Varimax rotation

As a rotating basis or rotation method is called in the multivariate statistics, a group of procedures that coordinate systems can be rotated until they meet a predefined criterion. The rooms in which those coordinate systems provide no special requirements. They are arbitrary n- dimensional, but ideally metric.

At your disposal are various methods, including:

• Varimax ( orthogonal)
• Oblimin ( oblique )
• Quartimax ( orthogonal)
• Equamax ( orthogonal)
• Promax ( oblique )

Use of rotational molding

Rotation procedures are often used in conjunction with the factor analysis or principal component analysis as an interpretation aid.

Figurative comparison: After were determined by the method of extraction factors that explain the variance of the variable that is used to attempt rotation, the factors the data " counter rotating " until only a few factors with a high charge are left. These can then assign unique hypothetical regularities, which is referred to as an interpretation aid.

The rotation does not increase the explained variance proportion. It merely helps to better understand the content factors.

In addition to the popular varimax rotation, there are other methods. Other orthogonal rotation method are quartimax and equamax (also Equimax ), which is a combination of varimax and quartimax. There also exist oblique ( oblique ) rotation method as oblimin and promax, loosen the acceptance of the orthogonality of the factors.

Varimax

As Varimax is called a mathematical calculation method by which can be coordinate systems rotate in n- dimensional spaces. The end of the 1950s, developed by Henry Felix Kaiser method is mainly used in statistical procedures and plays particularly in the factor analysis, an important role as a substantive interpretation aid.

Method

Varimax rotation is assigned to the process. When used in conjunction with the factor analysis, the factors as long rotated in continuous steps in the room until the squared variance of the charges of each factor is maximum. Thus this process has received its name. Medium charges are thus either less or more and can therefore be clearly assigned to their respective factors. In this case, an orthogonal design is used because the proponents of this method assume that the latent factors are independent.

Geometrically be the ( orthogonal ) coordinate axes relative to the old axes rotated in space, where the origin of the axes is the same. The component transformation matrix is formed from the cosine of the angle between the factors and the original coordinate axes. By multiplying this matrix with the unrotated factor loading matrix, the rotated factor loadings can be calculated:

• : Matrix of the rotated factor loadings
• : Matrix of unrotated factor loadings
• : Component transformation matrix