Why is it important to understand levels of measurement, and what are the different levels of measurement? Define and provide an example of each level of measurement.

Question

What will be an ideal response?

Answer 1

o Measurement involves following rules for assigning numbers to attributes.
o We assign numbers so that we can analyze and make better sense of information.
o We assign numbers to information, sometimes arbitrarily, and the numbers often serve as placeholders and do not represent amounts or magnitudes of anything.
o When we speak about level of measurement, we are referring to the relationship among the numbers we have assigned to the information.
o Understanding levels of measurement is critical to describing and interpreting psychological test and measurement results because we use numbers, at the test (or survey) item level, the scale level, and the test result level. Tests and surveys are made up of items to which individuals respond.
o After a person takes a test or completes a survey, we score the individual items. For tests, the individual items are typically combined to produce a score on a scale and/or combined to produce a test result, which may be a test score or a description. For example, a knowledge test may include multiple-choice test items that are individually scored as correct or incorrect, and then the number of correct items is summed to create an overall raw test score.
o A personality test may include test items where an individual must choose from one of two possible answers, and then the answers are scored to reveal a description of a person’s personality.
o How we describe or report test or survey item-level results and the claims we can make about psychological test results depends on the properties (the level of measurement) of these numbers.
o Nominal – numbers are assigned to represent labels or categories of data only.
o Ordinal – numbers are assigned to rank-order data, but the distances or values between numbers are not equal and can vary
o Interval – numbers are assigned to rank-order data, the distance between numbers is judged to be equal, but there is no absolute zero point indicating absence of what is measured
o Ratio – numbers are assigned to rank-order, the distance between the numbers is equal, and there is an absolute zero point indicated absence of what is measured.