About Analytical Consulting

Different strategic purposes, different research questions and hypotheses, different designs, different data models ... in the end, various statistical and other analytical approaches.  At MATRIX45, we are prepared to tackle these challenges - not just technically by analyzing your data rigorously, but also conceptually by translating results back to their strategic or clinical context.

What differentiates MATRIX45 in this area?

Rigor, for one: we do not just "run" analyses.  Our selection of analytical models and techniques is driven by the questions and hypotheses at hand, considers the characteristics (limitations and opportunities) of data and their distributions, and balances complexity with interpretability.  If several approaches exist, we explore these as well and interpret the relative advantages and disadvantages of each.

We look beyond the Xs and Ys, remembering throughout the analytical process that these mathematical symbols may represent patient characteristics, diseases, treatments, responses to treatment, etc.  We do not merely try "to fit data to statistical models" and judge success on the basis of "goodness-of-fit".  We focus instead on building strong and defensible statistical foundations for real-world concepts and contexts.

Our expertise goes beyond conventional biostatistics.  We are just as skilled in analytical models from other disciplines: decision sciences, health services research, behavioral and social science, measurement, business and economics, ...

We can tackle complex issues: multi-level separation and differentiation of (latent) effects and outcomes, risk adjustment, causal modeling and causal chain analysis with noncontrolled data, embedded pattern detection, ...

Through our academic affiliations in the USA and Europe, we have built alliances with some of the world's foremost experts in quantitative science and statistics.

Lastly, MATRIX45 provides analytical consulting services.  That, we believe, is different from being a "stats outfit".

Top of Page   Top of Section

Case Study

Throughout their careers, MATRIX45 principals provided a lot of analytical consulting.  In addition, they have contributed to the statistics and quantitative science literature (see Publications) .  We present one case study below, however much of the principals' prior work illustrates the analytical leadership that MATRIX45 can provide.

Socioeconomic Status and Health Outcomes

Top of Page

Low socioeconomic status (SES) predisposes individuals to illness, disease and death. SES interacts with other factors that determine health outcomes, including the biological factors of age, race and gender; environmental factors of physical and social resources; behavioral factors including positive and negative health behaviors; and access to and utilization of health care. We quantified the relationships between various health determinants through latent variable structural equation modeling using data from the 1993 National Health Interview Survey. The variables that most significantly impacted health status included: number of doctor visits in the previous year, age, proportion with Medicaid, and SES.  Further analyses showed that the poor use more health care and have worse health than the non-poor; and that the insured poor use more health care and have worse health than the uninsured poor. SES is an important determinant of health and healthcare utilization in both the poor and the non-poor.  

Return to list of examples

Top of Page