Analysis
Explanation
References
The line motion experiment offers students a chance to experience a novel illusion and then measure its magnitude.The illusion is the perception of line motion where there is none, an illusion that is created by preceding a line display with an attentional cue that is at one end of the line. For example, if the line below were displayed without an attentional cue to precede it, the line would be correctly seen to have appeared in a single piece, instantaneously on the screen. But if the line is preceded by a brief attention grabbing cue at one of its ends, the line is perceived to grow from that point, thus creating the illusion of left to right or right to left motion.
If you have the Authorware Web Player plug-in in the browser you are using to view this page, you can experience the illusion by clicking here. The demonstration presents a line instantaneously on the screen without motion, but there will be the perception of motion. The misperception is due to the attentional cue which appears to the right or left of the line just before it is displayed.
A question of interest in the study of illusory line motion is the temporal magnitude of the effect. A simultaneous line like the one in the demonstration has a “growth parameter” set to 0 msec."This merely says that it grows in zero time." But what if we alter that parameter, gradually incrementing it."Now the lines grow left to right or right to left across the screen. We do this in the Line Motion experiment by using the same line as before but we set it in motion like an arrow across the screen." It begins nearly off screen with just its leading edge showing." Then we move it onto the screen to an end position that is exactly the same as for the simultaneous line presentations." The only difference is that rather than achieving this position in 0 msec, it occurs in a time set by the line growth parameter. Some motions are extremely fast occurring in 15 msec thus creating a display that is just barely distinguishable from the simultaneous display. Others are relatively slow, occurring in 95 msec. The speed with which the line moves to its final position on the screen is the independent variable for our experiment." It has values ranging from 15 to 95 msec in steps of 10 msec.
The question to be answered by manipulating line speed (the
line growth parameter) is how slow can a line be drawn on the screen and still produce the
illusion that it grew not from its true side of origin but from the side indicated by the
attentional cue. That is, when a cue is presented in the 20 X 20 pixel space on the right,
and then the line moves from the left toward the position of the cue, will the observer
judge the line motion correctly or incorrectly? If it is judged to move from the cue
position, the illusion is present. If it is judged to move from the true position, the
illusion has disappeared, The question then is what line growth speed is needed to destroy
the illusion. This speed can be taken as an estimate of the temporal bias that is
introduced by cueing.
« Return to Top»
he data are best treated by separating the simultaneous line data from the growth line data.The simultaneous line data serve to demonstrate the presence of an illusion. One-fifth of the trials in the experiment will be simultaneous line trials.Thus if users elect 100 trials there will be about 20 simultaneous trials. Approximately half of these will have been preceded by a cue on the right and half by a cue on the left. The judgments that will have been made are that the line moved “Left to right”” “Right to left motion,” or “No Motion.” The dependent variable of interest will be the percent of judgments that reflect perceived motion from the side where the cue appeared. For most research participants this value will be near 100%.The data may look like this:
| Perceived origin of motion | Percent |
Cue side |
95% |
| Non-cued side | 0% |
No motion (Simultaneous) |
5% |
Once the simultaneous line data have been examined to verify that the research participant did in fact experience an illusion, the growth line data can be examined to determine the line speeds that sustained the illusion. Because an illusion can only be detected when cues appear on the side opposite from the origin of the line motion, three fourths of the growth line trials are “cue opposite” trials. Thus in a 100-trial experiment, on average there will be 80 growth line trials of which 60 will be trials on which the illusion might occur. The speed of line growth on these trials will be a speed selected from the 15, 25, 35, 45, 55, 65, 75, 85, and 95 msec set, but the probability of selection is biased to favor the faster speeds. Once again, the judgments will be categorical—“Left to right,” Right to left,” or “Simultaneous (No motion).” The dependent variable is the percent of correct judgments for each of the line speeds. The percent correct data may look like this:
Line Speed (msec) |
|||||||||
| Trial Type | 15 | 25 | 35 | 45 | 55 | 65 | 75 | 85 | 95 |
| Cue Opposite | 5% | 20% | 45% | 66% | 85% | 85% | 100% | 100% | 100% |
| Cue Same | 100% | 100% | 100% | 100% | 100% | 100% | 100% | 100% | 100% |
These data indicate that the probability of making a correct judgment on cue opposite trials gradually increased as line speeds slowed down. Alternatively, one could say that the strength of the illusion diminished as line speeds got slower.
In order to obtain a single summary data point for these data, one can estimate the line speed that would produce optimal conflict between true and incorrect perceptions, which is the line speed estimated to produce 50% correct and 50% incorrect judgments. This value can be interpreted as the value that reflects the temporal bias introduced by attentional cues. On an analogy with the concept of “point of subjective equality” used in classical psychophysics, this value can be called the “point of subjective simultaneity” (PSS).
To compute the PSS, begin by noting the line speed that produced the last percent value below 50%, which in the sample data is 35 msec. This value needs to be incremented by some value Dt to obtain the line speed that is estimated to produce 50% correct. The increment is not 10 msec because in the data 45 msec speeds produced 66% correct judgments It is a proportion of this 10 msec increment, a proportion that can be computed by taking he difference between 50% and 45% and putting it over the difference between 66% and 45%. Thus
|
This solution
assumes a monotonic, linear increase across line speeds in the probability that observers
will experience an illusion.
« Return to Top»
Hikosaka, O., Miyauchi, S. & Shimojo, S. (1992). Focal visual attention produces illusory temporal order and motion sensation. Vision Research, 33 (9), 1219-1240.
Hikosaka, O., Miyauchi, S. & Shimojo, S. (1993).Voluntary and stimulus-induced attention detected as motion sensation. Perception, 22, 517-526.
Downing, P.E. & Treisman, A.M. (1997). The line-motion illusion: Attention or impletion. Journal of Experimental Psychology: Human Perception and Performance, 23 (3), 768-799.
Stelmach,L.B. & Herdman,C.M.(1991). Directed attention and perception of temporal order. Journal of Experimental Psychology: Human Perception and Performance, 17, 539-550. (This is really a better reference for the temporal judgment study)
Zanker, Johannes M. (1997). Line Motion Illusion: Is Facilitation Responsible for the ‘Motion Induction’ Effect ? Retrieved Aug, 28, 2004, from http://cvs.anu.edu.au/johannes/mot_ill.html#linmot
http://cvs.anu.edu.au/johannes/mot_ill.html#linmot
http://www.illusionworks.com/html/line_motion.html llusion Works site
Last revised:September 07, 2004 09:06:07 AM
Copyright© 1995 The University
of Mississippi. All rights reserved.
Questions about this page?PsychExps