Research Support

The Research Support team at the Carroll School of Management assists faculty with data management and statistical analysis. Below is a brief overview of our services:

  1. Acquisition and data management.
  2. Application of contemporary statistical and econometric methods.
  3. Software support for Stata, Mata, SAS, R/S-Plus, M-Plus, SPSS, and MATLAB. 
  4. Development and debugging of computer programs.

Please contact us at carroll.research@bc.edu, even if your specific need is not described here. We would like to know what you are working on and how we might help!

Steve Lacey, Research Statistician

Steve Lacey joined the Carroll School of Management in 2008 as a Research Statistician. Steve’s statistical support focuses on structural equation, latent variable, and multilevel models. In addition, he develops programs to extract and restructure data from the Internet (databases, webpages) into forms suitable for statistical analysis. He works mostly in R and Python, but is comfortable with other statistical packages and programming languages.

Steve holds a PhD in Cognitive Psychology from the University of Michigan, Ann Arbor, and has long had a passion for data and statistical methods in the social sciences. 

Contact:
Fulton Hall 535
617.552.0591
steven.lacey@bc.edu

 

Ming Lu, Research Statistician

Ming Lu joined the Carroll School of Management in 2018 as a Research Statistician. Ming provides research support in data management, statistical analysis, and research paper replication. He has more than ten years’ experience on Compustat, CRSP, and other WRDS data. He works mostly in SAS, and can also work with other programming languages such as Matlab, R, and Python.

Ming holds a Master degree in Statistics from Columbia University in the City of New York, and a Master degree in Mathematics from Tsinghua University, Beijing. What’s more, Ming is a SAS Certified Advanced Programmer.

Contact:
Fulton Hall 552A
617.552.8598
ming.lu@bc.edu

Faculty Projects

  • Prepared guide to SSH Secure shell for running SAS and other programs on WRDS server.
  • Mapped longitude and latitude coordinates on to the world map using SAS GMAP procedure.
  • Computed quarterly and annual firm-market crash measures from the CRSP/Compustat databases.
  • Parsed and restructured data tables in portable document format (pdf) files and wrote results to comma-separated files (csv).
  • Extracted and restructured unbalanced nested patent data stored in HTML.
  • Compiled links to Securities and Exchange Commission (SEC) filings, downloaded and parsed filings, estimated language complexity and searched for words and phrases.
  • Reshuffled trading data in HDF5 format on Pleiades Linux server.
  • Automated conversion of csv, xlsx, and dta format files into SAS format and vice versa.
  • Reshaped a large wide format marketing dataset into a more usable long format dataset.
  • Acquired and managed data from the WRDS databases including Compustat, CRSP, I/B/E/S, OptionMetrics, RiskMetrics, and Audit Analytics.
  • Efficiently computed leads, lags and moving averages using SAS EXPAND procedure.
  • Computed biased-corrected and regular bootstrapped confidence intervals for beta coefficients and R2.
  • Performed Monte Carlo simulations and Bootstrapping to obtain sampling distributions of coefficients and conducted hypothesis tests.
  • Applied Stata gsem procedure to estimate Seemingly Unrelated Regression (SUR) model on unbalanced panel data.
  • Wrote SAS program estimating multivariate Fama-Macbeth regression model that identifies and ranks the best predictors of cross-sectional stock returns.
  • Presented options for addressing common method variance; implemented marker variable technique.
  • Developed intuition for why dummy-variable adjustment procedure leaves regression coefficients unchanged and researched procedure’s established limitations.
  • Outlined differences among intraclass coefficients (ICC) as measures of reliability in multilevel data and how best to report them in a revise-and-resubmit.
  • Clarified incomplete factorial design, alternative ways to estimate ANOVA, options for multiple comparisons, and how to read SPSS output.
  • Investigated analysis options for cross-classified multi-level data with small sample. Explored generalized mixed-effect models and generalized estimating equations, and found the later performed better, especially with non-normal outcomes.
  • Estimated multi-group Confirmatory Factor Analysis. Identified and addressed item invariance among groups.
  • Carroll School Research Reports.
  • Carroll School Teaching Reports.
  • Analysis of Carroll School Undergraduate Student Survey
Back To Top