1) is your test a scientist?
This test was created in a way empirical and experimental probably like the test d'Ishihara, the assistance of data processing was decisive.
I largely used the simulation programs of the vision of an existing daltonien on Internet.
the selection of the reactive colors was difficult especially to obtain a certain consensus while checking on vischeck and https://www.toptal.com/designers/colorfilter/.
| Test your vision of the colors |
|
2) have retreat for this test?
A statistical study on more than 12.000 answers is significant.
82% of the answers comprised any question randomly.
1st résultat:Il seems that the test is particularly sensitive.
3) sensitive in which direction?
Sensitive above all to the quality of the screen.
The error count is important:
by considering the 9364 which answered the 31 questions out of 12098 tests, there are only 27% which do not make any error and only 66% make less than 3 errors to the 31 questions.
There on the other hand, to be located in the 8% of this sample, it is necessary to make at least 10 errors out of 31 questions... it is surprising
4) and why 8%?
This is a value which approaches the commonly allowed proportions as the table below indicates it.
frequency of deficiencies of the vision of the colors among Europeans
(modified from Wright, 1953) |
type of color blind |
man |
woman |
| 1- monochromatism |
trés rare |
trés rare |
| 2- dichromatism |
2.105% |
0.08% |
protanopia |
1.0% |
0.02% |
deuteranopia |
1.1% |
0.01% |
tritanopia |
0.005% |
0.003% |
| 3- trichromatism |
5.9% |
0.40% |
protanomaly |
1.0% |
0.02% |
deutéranomaly |
4.9% |
0.38% |
tritanomaly |
assez rare |
assez rare |
total |
8.0% |
0.46% |
5) how to explain this difference?
I do not explain it completely, I could not find the origin of this table of proportions of daltoniens and by which method it was established in 1953.
With regard to my test, it is obvious that the bad quality of the screens or a false calibration plays a part inevitably, which can explain 3 or 4 frequent errors.
This test east cannot be not entirely a test of daltonism but a test of screen also.
In addition 14,9% of the testers think of being daltoniens without knowing the type +1,57% state to be it and know the type, which seems to indicate a strong proportion of testers daltoniens.
Another hypothése, given that the test does not separate man-woman, it is possible that women tetrachromates (see here ) appear in the statistics.
6) which are the most frequent errors?
- For the errors of the personal test of color :(49% of the errors)
here n° by order of most frequent with the least frequent: n°8-16-20-28-24-12-4 (particuliéres questions to highlight the deutéranomalie, the protanomalie, the deutéranomalie)
- For the errors for the test of squaring :(51% of the errors)
here n° by order of most frequent with the least frequent: n°26-31-17-1-2-19-14-22-23-4-6-21-13-29-10-5-18-25-7-27-30-3-15-9-11.(questions to highlight deuteranopia, protanopia, the tritanopie).
7) certain questions are not obvious why?
15% declare more than one question not obvious, probably because of the screen or the incomprehension of the name of the color.
To remember the proposal well: " choose by elimination, the answer nearest ".
| See also the table of the colors "not obvious". |
|
8) how to interpret the results?
If the proportion of 8% of daltoniens is true and corresponds to the panel, one would have to take into account only the tests with a score lower than 22/31.
On the other hand so at least 15% of the testers were well daltoniens as they annoncaient it, it is necessary to take into account a score inferior to 24/31.
| score higher than 25 |
score between 21 and 25 |
score lower than 21 |
| nonsignificant |
to confirm |
probable daltonism |
Will know that:
94% make less than 2 errors /20 on the tests of deuteranopia
95% make less than 2 errors /17 on the tests of protanopia
97% make less than 2 errors /15 on the tests of tritanopie
94% make less than 3 errors /7 concerning the deutéranomalie on the test of the names of colors.
98% make less than 2 errors /7 concerning the protanomalie on the test of the names of colors.
99% make less than 2 errors /7 concerning the tritanomalie on the test of the names of colors.
9) how is it made that certain indicators are contradictory?
I am not explained all the results, that occurs sometimes in the case of scores weak.
the software of simulation is theoretical and no daltonien sees in the same way.
it is necessary in this case to refer has the decreasing list of the most frequent errors (see question n° 7) and to allot less importance to them.
A hypothése, it is possible that women tetrachromates (see here ) appear in these contradictions (to be confirmed).
10) which are your last councils?
I repeat, to gauge its screen well, in case of doubt, to remake the test on a better screen and especially to remember that this test anything medical, and you must consulted your ophthalmologist regularly. |
