by Kevin Larson
translated by Juan Jose (Wally)
Leo at Microsoft Research a paper on how people recognize read the words as these ...
Editor's Note:
Before, I would like to clarify that this translation is very free compared to the original paper does not intend to overwhelm you with pages and pages on the subject, but rather give you an idea of \u200b\u200bhow humans have the read, the different models of recognition and some other data more. Therefore I recommend that if you could read the original paper will be a lot more complete than this brief summary.
Several studies
movable window
This technique restricts the amount of text visible to a certain amount of numbers and letters around the fixation point, and replace the remaining letters with a ' , x '. The task of the reader is simply reading the page of text. It is also possible to do the reverse and just replace the letters of the point of fixation with the letter 'x', but this is very frustrating for the reader. If you replace only the first three letters on the right and left of the fixation point with an 'x', the read speed drops to 11 words per minute. Some studies like McConkie and Rayner (1975) examined how many letters are needed to replace about fixing point to provide a normal reading experience. In the table below is a sample of what you would see the lectro your were reading a text passage and is set at the second ' e' of experiment. If the reader is provided with three points beyond the point of fixation, it can not see the word experiment completely, and your reading speed will as slow as 207 words per minute. If the reader is provided with nine points beyond the point of fixation, this may see the word experiment completely, and part of the word WAS. With 9 letters, reading speed is reduced only moderately. If it provides the reader with up to 15 points beyond the point of fixation, the speed will be as fast as if it were not moving window (changes character x). Up to 15 points is a linear relationship between the number of letters available to the reader and reading speed.
| Window Size | Prayer | Speed \u200b\u200bReading |
| 3 letters | xxx xxxxxxxxx xx experimxxx An | 207 ppm |
| 9 letters | An experiment wax xxxxxxxxx xx | 308 ppm |
| 15 letters | An experiment WAS condxxxxx xx | 340 ppm |
linear relationship between the number of letters available in the window and reading speed.
Through this study we learn that our perception space is about 15 letters. This is interesting because the average step size is 7-9 letters, or about half of our space perception. This indicates that while readers are recognizing nearest words to the point of fixation (fovea), use additional information farthest this to guide reading. It is important to note that we only take the information from the floor to the right of the fixation point, not the left. In the case of the table above, when the reader's fixation point is in the second
and of experiment, removing the word An, reading speed is not impacted.
The moving window study demonstrates the importance of letters in reading, but not as the only important element. The word model is also expected that reading speed decreases as the information on the form of the word disappears. This model makes the prediction that reading improves significantly if always maintained form of words. This turns out to be false.
The following diagram shows the reading speed when there is only three letters. This is equivalent to the speed of reading when the word set is shown in full, despite the whole word has more letters available 0.7 on average. When the word set and the next word are fully available, the read speed is comparable to when it has 9 nine letters. The speed is also equivalent to when there are three words or 15 letters. This means that reading is not necessarily more rapid when subsequent words have completely, can achieve reading speeds similar when you have a few letters.
| Window Size | Prayer | Speed \u200b\u200bReading |
| 3 letters | An experimxxx xxx xx xxxxxxxxx | 207 wpm |
| 1 word (3.7 Lyrics ) | An experiment xxx xxxxxxxxx xx | 212 wpm |
| | | |
| 9 letras | An experiment wax xxxxxxxxx xx | 308 wpm |
| two words (9.6 letters ) | An experiment was xxxxxxxxx xx | 309 wpm |
| ; |
|
|
| 15 letras | An experiment was condxxxxx xx | 340 wpm |
| 3 palabras (15.0 letras ) | An experiment was conducted xx | 339 wpm |
Links:
Documento original (en ingles) en Microsoft Research
0 comments:
Post a Comment