1. Traits versus Issues: Factor versus Ideal-Point Analysis of Candidate Thermometer Ratings
2. Those interested in seeing the methods developed here at work will find an application to the Chilean Senate given by Londregan (2000) Estimating Legislators’ Preferred Points
3. A few technical conditions are necessary: (i) the random variable ξνρ in inequality (2) can take on any value on the real line, albeit with a very low probability; (ii) there is “no guessing,” so that at very high levels of mastery the probability that the subject answers correctly converges to 1, while at abysmal levels it converges to 0; and (iii) random disturbances to the subject's response, caused by factors such as momentary distractions during the test, are uncorrelated across questions. Of these three assumptions, the second is the most problematic: we might expect low-ability subjects to guess.
4. This amounts to allowing the a parameter to differ among voters.