We have helped hundreds of corporate clients find solutions to complex business problems through the use of advanced statistical modeling techniques. More often than not, firms contact us to provide them with the solution to a particular obstacle or problem, as is the case at standard management consulting firms. Many firms have enormous amounts of extremely valuable data at their disposal, yet they lack the in house resources to optimally analyze their data.
What sets us apart is a core specialty in solving business problems that require statistical methodologies and data analysis techniques which are often beyond the scope of the client’s internal staff. While this is sometimes due to the problem requiring the manipulation of huge and hard to manage data sets, most of our clients come to us simply because the methods needed to solve the problem are complex and unknown to them. Engaging us for statistics consulting implies that you want a boutique firm that will provide you with demonstrable results.
Statistical Methodology and Data Analysis at the Forefront
Our core specialty is developing sound statistical methodologies to analyze business problems, and then carrying out the requisite data analyses to solve those problems. Ordinarily, firms come to us requiring the answer to a specific issue. Our first step is always to fully understand the problem and the data available to us in solving that problem. Given the type and amount of data available, we then determine the optimal statistical methodology to solve the problem. Alternatively, in the case of the firm having no relevant data, our first step would be to create the optimal methodology to collect the needed data.
In many business problems, the data available can be incomplete, disorganized, or unwieldy. In these cases, we often need to find a clever way to extract useful info out of the available data, and/or to supplement that data with additional data. At other times, we are presented with a vaguely-posed business problem, and our job is to specify the problem statistically, and to find a robust and practical solution that doesn’t succumb to data-snooping pitfalls.
In all cases, our consultancy differs from typical management consultancies in that our methods and solutions are always empirical and objective.
Many statistics consulting projects that we take on are brought to us after a traditional management consulting firm was unable to specify how to optimally solve the problem(s), or just generally unable to handle the quantitative rigor of the solution(s). We have a broad range of experience providing solutions through the application of a vast array of statistical and numerical methods, including survival and event history analysis, ARCH and GARCH models, neural nets, bootstrapping, Thurstone/Shapley analyses, structural equation modeling and multi-dimensional scaling, and Monte Carlo simulations.
Our comfort with virtually all higher-level quantitative methodologies and our expertise in applying these tools in a corporate setting enables us to assist with virtually any business problem. Here is a sample of some of the analytical tools with which we are familiar:
- Various forms of Regression Analysis, including Non-Linear Regression and Ridge Regression
- Survival Analysis including Kaplan-Meier Analysis
- Time Series Analysis, including Vector Autoregression (VAR), Vector Error-correction Models (VECM), and GARCH and its variants (NGARCH, EGARCH, etc.)
- Panel Analysis
- Survey Development and Reliability/Validity testing of existing surveys
- Statistical Power Analysis for Sample Size determination
- Multivariate Analysis (with multiple outcome variables), such as MANOVA and MANCOVA.
- Qualitative analytical methods including Phenomenological Analysis and Grounded Theory
- Optimization methods such as Linear and Nonlinear Programming, Genetic Search, and Simulated Annealing
- Markov Chain Monte Carlo and similar methods
- Nonparametric Methods
- Zero-Inflated Count Models
- TURF, Thurstone Scaling, and Shapley Values (used often in marketing-related projects)
- Conjoint Analysis, Choice Modeling, and Maximum Difference Scaling (MaxDiff)
- Structural Equation Modeling (SEM), Confirmatory and Exploratory Factor Analysis (CFA and EFA), Multidimensional Scaling, and Path Analysis
- Neural Nets
- Machine Learning and Artificial Intelligence (AI)
- Parallel Computing to handle very large datasets and for problems that would otherwise take excessive computing time
- Various Bootstrapping and Jackknife techniques
- Spatial Analysis
- All statistical software packages including R, Matlab, SAS, Stata, MPLUS, LISREL, EQS, PASS, Maple, Mathematica, and SPLUS
- Data Mining with Weka and RapidMiner