|
Correlation
Correlation has a significant place in six sigma. Correlation measures relation between two or more variables. It investigates the relationship between two quantitative and continuous variables. If the value of one variable increases when the value of the other also increases, they are said to be positively correlated. If the value of one variable decreases when the value of other variable is increasing, it is said to be negatively correlated. If one variable does not affect the other, they are not correlated.
Correlation is the link or relationship existing between two or more variables. In a positive correlation between two variables, an increase or decrease in one is matched by a similar change in the other. Conversely, a negative correlation make out one variable increase while the other declines. Several statistical methods are used to determine the strength of the correlation, that is, the correlation coefficient. In other words, statistical correlation means quantifiable relationship between two variables. It means that two things seem to be varying in a similar manner. For example, if variable A increases from two to five, variable B also increases from 10 to 20. Correlation analysis estimates the effect or cause relationships. The correlation consistently watches out how the value of a variable alters in cases when one make the changes in the value of the other. It can be concluded depending on the relationship exposed. For instance- the effect of advertising or promotions on the sales of a product or service sets as a good example. A scale of correlation is statistically calculated by the Coefficient of Determination (CoD).
Correlation test will discover the key inputs variable change in a process or products. This is done through mathematical process and by a skilled team. Correlation tests are used in Define, Analyze and Improve steps of the DMAIC process.
Scatter plot is the graph which is used while studying correlation. For creating a scatter plot, it is essential to mark the possible values for one data set along one axis and the possible values for the other set along the other axis. Then, each pair of data is to be plotted as a coordinate on the graph. If the plotted points come closer to form a straight line, the higher is the correlation between the two variables. If the line reaches from low X and Y values to high X and Y values, the correlation is positive. Alternatively, if the line goes from high Y and low X values to high X and low Y values, the correlation is said to be negative. There can be no correlation if the plotted points do not form a line. The number +1 or -1 is in perfect correlation and if the number is closer to either extreme, the relationship is stronger.
Correlation cannot be misunderstood as causation. It means that correlation cannot be validly used to infer a causal relationship between the variables. This aphorism should not be taken to mean that correlations cannot indicate causal relations. However, the causes for the correlation, if any, may be indirect and unknown. Consequently, to establish a correlation between two variables is not a sufficient situation to establish a causal relationship. When scatter plot shows correlation between two variables, it only denotes probable directions for further investigation. The three necessary points to be remembered using correlation studies are: - Correlation coefficient range from +1 to -1: The positive relationship can be made stronger by a closer proximity to +1 and the negative relationship is made by a closer proximity to -1.
- The correlation coefficient is a measure of linear relationship: It may not always be strictly linear in the relationship between two variables.
- Correlation does not mean causation: Correlation describes credible way for undertaking further investigation.
- It signifies that correlation cannot be used in a valid manner to conclude a relationship between the available variables. The aphorism should not be considered to indicate that correlations cannot signify casual relations. And all the causes taken place in relation to the correlation may be unknown and also not direct. So, a casual relation is not sufficient to ascertain a casual relationship between two variables. When scatter plot shows correlation between two variables, it only denotes probable directions for
The most widely-used correlation coefficient is Pearson r, also called linear or product - moment correlation.
Simple Linear Correlation (Pearson r): Pearson correlation (hereafter called correlation), assumes that the two variables are calculated on interval scales. It determines the degree to which values of the two variables are “proportional” to each other. The value of correlation does not solely depend on the specific measurement units used; for example, the correlation between height and weight will be indistinguishable despite the consequences whether inches and pounds, or centimetres and kilograms are used as unit of measurement.
|
