Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 19: Correlation and Regression - Coggle Diagram
Chapter 19: Correlation and Regression
Correlation and Regression
: important to understand if two factors in a process are related, is an input is related to an output, a strong correlation or regression does not indicate a causal relationship, learning this helps teams understand process elements and make predictions, correlation and regression do not prove causation
Correlation
: linear association between two variables, can be positive or negative
Positive: one variable increases as the other increases
Negative: one variable decreases as the other increases
can be approximated by viewing a graphical analysis of data from a scatter plot
Correlation Coefficient
: also known as R-value, number between -1 or 1, relation between two variable can be stronger
If R-value is equal to 1 then relationship is certain, positive trend
If R-value is equal to -1, relationship is certain, negative trend
If R-value is 0, no relationship, no trend
Calculation r value in excel is done by:
also can be done by CORREL formula or Data Analysis ToolPak in excel
Linear Regression Analysis
: Lets you create linear model to help predict how one variable behaves if the other changes
Coefficient of Determination
: known as r-squared value, a number between 0 and 1, tells us strength of linear relationships between two variable -x and y, lets us create best fit line
Can be used in excel by Data Analysis ToolPak
Using Correlation and Regression in Six Sigma
both are useful in six sigma in measure, analyze or improve phase of DMAIC
Continuous and ratio data
: correlation and regression works with continuous or ratio data and both sets of variables have to be quantitative
correlation requires two quantifiable data elements, regression requires data used as y=f(x)
used for helping verify relationships between inputs or help understand RCA in analyze phase
