SPSS Statistical graphs

The chart can show the main data information concisely and intuitively, and reflect the laws and associations of things. Of course, will inevitably lose the details of the data, fish and bear cake can not be both.

According to the number of chart rendering variables, it is divided into univariate, bivariate, multivariable, and then subdivided according to the test scale. Variables are divided into 3 categories: unordered, ordered, and continuous variables.

1, univariate diagram: Continuous type variable. including histogram, stem and leaf diagram, box diagram, p-map, etc.

2, univariate diagram: categorical variables. Includes pie chart, simple bar chart, Pareto chart.

3. Double variable diagram: Continuous response variable. That is, the strain amount is a continuous variable case. Note that the word "should" here is not a clerical error and the strain amount can be understood as the dependent variable in the mathematical equation.

This can be divided into three cases, when another active variable (independent variable) is:

(1) Unordered categorical variables: a simple bar chart.

(2) Ordered categorical variables: line graph, bar chart. It is used to visualize how variables rise or fall as the order variable changes.

(3) Continuity variable: scatter plot. The number of links between two consecutive variables is presented by the degree of dispersion and the trend of variation.

4. Double variable diagram: Classification should be variable. When the argument is:

(1) Categorical variables: Bar chart. According to the specific presentation mode, but also can be divided into: Duplex bar graph, Sectional bar chart and Massectoux 3 kinds.

(2) Continuous variable: There is currently no good graphics available. The common way to do this is to swap the self/response variable with a bar chart.

The above-mentioned two-variable diagram is only normal and common, in fact, can also take advantage of the characteristics of the single-variable graph, when the independent variable is a categorical variable, can be divided into categories to draw the corresponding single-variable graph to render, the common group box, duplex pie chart, histogram group.

5, multi-variable diagram. Only the 3 variable diagram is described here, do not make the chart too complicated, otherwise you will lose the chart "intuitive" advantages.

To show the correlation of 3 variables, it is best to use three-dimensional coordinates of the three-dimensional statistical chart, but because in fact still on the plane to the three-dimensional diagram, the three-dimensional map is not convenient to use.

(1) When a variable is a categorical variable, the two-dimensional graph can be expanded, so that the two-dimensional graph can show more information. For example, in a scatter plot with a point shape or color to distinguish between different categories, in fact, is rendered two consecutive variables and a categorical variable number of associated information. Similarly, there are multi-line graphs.

(2) When all variables are continuous variables, the above method is not available. The need for a high-dimensional scatter plot, SPSS provides a series of functions, such as scatter chart matrix, three-dimensional scatter plot and dynamic rotation.

6, other special use of the statistical chart.

(1) Meet the specific needs of a particular industry: such as statistical maps used to combine statistical data with geographical distribution, control charts for industrial quality control, and high and low charts for stock analysis.

(2) To solve a specific statistical analysis problem: the error bar graph used to describe the confidence interval or distribution range of the sample indicator, the ROC curve used for the analysis of the diagnostic test effect, and a sequence diagram for the pre-analysis of time series data.

Resources:

1. Zhang Wentong. The basic course of statistical analysis of SPSS. Page 166-216

SPSS Statistical graphs