Partial Least Squares (PLS)

How it works:

PLS (Partial Least Squares) is a highly developed statistical method for the assessment of so called causal models. Within the framework of a PLS analysis, individual, independent criteria are categorised into factors, i.e. into dimensions of quality of service. Afterwards, their influence on one or several dependent factors is determined. These dependent factors in turn consist of individual criteria. For example, the customer satisfaction index comprises three individual questions.

The algorithm used by Anovum has been optimised for the analysis of survey data. It assesses cause-and-effect connections through a combination of factor analyses and multiple regression analyses, with the main aim of maximising the model’s predictability with regard to the dependent factors.

Advantages:

LV PLS eliminates the weaknesses other statistical methods have when it comes to the analysis of survey data.

 Partial Least Squares

Uses:

Analyses of quality of service, customer satisfaction , reputation, customer loyalty , employee satisfaction etc.

Further information:

The PLS used by Anovum is based on a method published by Herman Wold that was adapted by J.-B. Lohmöller and applied in a software programme (LV PLS).

In customer and employee satisfaction research PLS became known through the work of Prof. Claes Fornell, amongst others, and through its application in the American Customer Satisfaction Index (ACSI).

To find out more about Partial Least Squares (PLS) please contact us.