Data Analysis


We have research and teaching strengths in a wide range of traditional and state-of-the-art statistical analysis methods. These include epidemiological measures of occurrence and effect for cohort and case-control (case-cohort or case-base) study designs, as well as notions of causality and confounding.

Methods in data analysis range from tabulation to stratified analysis, multiple comparisons, sensitivity and bias analysis, generalized linear (e.g. logistic) models and survival analysis. More advanced approaches are nonlinear and smooth (spline) regression and hierarchical regression (multilevel modeling). The latter approach can be applied for the construction of diagnostic and prognostic decision-making trees. All these procedures and Trellis graphics are available in the S+ environment or the Open Source R.

From Observed Frequencies To Modeled Probabilities

The idea is, like in the MarkStat logo, that the sigma letter symbolizes the collection of rough frequency data summarized as proportions, while the integral sign signifies their convolution into a smooth statistical model, which can be subjected to a stochastic analysis and inference.



Statistics for You

We specialize in epidemiologic study design & data analysis methodology.

Statistics Provided

Contact Markstat

MarkStat Consultancy
Kajanuksenkatu 4 A 2 | FI-00250 Helsinki
FINLAND | Tel. +358 505769290

Contact Details