| ISLE | Chapter 2
| Signal Detection Experiment
On July 3, 1988, Iran Air Flight 655 was shot down by a U. S. missile from the navy vessel, the U.S.S. Vincennes. The missile killed all 290 passengers and crew, including 66 children. Though the United States has never apologized to Iran, the US government gave $61 million to the relatives of the victims of the attack. The incident took place at a time in which there was heavy tension between the United States and Iran, and Iranian jets had previously attacked US navy vessels. However, in this case, the radar on the Vincennes mistakenly judged the civilian airplane, an Airbus A300, to be an incoming Iranian F-14 fighter. Why had the U.S. Navy made such a terrible mistake?
Consider the radar specialist examining the screen that depicted incoming objects. He or she must decide based on the information available on the radar screen if the incoming object is an enemy warplane (Iran had F-14s) or a harmless jetliner (Airbus A300). How does one tell the difference? In a war-zone with an object approaching you at 600 miles per hour, decisions must be made quickly.
From the U.S. Navy’s standpoint, there are two types of errors that can be made. The Navy could mistake a civilian airplane for a military jet (as they did in this case), or they could mistake a military jet for a civilian airplane. Both of these errors could have fatal consequences for innocent people. In psychophysical terms, the navy’s actual error is called a false alarm, in which a harmless signal is perceived as dangerous. The other error is called a miss, in which a harmful signal is perceived as harmless. In one case, the danger is to innocent civilians aboard the aircraft, whereas in the other case, the danger is to equally innocent personnel aboard the US Navy ship.
In such a situation, there are also two potential correct responses. A correct rejection is when a harmless signal is perceived as harmless, and a “hit” occurs when a harmful signal is correctly perceived as dangerous. In this case, the correct response should have been a correct rejection.
In this experiment, you will play the role of the radar operator. You will look at a circular screen. You will see several dots. Most of these dots will be harmless civilian aircraft, but on some trials you will see one, flying faster (or moving faster across your screen). This faster dot is the target aircraft. You only have a brief period of time. After a few seconds, the screen will go blank, and you must make your decision.
To see the illustration in full screen, which is recommended, press the Full Screen button, which appears at the top of the page.
On this tab you can adjust the parameters of the stimuli to see you they alter your ability to detect the target. The settings include the following:
Number of Background Dots: the number of harmless, nontarget dots that will be presented on each trial.
Relative Speed of Stimulus: how much faster the target is than the background dots (in pixels per update of the screen).
Stimulus Color: how different the stimulus color is from the background dots.
You can control the amount of red in the stimulus.
Duration of Stimulus: how long the moving dots will be presented (sec).
Reset: at the top of the settings page is a Reset button. Pressing this button restores the method settings to their default values.
On this tab you can adjust how the method will work. The settings include the following:
Number of Trials: how many trials will be presented.
Percentage of Trials With Signal: what percentage of the above trials will be have the target stimulus.
Be careful here. Use this value. If the stimulus is rare and you are unsure, what is your best guess?
Reset At the top of the settings page is a Reset button. Pressing this button restores the stimulus settings to their default values.
On the Experiment tab, press the space bar or the Start button
on the screen to start the experiment.
When a trial begins, dots will start moving from the outer edge. On some trials, there is a target dot that moves
faster than the others. When the trial is over, the dots are cleared and you will have buttons to record your responses.
Repond Yes if you see the target resonse and no if you did not.
If you are unsure you must guess, but do so
that gives you the best chance of being correct. If you have a keyboard the Z key will work
for Yes, and the
/ key will work for No.
When you have finished all of your trials, you will be instructed that you can view your results.
Your data will be presented on this tab.
You will see your results in a table.
"His" will be in the upper left box of the table when there was a target
and you said there was.
"Misses" will be in the lower left box of the table when there was a target
and you said there was not one.
"False Alarms" will be in the upper right box of the table when there was not a target
and you said there was.
"Correct Rejections" will be in the lower right box of the table when there was not a target
and you said there was not one.
To see your d' and criterion values, click the button that says Show Threshold.
See the text and other activities for an explanation of these values. To display the data from
the table in a comma-separated format click the Show Data button.
Change the settings below to alter the stimulus parameters in this experiment.