Measurement ScaleBefore planning a questionnaire or survey, test developers identify the scale of measurement of their data to make sure that they are in alignment with the objectives of the study. Th

More on Scales of Measurement PSY3700 Multimedia Assessment and Psychometrics ©20 16 South University 2 More on Scales of Measurement Scales of Measurement Scales of Measurement Nominal : This scale of measurement is used when you have clear categories, and people fall into one or another, but cannot fall into more than one. A good example of nominal data is sex. Sex is composed of just two categories, male and female. People are biolog ically male or they are biologically female. You cannot be both. This is contrasted with the psychological variable of gender. Gender is measured on a continuous scale, and people may fall anywhere along a continuum of masculine to feminine. In the population, people (males or females) identify themselves anywhere from extremely masculine to extremely feminine. In the middle, there is a construct we call androgyny, where they are balanced in terms of masculinity and femininity. So, with gender ther e are no set categories and we measure gender on a continuous (interval or ratio) scale rather than a nominal scale. This is true with most psychological variables, such as intelligence and attractiveness. People do not fall neatly into categories. Int elligence, for example, is most often measured on a continuous scale which is converted to a standardized IQ score. This score can range anywhere up to 150 or more, and a person's score, although generally steady, can change from year to year. Other examples of nominal scaled variables include grouping variables, such as race, ethnicity, experimental/control group. You will find that nominal scales of measurement are used most often in psychology to group research participants in one way or another. Ordinal : This scale of measurement is used to rank order a variable in some way. A good example of an ordinal, or ranking, variable is year in high school. Here, students are ranked by whether they are Freshmen, Sophomores, Juniors, or Seniors. As with categories, each individual or piece of data falls into just one. A person cannot, for example, be both a freshman and a sophomore. Ranking differs in that things are generally ordered from lowest to highest; you have a definite hierarchy in your data. You can think of ordinal data as categories placed on a hierarchy. An example on a questionnaire or survey might be when a participant is asked to rank order their liking of different products or traits according to most liked, average, or least liked. As with nominal data, you have no variability in the data, something that is very important when running statistical analyses. You can use ordinal data to group individuals, but it most often does not make sense to find a measure of variability. Inter val and Ratio : Both interval and ratio scaled variables are variables that are measured on a continuous scale. You might, for example, have several questions about attractiveness that you ask participants to respond to on a Likert -type scale of 1 to 5. Those responses will be added together to find a total rating of attractiveness. Each individual will vary in their responses to this type of scale, based on individual differences. What this means is that you have scores ranging along a continuum - ther e are no neat categories to fit them into. Most psychological constructs are measured using continuous scales so the data can be quantified. These scales differ in that the Ratio scale has an absolute zero. PSY3700 Multimedia Assessment and Psychometrics ©20 16 South University 3 More on Scales of Measurement Scales of Measurement © 201 6 South University