In the process of measurement, it is considered that the measuring tools and instruments used are standard and used to calibrate and verify the measured tools and instruments, so as to measure and ensure the reliability of the measurement results obtained when using the measured tools and instruments.
There are three data distribution characteristics of measurement data: concentration trend (involving quantity: mean, geometric mean and median), dispersion degree (involving quantity: range, percentile and quartile spacing, variance, standard deviation and coefficient of variation) and distribution shape (normal distribution and skewed distribution); Statistical inference of measurement data includes parameter estimation and hypothesis testing.
Extended data
Statistical description:
1, concentration trend
(1) arithmetic mean is the average level, which is widely used and is most suitable for symmetric distribution, especially normal distribution.
(2) The geometric mean is the average increase (decrease) times, which is applied to the equal ratio data and lognormal distribution;
(3) The median is an observation level with a middle rank, which is suitable for skewed distribution, unknown distribution or distribution with no definite value at the end.
2. Dispersion
The commonly used indicators to describe the discrete trend of a group of measurement data are range, percentile, interquartile spacing, variance, standard deviation and coefficient of variation, among which variance and standard deviation are the most commonly used.
Concentration trend and dispersion index reflect different characteristics of data respectively. As summary statistics of data, these two types of indicators are required to be used together in statistical description, such as mean standard deviation of normal or nearly normal distribution data, median and quartile interval of skewed distribution data.
Baidu Encyclopedia-Measurement Data
Baidu encyclopedia-measurement