Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
tutorials:sdt [2016/04/11 18:37]
justin
tutorials:sdt [2016/04/11 18:39] (current)
justin
Line 8: Line 8:
  
 A key idea that motivates the theory of signal detection is that you want to determine sensitivity (how good are subjects at detecting that faint candle light), but that there is an unaccounted for cognitive factor - what criteria the subject uses. A subject can have a very conservative criteria (only say the light is a candle if you are very sure). This will lower false-alarm rates, but then you may make more misses. A subject alternatively can change their criteria so that they are less prone to missing, but then they will make more false-alarms. Signal detection theory allows you to compute sensitivity and criteria separately from subject responses (i.e. the hit and false-alarm rates) so that you can determine how sensitive a subject is regardless of what arbitrary criteria they used. A key idea that motivates the theory of signal detection is that you want to determine sensitivity (how good are subjects at detecting that faint candle light), but that there is an unaccounted for cognitive factor - what criteria the subject uses. A subject can have a very conservative criteria (only say the light is a candle if you are very sure). This will lower false-alarm rates, but then you may make more misses. A subject alternatively can change their criteria so that they are less prone to missing, but then they will make more false-alarms. Signal detection theory allows you to compute sensitivity and criteria separately from subject responses (i.e. the hit and false-alarm rates) so that you can determine how sensitive a subject is regardless of what arbitrary criteria they used.
 +
 +For more background, check-out David Heeger'​s [[http://​www.cns.nyu.edu/​~david/​handouts/​sdt/​sdt.html|signal detection theory handouts]].
  
 We will simulate a signal-detection experiment. On each trial, our observer sees an element sampled from either the signal present gaussian distribution or the signal absent distribution,​ which is also gaussian with the same standard deviation. The observer chooses to say "​signal present"​ when the signal they see on that trial is above criterion and "​signal absent"​ otherwise. The picture you should have in your head is this: We will simulate a signal-detection experiment. On each trial, our observer sees an element sampled from either the signal present gaussian distribution or the signal absent distribution,​ which is also gaussian with the same standard deviation. The observer chooses to say "​signal present"​ when the signal they see on that trial is above criterion and "​signal absent"​ otherwise. The picture you should have in your head is this:
Line 14: Line 16:
 {{:​tutorials:​nobias.png|}} {{:​tutorials:​nobias.png|}}
 <​html></​div></​html>​ <​html></​div></​html>​
- 
  
 ====== Make signalPresentAbsent array ====== ====== Make signalPresentAbsent array ======