If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

# Signal detection theory - part 2

Signal detection theory explores how we distinguish signals from noise. The video discusses two key variables: d-prime, the difference between noise and signal distributions, and C, the individual's strategy. Different strategies (B, D, C, and beta) are used to set thresholds for signal detection. Ideal observers use the C strategy, while others may be more liberal or conservative. Created by Ronald Sahyouni.

## Want to join the conversation?

• What are you talking about?
• This is a terrible video, please take it down or redo it
• if d'=1 B=2, D=b'- B, then shouldn't D= 1-2??
• So apparently, B is just an arbitrarily chosen threshold. The mistake he made (I think) is that he decided to change the value for B just for strategy D, simply because d' already equaled 1, and he just decided to change it randomly. For the rest of the strategies, he observed B = 2 as he had originally started.
• If they screwed up the video with this many mistakes they might as well redo it.
• love khan, but this one very confusing
• This video might be helpful.

• C = A measure of the observers ability to correctly identify a stimulus (e.g. a sight, a sound.... w/e), based on a given "signal", in the presence of and equally present "noise" (overlapping but undesired/interfering stimulus). "B" is just a quantitative value for where (on an intensity scale) that individual can differentiate the signal from the noise. d' is just a measure of how similar the signal and the noise are.
• I'm glad everyone else thought this video didn't make sense
• Do we really have to know this stuff? I've never heard of it before and it doesn't seem like the type of thing they'd typically ask