Making sense of seemingly unrelated or complex data sets is what we do. FMG’s statisticians, database managers, software developers, and data scientists are responsible for large-scale data management and statistical analyses to convert big data into usable information.
FMG has recognized expertise in popular and advanced statistical and data analysis approaches. We develop complex sampling plans, automate the production of large quantities of descriptive statistics, and conduct complex statistical analyses with multilevel regression with poststratification. FMG conducts simulations and develops semi-automated, content-coding platforms using natural language processing (NLP) techniques and supports vector machine classification, providing efficiencies in content-coding tracking studies.
To work with a wide variety of markets and industries, we employ a suite of analytic software, such as SPSS, SAS, NVivo, Stata, MPlus, MS SQL Server, Python, ArcGIS, and R, and develop efficient parallelized and cluster computing-based analysis using big data tools such as Apache Hadoop and Spark.
With this technology and expert staff, there is no data challenge FMG does not solve.
Partner With FMG
Vice President, Statistics and Research Methods
Examples of Our Methods
- Regressions (e.g., Linear, Logit, Multinomial)
- Conjoint Analysis and Experimental Design
- Panel and Multilevel Modeling
- Hazard Modeling/Survival Analysis
- Time Series Analysis
- Bayesian Modeling (Naïve Bayes, Complex Random Effects)
- Simulation Methods (including Monte Carlo and Agent-based Modeling)
- Power Analysis and Effect Size Computation
- Machine Learning/Data Mining (e.g., Support Vector Machines, Neural Network Analysis, Market Basket Analysis)
- Decision Trees and Ensembles (CHAID, CaRT, Random Forest, Boosting)
- Structural Equation Modeling
- ANOVA (including Repeated Measures, MANOVA, and ANCOVA)
- Clustering/Latent Class Analysis
- Psychometric Analysis and Data Reduction
- Merging and Harmonization of Multisource Data
- Feature Engineering, Data Cleaning, Recoding
- Missing Data Methods
- Predictive Model Development and Validation