This you can find the 4 scales of measurement – the Nominal scale, ordinal, interval and ratio scale.
Scores on a test, questionnaire, or assessment instrument represent the assignment of numbers to individual items or the total test in order to classify the individual or represent the standing of an individual on some variable. That is measurement – the assigning of numbers to the attributes of individuals, events or objects according to a set of logical rules. Specific sets of rules define a scale of measurement; four of these scales – nominal, ordinal, interval and ratio – are used to quantify variables in the behavioral sciences.
When numbers stand for names or categories that represent the way individuals differ, that is nominal measurement. Numbers in the nominal scales are arbitrarily assigned to code a specific variable. Here are some from interview schedules.
– never married
Highest educational level completed:
– none of the following responses
– high school diploma or GED
– vocational, business, or technical school
– associate of arts or sciences degree
– bachelor’s degree
– graduate degree
When individuals or objects are ordered or ranked according to some characteristic, which is ordinal measurement. For example, a counselor who is a group facilitator might order his students along an introversion/extroversion continuum. The client who is the most introverted might be assigned the rank of 1 and the most extroverted client the highest rank. Or the percentile of individuals on an achievement test could be ranked from high to low or low to high.
Many items on schedules and questionnaires require a personal ranking on a certain characteristic. Here is just one example. Please rate you degree of competence in the listed behaviors, using the following key: 1 = low, 3 = average, 5 = extremely high.
1. Oral communication skills 1 2 3 4 5
2. Written communication skills 1 2 3 4 5
3. Listening skills 1 2 3 4 5
There are problems with the ordinal scale because the competencies may not be equally spaced. An individual might be extremely strong in certain areas, with minute differences distinguishing those competencies, and extremely weak in some others. Problems also can occur in comparing rankings across groups. The persons with number one rankings in each group might vary tremendously on the variable being ranked. Thus, in rank ordering, the length of the intervals can be unequal. And the numbers used for ranking do not reflect anything quantitative about the variable being ranked. Caution is needed in performing arithmetical operations on measurements derived from ordinal scales and in using and interpreting the information.
When a scale differentiates among levels of an attribute and has equal distances between those levels, which are interval measurement. It uses numerical values that are equally spaced. Examples of interval scales occur in educational and psychological measurement. We often treat the scores on aptitude and ability tests as belonging to the interval scale; we assign numbers to levels of an attribute and assume that equal differences in the numbers correspond to equal differenc3s in the attribute. That approach may be meaningful with a variable like temperature, but we can have problems with test results. Can we say a score of 0 indicates an absence of a trait or characteristic or that someone with an IQ of 100 has twice as much as intelligence as a person having and IQ of 50?
When a true or absolute zero point exists in addition to ranking and equal intervals, which is ratio measurement. Measure such as height and weight are examples of ratio scales. Response time is often an important ratio measurement in interpreting standardized tests: Two minutes are twice as much as one, and there is a true zero point. Both the interval and ratio scale have interval properties and utilize the same statistical procedures.