parent nodes: Basic Statistics | Choosing a statistical test | Compare with | Distribution fitting | Graphs | index | IndexBasicStat | Indexch | IndexNews | indexnl | IndexTools | Measuring and data collection | Regression | Result table | Sample size Calculations | Sitemap | Six sigma | Statistical software | Support | Whatsnew | What to do with not normally distributed Data


Linear Correlation

Main Help

Basic Statistics
Mean
n
Median
STDEV
Cp Cpk % out of tolerance
Min/Max
Compare with
Calculate Difference
Diff mean normally
Diff median not normally
Diff variation normally
Diff variation not normally
Normally test
Correlation
Regression
Kurtosis
Skewness


Full version of Develve
For commersial use
75 EURO

If Linear Correlation test is selected, the correlation test is calculated. For this calculation there must be at least 3 or more observations. This calculation calculates if there is a significant correlation between the two compared columns. When there is a relationship (correlation) between the two variables this relationship can be calculated. See Regression.
The correlation does not work in DOE mode because there are no pairs in this mode.

Interpretation

Colors of the cells

Linear Correlation No correlation

Formula for calculating the correlation r value.

formula

Formula to convert the r to the student t value

formula

With the t value and the degrees of freedom minus 2 can the program calculate the p value.

Sample size

formula
The result is the minimum sample size of pairs.

Legend

n = Amount of pairs
formula= correlation
formula= significance 0.05 in the sample size calculation (2 sided)
formula= Power 0.80 in the sample size calculation

Example

Between data-sets A and D is a correlation (Row Correl p <0.05), and the sample size is big enough.
Between data-sets A and C is no significant correlation and the sample size is to small.

Data file

Multiple Linear Correlation

Main Help

Tools
Proportions
Chi Square test
One way Anova
Kruskal-Wallis Test
Variation Levene test
Multiple Correlation
Multiple linear Regression
Generate Distribution
Calculations
Filter
Box-Cox transformation
Gauge R&R
Distribution fitting
Graphs


Full version of Develve
For commersial use
75 EURO

Basically the same as the Linear correlation but then for more than 2 data sets.

Interpretation

Colors of the cells

Example

To use the Multiple Correlation test "Menu: Tools=> Multiple Correlation". Then select the data-sets to test. Only between data-sets A and D is a correlation (Row Correl p <0.05). When clicking on Graph the a matrix of correlation graphs will be displayed. In graph it is visible that there is a correlation between A and D.

Data file

External links