Small article modifications undergone two days, I found the original excel can do very basic data analysis.
1: Scatter do linear regression
Very powerful, if it is a scatter plot, the label can not change the axis of abscissa.
Coordinate axes may be adjusted, the mark, the name, scatter pattern, drawing the regression line, regression equation, error bars, grid lines, the legend, the linear parameters and the like.
2: Line Chart
The axis of abscissa of FIG tab fold line may be modified, may be made to the series line "smoothing" process.
For example, do a trend analysis, or to make a frequency distribution, are plotted against a variable. For example, the number of occurrences.
In doing frequency distribution, to do an interval division. FREQUENCY function to be used here, while using ctrl + shift + enter, into the data array.
3: determine the correlation coefficient
There are three ways, one is the PEARSON function, a function is CORREL, a data analysis of the "correlation coefficient" option.
The "Data Analysis" module default excel not open, we need to from the "File" - Open the "Analysis Library Tools" - "Options" - "add-ons."
4: do T test
If the comparison is whether the two samples are different mean, T-test may be applied to the case of the sample less than 30,
TTest can use the function directly.
5: analysis of variance
If you have more sample groups (> = 2), if there are significant differences in the results to compare different treatment, if a single factor is the treatment process is the analysis of a variety of multi-factor analysis of variance.
T-test error is smaller than, but also simple, no-one to do the T-test.
For example, using three different fertilizers, using four kinds of seed, yield different varieties of contrast, look at what factors had a significant effect on yield
6: Chi-square test
In front of a diary recording the chi-square test, but more trouble is, you need to do the math chi-square value, and then use the function CHIDIST, calculated p value. Do not have to go to the poor chi-square table.
Chi-square value is calculated very much trouble, or use the R language to count it, just two sentences.
tablefaw<-matrix(c(2,3468,154,45348),nrow=2,ncol=2)
chisq.test(tablefaw)
7: Other For example, time series analysis, multiple regression analysis, non-parametric statistical temporarily useless.