MEASUREMENT OF THE

ILLUSION OF
MÜLLER-LYER
LIDIA ESTEBAN BELCHI
DNI: 48693580 G

1. INTRODUCTION
This practice allows us to delve into classical psychophysical methods and come closer.
to the world of visual illusions through the famous Müller-Lyer illusion.

The Müller-Lyer illusion is a type of geometric illusion that consists of perception

incorrect length of an image with arrows, that is, it is the optical illusion that
presents when observing two lines with arrows at the ends with different orientations.
On one side of the line the arrows point inward, while on the other side the arrows
They lead to the ends of each side, as can be seen in figure 1.

Figure 1. Müller-Lyer illusion.

Although it is still not exactly known what causes the effect of the illusion of
Müller-Lyer, various authors have contributed different theories, the most popular being the
theory of perspective.

In the three-dimensional world, we often tend to use angles to estimate the

depth and distance. Our brain is accustomed to perceiving these angles
like nearby or distant corners, at greater or lesser distance; and this information
it also uses it to make judgments about size.

The arrows of the Müller-Lyer illusion are interpreted by our brain as

distant and nearby corners, canceling the information from the retina that tells us that
both lines are of the same length.

In this experiment, we will consider the degree of as independent variables.

opening of the arrow tips: 30, 60, 90, 120, and 150, and the position of the arrows,
horizontal or vertical; while the dependent variable is the perception we have
about the length of the segments. This perception is measured through the difference of
length of the line that we modified or traced based on the segment
model.

It is expected, on one hand, that the subjects draw longer lines when the degree of the
the tip of the arrow is less than (<90) and, on the contrary, they will draw a line of lesser
length when the opening degrees of the tips of the variable segment are greater
(>90).

2. METHOD
In this experiment, the psychophysical method used to measure the illusion is the method
for adjustment. A software program was used: [Link], which presents
different stimuli model the subject, initially placed horizontally and later
vertical, at whose ends there are two arrow tips with a certain amplitude
between 30º and 150º, from this model, the subject must adjust the variable stimulus to
that both segments have the same length.

The degrees of the arrow opening and the position of the segment are the variables.
independent of the experiment. On one hand, the degree of openness has five values
different: 30º, 60º, 90º, 120º and 150º. On the other hand, the values of the position are
horizontal or vertical. Therefore, there are 5 different experimental conditions that are
they have used for each position, with 10 trials each (100 in total) presented from
random forms, which are administered intrasubject.

The dependent variable in this experiment, as has been mentioned in the

introduction, it is the perception that the subject has about the length of the segments, the
which manifests itself through the length of the drawn line as a function of the segment
model. This variable is measured in pixels.

Finally, the controlled extraneous variable is the screen resolution as it is the same.
for all subjects. However, there are other strange variables that have not been
I have been able to control in this experiment, such as fatigue, the vision problems of the
subjects or possible distractions during the session.

3. RESULT AND DISCUSSION

For the analysis of the results obtained in the experiment, the

mean, median, and standard deviation.

In the obtained data, it has been necessary to eliminate "outliers." The "outliers" are data
that stray far from what is expected, and this affects the average. These data can be given
when the subject answers incorrectly and they must be removed and the mean recalculated. In
In this case, there were 2 incorrect results in the horizontal position and 1 in position.
vertical.

a.) Analysis of accuracy

Table 1. Average horizontal position measurements

HORIZONTAL 30 60 90 120 150

Lidia 141.75 130.40 99.70 85.3 81.1
Casandra 140 125.40 98.2 76.5 72.7
Daniela 144.3 121.7 100.1 71 61.6
Angela 149.3 128.5 101.1 83.9 78.22
MEDIA WHORE143.8375 126.5 99,775 79,175 73,405
Table 2. Average measures vertical position

VERTICAL 30 60 90 120 150

Lidia 148.5 133.2 102.2 99,8 90.1
Casandra 133.8 116.3 98.9 76.9 67.9
Daniela 142.9 125.3 105.4 80.7 68.8
Angela 139,5 126.8 103.8 94 82.4
MEDIA TO SEE 141,175 125.4 102,575 87,85 77.3

Figure 2. Subjective equality point graph

As we can see in Figure 2, the group mean is positioned above the line.
red, which represents the length of the model segment (100 pixels) when the opening
the angle of the variable segment is 30º and 60º and, therefore, the degree of opening
the model segment is larger (150º), with the 30º being the highest average (141.175)
(pixels), this is true for both horizontal and vertical positions. This pattern is repeated
until reaching 90º, where the subjects' responses are more similar to the model.
When the opening degree of the variable segment arrow is less than 90º, and, for
Thus, the degree of opening of the model segment is greater than 90º, the subjects tend to
respond to shorter lengths and the group average lies below the line
red.

b.) Precision analysis

Table 3. Standard deviation of horizontal position

HORIZONTAL 30 60 90 120 150

Lidia 15,30 8.20 5.29 9.55 6.27
Casandra 20.78 5.50 1.47 3,68 4.69
Daniela 6.96 7,57 4.79 3.19 6.04
Angela 7.31 8,16 4.58 7.88 7.37

Table 4. Standard deviation vertical position

VERTICAL 30 60 90 120 150

Lidia 6.9 8.71 4.91 7.39 4.01
Casandra 4.31 3.86 2.23 4.01 4.67
Daniela 8.06 5.05 5.64 6.41 4,98
Angela 7,13 5.26 7.08 5.61 5.21
Figure 3. Horizontal position accuracy graph
 Figure 4. Vertical position accuracy graph

Comparing figures 3 and 4, we observe that the subjects show more accuracy.
when the segment is in a horizontal position, we can also highlight
how subjects have a lower standard deviation when the degree of openness of the
The arrow tips are 90º, which reflects greater accuracy.

Observing the data, we conclude that, although all subjects have had a high
standard deviation we can say that Casandra would stand out among all for being the one who
lower score has in the standard deviation, therefore, the one with the greatest accuracy has
held in the experiment.

c.) The effect of the Müller-Lyer illusion

The obtained data corroborate the initial hypothesis. As we have observed in Figure
the subjects tend to draw longer lines when the degree of opening of the
the variable segment arrow is less than 90º, while when the degree of opening
is greater than 90º, subjects tend to draw shorter lines.

The difference between the subjects' data, although not very noticeable, is indeed
they appreciate some, as in the case of Casandra which has a higher accuracy than the
rest of the subjects.

After analyzing the data, we can confirm that the Müller-Lyer illusion affects both
accuracy as with precision, as we observe how the results show that the
variable segments differ significantly in all instances from the segments
models in both accuracy and precision.

4. CONCLUSION

In conclusion, we can confirm that in this experiment we have obtained the

expected results, the Müller-Lyer optical illusion will appear with a longer segment if
the arrow at the end points outward, and is smaller if it points inward,
and therefore, the Müller-Lyer illusion holds true and affects our perception.

5. REFERENCES

Pedraja, M. J. (2023). Practice 2: On the Müller-Lyer illusion. Slides of

PowerPoint. University of Murcia, Murcia.

Aso, U. (2019) Psychology and Mind. Müller-Lyer Illusion: What it is and why it is
produce. [Link]

6. ANNEXES

HORIZONTAL POSITION
Opening (in degrees) of the tips of the E variable
30 60 90 120 150
136 97 85 74
155 140 92 78 84
Longit
145 139 100 103 78
of the
154 120 96 81 94
the
137 108 95 76
line
152 120 95 78 87
trazad
117 126 96 85 74
a 154 120 105 95 79
pixels) 122 135 105 81 82
135 131 103 72 83
Media141.75 130.4 99.7 85.3 81.1
Media
na 148.5 133 98.5 83 80.5
Desv.15,30406108 8,2083967848 5,2925524193 9,5574287569 6,2795965900
tpica5682 6474 0688 6411 1542
 VERTICAL Position
Opening (in degrees) of the tips of the E variable
30 60 90 120 150
138 147 107 110 96
145 126 97 97 83
Longitude
151 134 99 106 87
of the 149 131 94 107 93
line 163 138 105 91 94
traced 149 100 92 92
(en 145 125 100 98 86
pixels 156 145 109 90 89
144 122 104 107 92
145 131 107 100 89
133.2222222222
Media
148.5 22 102.2 99.8 90.1
Median
a 147 131 102 99 90.5
Desv.6,996030620605 8,714610974934 4,917090377222 7,390684827688 4,012480529547
tpica12 25 87 31 78