collier

• look at average responses
• look at many different topics

• puts numbers on individuals
• usually are unidimensional

• advantages: dont limit the responses
• disadvantages: have to be coded by th researchers which is difficult and time consuming

• advantages: dont have to coded
• can be directly analyzed statistically
• disadvantages: limit the responses to what the options given by the test designer

• items do not jave the same numver of responses to compute the correlation between each other
• convenient to have many items with the same set of responses because they can get used to the answer scale and dont have to read it every time

easier to compute when the items have the same number or responses

• bipolar resonse sets muxt be symmetrical
• ordered response sets should be in the right order
• responses need to be mutually exclusive and exhaustive

• the responses dont overlap
• more than one cant be true (unless able to pick more than one response)

all possible answers are included

• you dont want the question to have multiple interpretations
• otherwise different participants might interpret the quesrions differently and your responses will be a mixture

• these are two questions packed into one
• each question might have a different answer

ex. do you not like spicy foods

ex. have you stopped beating your wife?

• reversed scored items are pairs of items that are scored in the opposite direction of each other
• sometimes used to check for aquiescence
• it's helpful to revesrse score some items so that they all "point" in the same direction

the more specific, concrete and the more recent the question, the more reliable and valid the answer will be