Problem
A researcher was interested in studying trait hostility in postal workers. The researcher had each postal worker nominate two people to provide ratings of that postal worker's trait hostility. Over all the participants, the pairs of raters correlated .40.
i. What is the reliability of individual coders?
ii. If the researcher averages the two ratings of trait hostility for each participant, what would the reliability of the average be?
iii. In a future study, how many coders would the researcher need to collect ratings from in order to achieve a reliability of .80?