Experiment-
Date-
Attentional blink
Theory
Cognitive mechanisms can handle only limited amounts of information. In many situations,
there are more stimuli and mental events than resources for processing. As a result, some
stimuli are processed and some are not. Selectively choosing some stimuli and ignoring
others is called attention. Many studies of attention investigate how mental resources are
switched from one stimulus to another. This experiment explores some properties of attention
with rapidly changing stimuli. It shows that there is a brief time after paying attention to one
stimulus when attention cannot be focused on a subsequent stimulus. This duration is called
an attentional blink because it is analogous to being unable to see objects during an eye blink.
In the experiment, many letters are shown in rapid succession, with each letter overwriting
the previous letter. The observer's task is to watch the entire sequence and then indicate
whether certain target letters were in the sequence. The sequences are carefully constructed to
systematically vary the temporal separation between two target letters. Thus, if the target
letters are J and K, a sequence with the letters ". . . JXTVRK . . ." places K five letter-spaces
beyond J.
The notable finding is that identification of the second target letter is very low when it
quickly follows the first target letter. As temporal separation increases, identification of the
second letter improves. This finding suggests that when the observer sees the first target
letter, he/she must attend to it to ensure that it will be remembered later. The focusing of
attention to that letter apparently requires time, and if the second target letter appears during
that time, it is not attended and not reported. By looking at recognition of the second letter as
a function of separation, we can estimate the time required to focus and break attention for
stimuli.
Instructions
A window will fill the entire screen, and a smaller window will appear with abbreviated
instructions. Close the instructions window. You can open it again later from the Lab Info.
menu.
Start a trial by pressing the space bar. A sequence of 19 letters will appear, with each new
letter overwriting the previous one. Each letter is presented for only 100 milliseconds. After
the letter sequence is finished, you will be prompted to make a response. Do not enter a
response before the sequence is finished, or the trial will have to be discarded (and run again).
Your task is to determine whether the letter J and/or K were in the sequence of letters just
presented. You indicate the presence of a J by pressing the j key and you indicate the
presence of the K by pressing the k key. If you see both letters, press each in any order. If you
did not see either letter, do not press either key. The screen will present the key you press, but
this is only so you know the program received your key press; it is not feedback on whether
you were correct. When you are ready for the next trial, press the space bar again.
If, for some reason, you need to discard a trial (e.g., you sneezed, your attention drifted, you
"zoned out," et cetera), press the t key before pressing the space bar for the next trial. This
will discard the just run trial (It will be re-run later in the experiment).
At the end of the experiment, the experiment window will close and a new window will
appear that displays your data as a table and a plot (if appropriate) and provides an
explanation of the experiment and results. You can print this information, save it as an html
file, or save it in CogLab format. The latter format can be re-opened with CogLab on a CD
and by your instructor who may want to combine data from students in your class.
Results/Observations:
subject 1
Separation of targets First target Second target
0
2
4
6
8
subject 2
Separation of targets First target Second target
0
2
4
6
8
The expected result is that the percentage of reports for the first letter do not vary much with
separation. For the second letter, there should be few reports at separation zero, and the
percentage of reports should increase with separation.
Conclusion:
