.