SPSS data charts

　　"Achilles heel" of data, is its visualization.

　　We like painting, like articles with pictures, like sophisticated charts statistically. Visualization of historical data dating back to the 10th century, while the 17-1990s, Playfair's book "Business and Political Illustration set" for the first time using a line graph to represent the difference between imports and exports of the country, to continue this practice today! These interesting stories, you can refer to "the Image Statistics" .

Experiment name: Data charts.

Purpose: 1. Learn to use SPSS is easy to operate. 2. 7 kinds of master data in the chart.

Experiment: 1. generating a frequency distribution table and contingency table; FIG. 2 production stem; 3 plots boxplot; FIG. 4 drawing vertical; FIG. 5 plots the error; 6 plotted scattergram......

　　And generating a frequency distribution table 1. The contingency table: frequency of use described process, if you want to know the distribution of a degree project respondents, the process may be used to output a corresponding frequency frequency table. "Analysis" → "Descriptive Statistics" → "Frequency" → select the variable you want to analyze . The variable X opt-in "Variables" list, click "OK" button. Experimental Procedure:

1 FREQUENCIES VARIABLES=math english chinese
2   /ORDER=ANALYSIS.

frequency

　　Use the Crosstabs procedure described, select "Analysis" → "Descriptive Statistics" → "Cross-table" , will open the dialog Crosstabs procedure. After you select the variables you want to analyze, click "OK" button.

1 CROSSTABS
2   /TABLES=gender BY math BY city
3   /FORMAT=AVALUE TABLES
4   /CELLS=COUNT
5   /COUNT ROUND CELL.

Crosstabs

　　2. FIG leaf production: due to the need to group the data summary for the histogram, so for the case of small sample size, the histogram will lose some information, it may be used at this time to FIG stem more precise description. Shape and function is very similar to the histogram of FIG stem, but he is a text-based graphics. Operation: "analysis" → "Descriptive Statistics" → "explore" to the variable math, for example, use the discovery process to draw out the stem and leaf diagram. As can be seen from FIG. 7, FIG stem is actually the result can be approximated as the traditional histogram placed horizontally, the entire pattern composed entirely of text output, the content is divided into 3: 1 as the frequency, row represents the frequency of the observations. As the second stem, the stem denotes the integer part seen from below the actual width divided by FIG. 3 as a leaf representing the actual observed value divided by the fractional portion of the width of the stem. There stem and leaf diagram below indicate the number of cases leaves each number represents.

1 EXAMINE VARIABLES=score
2   /PLOT STEMLEAF
3   /STATISTICS NONE
4   /CINTERVAL 95
5   /MISSING LISTWISE
6   /NOTOTAL.

Stem and Leaf

　　3. Draw boxplot: Boxplot and histograms are used to describe the same distribution of continuous variables. However, a histogram focused on the distribution of continuous variables were examined in detail, and box plots to focus more on percentiles outline the main indicators of statistical information on. FIG box facilitates the use of a plurality of continuous variables simultaneously inspect, or to inspect packets of a variable, and therefore more flexible than the use of the histogram, and more widely used.

　　( 1) "analysis" → "Descriptive Statistics" → "exploration" , in the dialog box "map" , there is a box plot options.

　　( 2) "Graphics" → "Chart Builder" , select the gallery "FIG box" group , select the applicable box plots, the icon into the canvas. Y found in the list of variables, drag it into the canvas of the longitudinal axis of the box. The X abscissa dragged into the box. Click the "OK" button. (Note that the horizontal axis)

　　( 3) "Graphics" → "old box" → "FIG box" , the "box plot" dialog box, click the "definition", the next dialog box to select a good "variable", "category axis" and the like, and then click OK. And the first to use the same kind of approach works.

1 EXAMINE VARIABLES=score BY class
2   /PLOT=BOXPLOT
3   /STATISTICS=NONE
4   /NOTOTAL.

Boxplot

　　4. Draw FIG vertical: vertical FIG, each belonging to the same class scatter connected, with the vertical length may reflect differences in the size of the time variation values of the changes. Thus, compared to some of the vertical chart drawing more emphasis on the value of several variables where the difference with another variable changing circumstances. Operation: "Graphics" → "old box" → "line chart" in the dialog line graph, select "vertical line" → "values for a single case of" → "definition" , in the dialog box, " point that " elected to be analyzed variable " category tag " click" variable "category and elected to represent variables, and finally, click OK. Note: The "dot indicates" where select at least two variables, and move it in the "dot stands for" field. These variables must be numeric. Generating for each field labeled variable tag in each case connected by a vertical line. Select an option in the "category tag" group, according to case number or variable mark category. If you select a variable, you must select the variable from the target list, and move within a "variable" field. This variable may be a number, a string or a long string.

1 GRAPH
2   /LINE(DROP)=VALUE(a b c d) BY type.

Vertical map

　　5. Draw FIG error: error map display data for confidence intervals, standard deviation, standard error of the mean, or to estimate the dispersion. By a single formula for categorical variables and error bars bar duplex FIG.

　　( 1) "Graphics" → "old box" → "error bars" , in the dialog box, select the "simple" and "groups of cases summary" , click Define. Choose a good variable and category axis. Finally, click OK.

1 GRAPH
2   /ERRORBAR(CI 95)=score BY school.

Simple error bars

　　( 2) "Graphics" → "old box" → "error bars" , selected in the dialog box "clusters" and "groups of cases summary" , click Define. Choose a good variable category axis and defined in terms of clustering. Finally, click OK.

1 GRAPH
2   /ERRORBAR(CI 95)=score BY school BY sex.

Error bars tufted FIG.

　　6. Draw scattergram: simple Scatter Examining the correlation between the trends of two or more variables.

　　( 1) "Graphic" → "Chart Builder" , select the gallery "Scatter / point map" , the simple standard scattergram onto the canvas, the variable onto the corresponding axis.

　　( 2) "Graphics" → "old box" → "scatter / dot map" , click "simple scatter plot" In the dialog box, click the definition in "simple scatter plot" dialog box, onto the corresponding variable corresponding coordinate axis, and finally click OK.

1 GRAPH
2   /SCATTERPLOT(BIVAR)=x WITH y
3   /MISSING=LISTWISE.

Scatter

Summary and Experience

　　Secondary chart is not the only commonly used "bar", "line graph", "pie", it also "box plots", "error map", "Cross-Tab" and the like. Flexible use of charts, to enhance data visualization, statistics also a great pleasure, use the chart is also a major feature of it is different from other branches of mathematics. However, we must be based on characteristics of the data, choose the appropriate chart to visualize the nature of these data, otherwise, it will not take Ngau Tau horses mouth. So, before drawing the first painting at these charts in the brain, it can be used with.