Correlation Analysis Method (Pearson, Spearman)

Sometimes we need to study the correlation of certain properties and specified attributes in the dataset, obviously we can use the general statistical method to solve the problem, the following is a brief introduction of two correlation analysis methods, not detailed methods of the process and principle, but simply to do an introduction, because the understanding may not be very deep, I hope you understand.

1. Pearson correlation coefficient

The most commonly used correlation coefficient, also known as the product difference correlation coefficient, takes the value 1 to 1, the greater the absolute value, indicates the correlation stronger. The coefficient is calculated and tested as a parametric method and the applicable conditions are as follows: (suitable for continuous variable correlation analysis)

(1) Two variables are linearly related, if the curve correlation may be inaccurate.

(2) Extreme values can have a greater impact on results

(3) The two variables conform to the two-variable joint normal distribution.

2. Spearman rank correlation coefficient

The distribution of the original variable does not require, the scope of application is wider than Pearson correlation coefficient, even if the grade data, can also be applied. But it belongs to Nonparametric method, the test efficiency is lower than Pearson coefficient. (suitable for classes containing

Variable or all of the correlation analysis of the rank variable)

3. Correlation of disordered categorical variables

The most commonly used is chi-square test, which is used to evaluate the correlation of two unordered categorical variables. The indicators derived from the chi-square values include the number of contacts, Phi, Cramer V, Lambda coefficients, uncertainties, and so on.

Or, RR is also an indicator of the degree of correlation between two variables.

The chi-square test is used to examine whether the two groups of data are statistically different, thus analyzing the correlations between the factors. Chi-Square inspection has Pearson Chi Square inspection, calibration test, etc., different conditions under the use of different chi-square inspection party

method, such as satisfying double greater than (40,5) conditions to use the Pearson Chi-square test method, in addition to the use of calibration Chi Square test method.

　　not much, just want to know the difference between them when they use the relevant methods, and what are the conditions for different methods .

Correlation Analysis Method (Pearson, Spearman)