# Inter-subject signal sensitivity estimation?

I am using paired comparison methods to estimate estimate respondents’ sensitivity to distinctions in words used to characterize the construct of agreeableness. (e.g, Undemanding = .34 versus Peaceful = .72). The contrast between the words' factor loadings is the signal to be detected. The stronger loading was alternated between being on the first and second word. I did find that the contrast predicted probability of correct detection (r = .20*).

# Experiment background

The stimulus is paired comparison words with each word having a different factor loading (from prior research) based on it's measurement of agreeableness. So each word taps the construct more or less deeply than the other. The contrast between the words in terms of their factor loading is the signal. A set of 122 such paired comparison tasks was developed and the task before the respondent is to select the more agreeable word.

The respondents are students recruited for the research having varying degrees of sensitivity. Each word in the pair only has a factor loading as the estimate of signal and the contrast should be the salient feature in the task.

# How to get sensitivity estimate?

I have been using the number of correct “larger” loading detections as the respondent’s score. How might I estimate respondents' sensitivity to the presented signals?

• Estimating sensitivity is different from enhancing sensitivity. Which are you asking about? Jun 18, 2017 at 14:36
• I am most interested in estimating the sensitivity given the data I have to work with (i.e. , 122 paired comparing trials with different signal strengths at each trial) Jun 19, 2017 at 15:33
• I guess I'm on my own.... Jun 22, 2017 at 1:22
• I do not understand how the experiment is developed, but to know the sensibility, like the Theory Detection Signal, you should do a table of double entry: stimulus x answer, with yes and not for both and from this, you must analyze the probability and variance. Jul 28, 2017 at 23:57
• @hexadecimal that sounds like an answer Jul 29, 2017 at 18:01