How does an implicit reaction time test work?

The commercial test is itself based on the evaluative priming paradigm in academic research (e.g., Fazio, et al., 1986)1. The first phase of the test is to detect target emotion words as belonging to either one category (e.g., Happy) or another (e.g., Sad).  On each trial in this first phase, the word appears and the respondent has to press one key for happy words and another key for sad words. This is a very easy task, and respondents can do this very quickly with few errors.