|
|
 |
 |
|
|
|
| Quantitative methods are normally applied when empirical observations of a few selected characteristics are required. This method provides a platform for the testing and statistical validation of a hypothesis. Some examples of when this type of method might be used include:
|
|
 |
the statistical evaluation of qualitative results or planned campaigns |
|
examinations where the statistical accuracy is more important than in-depth assessment (such as for customer satisfaction analyses, analyses of the competition etc.) |
|
| |
| The analysis of quantitative data is carried out by means of univariate, bivariate or multivariate procedures. It is geared to the client’s individual information needs. Amongst others, we offer the following multivariate techniques: |
|
|
regression analysis |
|
factor analysis |
|
multidimensional scaling
|
|
conjoint analysis |
|
cluster analysis |
|
Analysis of structural equation modeling |
|
Correspondence analysis |
|
|
|
 |
|
|
Statistical
procedures
Regression analysis
It aims at the description and explanation of coherences between
one dependent and one or several independent variables. Alternatively,
a regression analysis can be used for the calculation of prognoses.
Example:
Description of the influence of advertising expenses, number
of shops and national income on a product’s quantity of
sales
Prognosis of the development of sales in case of changes in
prices, advertising expenses etc. |
|
|
|
|
| |
Factor analysis
It is deployed when a question is surveyed through many different
variables, which are to be bundled to so-called factors. In
this context, the following question is in the spotlight: Is
it possible to reduce the characteristics to few factors?
Example:
Compression of the numerous technical features of cars to few
dimensions like performance and security |
|
|
|
|
| |
Multidimensional scaling (MDS)
Mainly, multidimensional scales are used for positioning analyses.
Thereby, global similarities between objects are inquired. With
the help of this technique, the dimensions of perception underlying
these similarities are derived. Multidimensional scaling is
particularly utilized whenever there is none or barely any knowledge
about characteristics relevant to the subjective judgement of
objects.
Example:
Inquiry about the global similarities of brands
|
|
|
|
|
| |
Conjoint analysis
It is used for the measurement of preferences. Its aim is to
determine the contribution of individual product features to
the total utility of the product. The conjoint analysis is an
important field of application for the configuration of new
products.
Example:
Analysis of the influence or contribution of alternative product
features (materials, forms, colors, prices) to the judgements
made with respect to the product’s advantages
|
|
|
|
|
| |
Cluster
analysis
In this case, similar objects are clustered in individual groups
rendering the differences between the groups as big as possible.
Example:
Differentiation of competing groups e.g. based on range of products
and target groups |
|
|
|
|
| |
Analysis of structural equation modeling
Within the area of market and opinion research, different questions are formulated regarding the complex relationship structures of latent characteristics. The term “structural equation modeling” does not indicate one individual technique, but rather a whole group of models for multivariate, statistical data analysis, which examine these characteristics. They are suitable for displaying explanatory, confirmative or causal dependencies between the characteristics in question.
Example:
Modeling of purchase decision processes for end customers
|
|
|
|
|
| |
Correspondence
Analysis
A correspondence analysis is an explorative, multivariate method which aims to display the complex facts about different characteristics in clear scatter diagram. The characteristics are placed within a two dimensional range which has been previously defined so that the positioning and distances between the different points can be interpreted for their content. In this way, the relationships between the characteristics, which could not be seen within the mass of data, can be revealed.
Example:
Image and positioning analyses, market segmentation the form of target groups and product identification using selected characteristics (product/target groups).
|
|
|
|
|
|
|
|
|
 |
|
Practical example for a quantitative analysis
Read more about:
- EEG Study 
- PV in Germany
|
|
|
|
|
| |
 Thomas
Olbrecht
t.olbrecht@eupd-research.com
+49 228 97143 59 |
|
|
|
|
