Statistical Data Measuring Scales

Data Measuring Scales 

In statistical data, there are four types of measuring scales. Types are 

1. Nominal

2. Ordinal

3. Interval

4. Ratio

1. What is nominal data type?

It is qualitative. No mathematical operation. Can not be measured. Only category/classification is possible. No ordering here.

Example,

Male, Female, Divorced, Married, widow, eye colour, Gender, Religion etc.

It is impossible to order religion. Rather helpful for getting a Pie chart and frequency.

2. What is the Ordinal data type?

Nominal data type features + more features.

It is Qualitative. It is possible to sort data. 

Example,

Very unhappy, Unhappy, Ok, happy, very happy

But the problem is that it is impossible to measure the amount between very unhappy and happy. 

Roll 1, 2, 3, 4, 5

Here, the roll number is ordered but the number gap is not equal between rolls 1 and 2 etc.

3. What is the Interval data type?

It is quantitative. Categorized, Ordered, Addition and Subtraction are possible.

All features of ordinal + more features. 

The thermometer is an example of interval-type data. 0 degree, 10 degrees, 20 degrees, 30 degrees etc.

30-20= 10 degree. 

Here interval same but it does not ensure that the comfort level is not a multiplication of 20 degree, 30 degrees and at 40 degrees.

Say, for example, One student can not answer of 10 questions in a math exam. And got 0 marks on the exam. That does not mean that student does not know mathematics. So here o marks are not true zero.

4. What is the Ratio data type?

It is quantitative. Multiplication and Division are possible. All mathematical information is possible. i.e. +, -, * and / are possible. Here 0 = True zero. For example speed =0 means it is stationary. 

All features of interval + more features. 

Highest and most informative scale. 

Most measurements in the physical sciences and engineering are done by ratio scale.

Contain all qualities of the scales like nominal, ordinal and interval scales.

Example-

height

weight

marks

distance

Here if I tell that height = 0 then actually no height

Here if I tell that weight = 0 then actually no weight

Here if I tell that marks = 0 then actually obtained a 0 mark

Here if I tell that distance = 0 then actually no distance

Here 0 means absolute zero.

***************************************************************

When handling SPSS software and viewing variable view window then 

Nominal

Ordinal

Scale ( Both Interval and Ratio)

Classify variables in the statistical scale of measurement:

data variables	Nominal	Ordinal	Interval	Ratio
sex (male, female)	Y
group (science, commerce)	Y
occupations (farmer, business, job, other)	Y
meaningful order	X	Y	Y	Y
social order (very satisfactory, satisfactory, neutral, unsatisfactory, very unsatisfactory	X	Y	Y	Y
meaningful difference	X	X	Y	Y
true zero	X	X	X	Y

 www.rashid-bd.info