Measurement in Education, as in any human field is essentially concerned with assigning numerical values to observations. This assignment of values is usually done in accordance with well specified rules. Hence it is systematic.
It is done at various levels depending on what is being measured, how it is being measured, what instruments are involved and desired level of accuracy and precision. These various levels at which measurement can be carried out are described as scales of measuring instruments or levels of measuring scales. These are nominal scale, ordinal scale, interval scale and ratio scale.
Data obtained on these scales are also described as nominal, ordinal, interval and ratio data, respectively. The type of scale determines the level of refinement of the data obtained and the appropriate statistical treatment to which the data can be subjected.
Norminal scale, this is the simplest and least refined scale. Measurement at this level only involves assignment to classes or categories. No category is greater than or less than the other. The scale lacks the property or magnitude.
Ordinal scale unlike the norminal scale where there is no order of magnitude. The ordinal scale possesses the property of magnitude. On this scale, it is not only possible to classify members of a group, it is also possible to compare any two such members in terms of relative magnitude or size that is less than, equal to or greater than.
Consider the grades received by two students in school certificate mathematics. The grade of 1 is higher than grade 2, as one who made grade 3 did better than the one who made grade 4. Similarly one who scored grade 2 is not necessarily twice as good as one with grade 4. In other words equal intervals do not represent equal qualities.
Interval scale, the property of order is present. In addition equal intervals on the scale represent equal amounts of the attributes being measured. However this scale does not possess an absolute zero.
Ratio scale, this is most refined scale. It has the properties of order, magnitude and equal interval. The property of absolute zero is present. For instance, measuring height with a meter rule. There is zero on the rule which corresponds to complete absence of the attributes(ie height) height measured.
statiscal tools have done more for us than make it practical to measure the reliability of results. They have provided methods of testing technique and have made possible the design of more efficient and more comprehensive experiments.
Content created and supplied by: Nedved (via Opera News )