Ratio scale has all the characteristics to make it a pure scale especially for the research related to pure sciences. Ratio scale has all the characteristics of interval scale plus the additional benefit that it has a well-defined zero. It also takes into account the ratios between different values of the variables.
Characteristics of ratio scale
Ratio scale has order and equally distanced values between units and at the same time has an absolute zero as well. It can quantify the difference between each value and ratios can also be calculated. Though, unlike interval scale, this scale does not have negative numbers because of absolute zero. To know whether a variable can be measured on the ratio scale one should see that whether it meets all the requirements for the interval scale if it does and it has an absolute zero then it can be measured on the ratio scale.
- A ratio scale has a rational zero point where the value of the variable has no value. For example, weight and height can be measured on ratio scale because they have a real zero value. On the other hand, the temperature cannot be measured on ratio scale because a zero degree centigrade temperature does not mean that there is no heat or cold. Most of the scientific variables like height, weight, age, density, strength, time, volume, speed can be measured on a ratio scale.
- A rational zero is necessary to measure the ratio between two values of a variable, in the absence of a real zero there is no ratio. In the absence of a zero, does it make sense to say that a student who got 100 marks in mathematics has got double as compare to the one who got 50?
- A scientific variable cannot be measured precisely on nominal, ordinal or interval variable, in the absence of a ratio scale, scientific variables cannot be measured.
Statistical analysis of ratio scale
A wide range of all descriptive and inferential statistics can be applied on the ratio scale. All statistical analysis including mean, median, mode, variance, standard deviation, and the range can be calculated on a ratio scale. In addition, statistical tests like t-test, f-test, correlation, chi-square etc can also be calculated on ratio scale variable.
For detailed overview of the measurement scales read Types of Measurement Scales in Research
- Gideon Vigderhous, The Level of Measurement and Permissible Statistical Analysis in Social Research, Pacific Sociological Review, Vol. 20, No. 1, 1977, pp. 61-72.
- Campbell MJ, Machin D, Wiley J. Medical Statistics: A Commonsense Approach, Vol. 2. London: Wiley; 1993.