The table can be rewritten with these and Hit rate H : proportion of YES trials to which subject respondedįalse alarm rate F : proportion of NO trials to which subject responded These are conventionally the Hits and False Alarms, and these are then givenĪs proportions of the row totals, which are in turn viewed as estimates of
#Signal detection theory explained plus#
Plus the total numbers of trials, are needed to characterize a subject's performance. So, only 2 of the 4 numbers in the table (1 per row), if there areĢ0 different trials, and a subject has 5 hits, then that subject must The value in one column from the value in the other. The NO response column, and neither number can be known in advance.īut, if you know the number of YES and NO trials in the experiment, you know YES response column will not necessarily be the same as the total number in Row (or, more generally, the total for each row is known in advance from theĭesign of the experiment) however, the total number of responses in the Number of responses in the top row will equal the total number in the bottom Notice that if the number of Signal (Different)Īnd No-signal (Same) stimuli are the same in the experiment, then the total “Yes” here represents the presence of a signal orĭifference to be detected our different and same labels areĪdded for convenience in thinking about AX discrimination: Viewing such an experiment, and naming the possible outcomes, is as follows. Who try to detect all and only the signals. An experiment presents signals and non-signals to subjects, To another signal), and model how a perceiver decides whether a signal is Using detection theory, we conceive of sensitivityĪs (broadly) detecting a signal (e.g. Macmillan and Creelman's 1991 Detection Theory: A User's Guide (known The presentation that follows comes directly from We are interested in, while bias is what we have to take into account to (Signal) Detection theory attributes responses to a combination of sensitivity and bias.
To the same pairs can be used as an indication of response bias.Ģ. It becomes meaningful when interpreted in terms of the listener's responseīias, or tendency to respond "same" or "different". Pairs alone is not a very meaningful measure of discrimination. The point is that % correct on the different This pattern of results is common in AX discrimination Some of the different pairs (small step sizes, within-category pairs).Ĭlearly you might nonetheless be discriminating better than a person who On the same pairs, but you will probably have 0% correct on at least Not sure you consistently respond "same". Is, you don't respond at random when you do not hear a difference, or are Only do so when you are quite sure that the stimuli are different. One - suppose that you are very conservative in answering "different", and The pairs very well? Clearly not you don't even have to have listenedĬompare this response strategy with an opposite Does this result, 100% correct, mean that you discriminated Same pairs are not analyzed at all, and this response strategy would You would of course also get 0% correct on the same pairs,īecause you answered "different" to all of them.
To every item, and you would then get 100% correct on the different Task and you wanted to show 100% discrimination. Suppose you were a subject in a discrimination But there is no guarantee that listeners will do that.
Sure, they respond "same" or "different" randomly, so that performance isĪt chance. File assumes that when listeners do not hear a difference, or are not
0 kommentar(er)
