How to look at the results of multiple logistic regression analysis

Logit regression analysis is used to study the effect of x on y, and there is no requirement for the data type of x, x can be categorical data (can be set as a dummy variable), or it can be quantitative data, but y must be categorized data, and according to the number of options for y, the corresponding data analysis method is used. Logit regression analysis can generally be divided into three categories, namely binary logit regression, multi-categorical logit regression, and ordinal logit regression

1) Binary logit regression analysis, the dependent variable was a dichotomous variable.

2) Multi-categorical logit regression. The dependent variable is categorical data, multiple groups, and unordered.

3) Ordinal logit regression, where the dependent variable is multiple groups of categorical data and orderly.

Binary logit regression analysis is used to study the relationship between the influence of x on y, where x is quantitative data or categorical data, y is dichotomous categorical data, (the number of y must only be 0 and 1), such as willing and unwilling, yes and no, etc.

1) If x is a categorical data, such as gender or education, etc. Then you need to do virtual dummy variable processing on them first, using the spssau "data processing" - "generate variables" function. The operation is as follows:

The dependent variable y can only include the numbers 0 and 1, if the original data of the dependent variable is not like this, then the data encoding is required, set to 0 and 1, using the spssau "data processing" - "data encoding" function, the operation is as follows:

2) Multi-categorical logit regression.

As long as it is logit regression, it is to study the influence of x on y, the difference is that on the dependent variable y, if y has multiple options, and there is no comparative significance between each option, for example, 1 represents "Heilongjiang Province", 2 represents "Yunnan Province", 3 represents "Sichuan Province", 4 represents "Shaanxi Province", the value only represents different categories, and the numerical size does not have comparative significance, then multicategorical logit regression analysis should be used. If there are many categories of the dependent variable y, such as 10, it is recommended to combine the categories to reduce the number of categories as much as possible to facilitate subsequent analysis. This step can be done using the data encoding function of the SPSSAU data processing module.

In the "Advanced Methods" module, select the "Multi-Classification Logit" method, put the Y variables in the upper analysis box, and the X quantitative variables in the lower analysis box, and click "Start Analysis".

Ordered logit returns:

As long as it is logit regression, it is to study the influence of x on y, the difference is that on the dependent variable y, if y has multiple options, and each option has a comparative meaning, for example: 1 represents dissatisfaction, 2 represents average, and 3 represents satisfaction, then an ordered logit regression analysis can be used.

In the "Advanced Methods" module, select the "Ordered Logit" method, put the Y variables in the upper analysis box, and the X quantitative variables in the lower analysis box, and click "Start Analysis".

<>The steps are as follows:1. Import the data you need to analyze to SPSS, click the file in the upper left corner to open it, and select the data in the pop-up dialog box.

2. Click Analyze on the toolbar, select Regression in turn, and then select "Multinomial Logistic" Multiple linear regression analysis and logistic regression analysis are available.

3. Move the variables to the Dependent, Factor, and Covariate boxes on the right.

4. You can see the measurement data in the metric standard.

5. Then set the model, statistics, conditions, options and save of multiple logistic regression.

6. Click OK to use SPSS to do a good job of multivariate logistic regression analysis.

Multivariate logistic regression refers to the study of many factors, such as binomial logistic regression and multinomial logistic regression.

How would you rate this? Collapse.

Binarylogistic regressionThe result interpretation is used for the dependent variable to be the categorical rolling variable. When investigating the effect of x on y, multiple linear regression is used if y is quantitative.

Analyze the linear regression in the SPSSAU general method, if Y is the qualitative data, then use logistic regression analysis.

Characteristics of binary logistic regression results

Logistic regression analysis can be divided into three categories, namely binary logistic regression analysis, multiple ordinal logistic regression analysis, and multiple disordered logistic regression analysis, which is used to study the effect of X on Y and the data type of X.

There is no requirement and x can be either qualitative or quantitative.

However, y must be a fixed type of data, and the corresponding data analysis is used according to the number of options in y.

Methods, in the study of the relevant factors on whether the sample is willing to buy wealth management products in the future.

Gender, major, etc. are all influencing factors, while the gender and major are qualitative data, so it is necessary to set virtual dumb variables, and data processing can be used.

The generation of variables is complete.

Logistic regression analysis is a generalized linear regression analysis model, which is commonly used in data mining, automatic diagnosis of diseases, economic analysis and other fields. For example, risk factors that cause disease, and the probability of disease occurrence according to risk factors, etc.

Multiple linear regression directly uses w'x+b as the dependent variable, i.e., y = w'x+b, while logistic regression uses the function l to correspond w'x+b to a hidden state p, p =l(w'x+b), and then determines the value of the dependent variable according to the magnitude of p and 1-p. If l is a logistic function, it is a logistic regression, and if l is a polynomial function, it is a polynomial regression.

Applicable conditions for logistic regression models

1. The dependent variable is a categorical variable of dichotomous or the incidence of an event, and it is a numerical variable. However, it should be noted that the double count phenomenon indicator is not suitable for logistic regression.

2. Both the residuals and the dependent variable should obey a binomial distribution. The binomial distribution corresponds to a categorical variable, so it is not a normal distribution, and instead of using the least squares method, the maximum likelihood method is used to solve the equation estimation and testing problem.

3. The relationship between independent variables and logistic probability is linear.

Refer to the above content: Encyclopedia-logistic regression.

Hello, I am a partner lawyer of the platform and have received your question.

Through market research and data access, find out the relevant influencing factors related to the ** goal, i.e., the independent variables, and select the main influencing factors.

2. Establish a model: calculate according to the historical statistics of the independent variable and the dependent variable, and establish a regression analysis equation on this basis, that is, the regression analysis model.

3. Correlation analysis: Regression analysis is a mathematical statistical analysis of causal factors (independent variables) and ** factors (dependent variables). The established regression equation only makes sense if there is some kind of relationship between the independent variable and the dependent variable.

Therefore, whether the factor as an independent variable is related to the ** object as a dependent variable, the degree of relevance, and the degree of judging the degree of correlation are issues that must be solved in regression analysis. Correlation analysis usually requires correlations, and correlation coefficients are used to determine the degree of correlation between the independent and dependent variables.

4. Calculation error: Whether the regression model can be used in practice depends on the test of the regression model and the calculation of the error. The regression equation can only be used as a model by the regression equation, and only if it passes various tests and the error is small.

5. Determine the value: use the regression model to calculate the value, and conduct a comprehensive analysis of the value to determine the final value.

Application of regression analysis: 1. Correlation analysis studies whether phenomena are related, the direction and closeness of the relationship, and generally does not distinguish between independent variables and dependent variables. Regression analysis, on the other hand, analyzes the specific forms of correlation between phenomena, determines their causal relationships, and uses mathematical models to express their specific relationships.

For example, from correlation analysis, we can know that the "quality" and "user satisfaction" variables are closely related, but which variable is affected by which variable between the two variables, and to what extent, needs to be determined by regression analysis.

2. Generally speaking, regression analysis is to determine the causal relationship between variables by specifying the dependent variable and the independent variable, establish a regression model, and solve the parameters of the model according to the measured data, and then evaluate whether the regression model can fit the measured data well; If it fits well, it can be further based on the independent variables.

I hope my reply will be helpful and useful to you.

Univariate statistics: Univariate analysis refers to the analysis of a variable at a point in time.

Many phenomena in reality can be divided into two possibilities, or boiled down to two states, which are represented by 0 and 1, respectively. If we take multiple factors to causally relate a phenomenon represented by 0 1.

explanation, it may be applied to logistic regression.

Logistic regression can be divided into two types: binary logistic regression and multivariate logistic regression. First, binary logistic regression is described with examples, and then multivalued logistic regression is further illustrated.

Application of a univariate completely randomized experimental design

1. Single-factor pot test; The design should be applied to experiments in greenhouses and laboratories.

2. If the data obtained in the experiment are equal in the number of replicates of each treatment, the analysis of variance method of one-factor data with equal number of replicates is used.

3. If the data obtained in the experiment are not equal in the number of replicates of each treatment, the one-factor data analysis of variance method with unequal number of replicates is used for analysis.

The above content refers to: Encyclopedia - Univariate Analysis.

Single factorlogistic regressionRefers to the analysis of a variable at a point in time. Many phenomena in reality can be divided into two possibilities, or boiled down to two states, which are represented by 0 and 1 respectively, if we use multiple factors to causally relate a phenomenon represented by 0 to 1.

explanation, it may be applied to logistic regression.

Logistic regression is divided into two categories: binary logistic regression and multivariate logistic regression. <>

Application of a univariate completely randomized experimental designIf the data obtained in the experiment are equal in the number of replicates of each treatment, the analysis of variance of the data obtained in the experiment is used to analyze the data of the same number of replicates, and if the data obtained in the experiment are not equal in the number of replicates of each treatment, the analysis of variance of the data of the single factor data with the unequal number of replicates is used.