PBS Services Advanced data analysis

Advanced data analysis

Marketing is a very complex phenomenon. Market research is a tool for the introduction of order in a huge number of variables and, above all, assess how these variables are interrelated.

For this purpose, the raw data is applied to various statistical procedures. Some of them, such as simple correlation analysis of two variables, are easy to intuitive concepts, even by a layman. However, more complex multi-variable procedures, such as canonical analysis, require long preparation and training. However, the ability to carry out the most sophisticated procedures, such as latent class analysis, is held by only a few, professional statisticians.

Working with them opens up completely new horizons to the researcher, and the possibility of very deep penetration into the data structure.

Some of the analysis offered by the PBS:

• Conjoint Analysis (full-profile, and the adaptive choice based)
• Analysis of classification and regression trees
• Analysis of discriminant function
• Analysis of data quality (including the imputation of missing data)
• Correspondence analysis
• latent class analysis
• Time Series Analysis
• Analysis of simple and multiple regression (classical and logistics)
• Analysis of contractual states (PSM)
• Analyses of psychometric test results and test items (in the classical theory test, theory and universalization irt)
• Dual scaling
• MaxDiff
• hierarchical models (analysis of complex samples; curve analysis and development / changes)
• Structural Equation Models (including confirmatory factor analysis, confirmatory factor analysis for discrete, Path Analysis, a full model SEM)
• Loglinear models
• Limited hierarchical cluster analysis
• Randomized response
• Multidimensional scaling
• Estimating the effect of the sample plan (Deff)