Читать книгу Statistics and Probability with Applications for Engineers and Scientists Using MINITAB, R and JMP - Bhisham C. Gupta, Irwin Guttman - Страница 44

Interval data are numerical data, more informative than nominal and ordinal data but less informative than ratio data. A typical example of interval data is temperature (in Celsius and Fahrenheit). Arithmetic operations of addition and subtraction are applicable, but multiplication and division are not applicable. For example, the temperature of three consecutive parts A, B, and C during a selected step in a manufacturing process are F, F, and F, respectively. Then we can say the temperature difference between parts A and B is different from the difference between parts B and C. Also we can say that part B is warmer than part A and part C is warmer than part A, but cooler than part B. However, it is physically meaningless to say that part B is three times as warm as part A and twice as warm as part C. Moreover, in interval data, zero does not have the conventional meaning of nothingness; it is just an arbitrary point on the scale of measurement. For instance, F and C (=F) have different values, and they are in fact the arbitrary points on different scales of measurements. Other examples of interval data are year in which a part is produced, students' numeric grades on a test, and date of birth.