Data Handling

SKIM's general data handling services

Data quality check
some respondents apply "efficiency" strategies such as random answering or rushing through the questionnaire by selecting same response options every time. In most cases the time-span of the interview gives a good indication of the effort your respondents have put into their answers. Given your indications we can help identify these respondents and exclude them from analysis.

Cleaning data files
based on identification criteria we can remove specific respondent files from source data.

Recoding / merging data files
Some surveys are conducted using different methods of data gathering (Web, CAPI, P&P) and different versions (usually per country). We can help you build one single data file.

Alternative HB estimation runs
Individual utilities can be derived based on all kinds of parameters to tune the estimation using Hierarchical Bayes. We can produce several sets of utilities using different parameters. Think of using monotonicity constraints, absolute and conditional pricing.

Latent Class runs
Latent Class is an excellent technique to discover base preference patterns in your sample. It is a valuable and manageable way to illustrate heterogeneity in your sample.

Advanced Data Handling Services

Recoding/merging .CHO files
In order to fine tune the experiment or gain multidimensional insights from the conjoint task, we can help recode or merge conjoint experiment files in order to squeeze out all the information it contains.

Running scenario simulations
Given your specification we can run multiple scenario simulations and structure outputs in order to speed up your analysis.

Sensitivity / pricelines / price sensitivity
Based on a scenario we can test and depict (price) sensitivity in various ways.

Optimization runs
If you want to identify an optimal product or product range we can help you optimize your scenario.

For more information on how we work on a project, please take a look at our project guidelines.