Personal tools
You are here: Home PRI Programmers Data Management and Statistical Programming

Data Management and Statistical Programming

An overview of the core and the services it provides

Who we are

  • 6.75 FTE staff,  1 faculty advisor, 1 PRI funded graduate assistant (vacant), 1 graduate assistant shared with College of Education
  • Experience in the unit: 1994-2002
  • Experience in the field: 10-30yrs Coming to us from: Federal government agencies (NCI, DOE), private industry (Bethlehem Steel, MITRE, Spectrumedix), other PSU depts (ITS, student aid, colleges).
  • Academic background: Computer Science, Mathematics, Library & Information science, Plant Pathology, Economics, Biology, Civil Engineering.

Statistical programming

  • Statistical programming: missing data imputation, variable reduction, causal analysis, mixed methods, nonparametric models, sampling, power analysis.
  • Representative projects: Scanlon, Farkas, Johnson
  • KDEplothisto

Web programming

  • Online statistical analysis
  • Online data collection & dissemination
  • Team webspaces for collaborative writing, project management
  • Representative projects: Grouphomes

Data management & database design

  • Large-file manipulation: images, genetic, census
  • Relational database design
  • Lab data, survey data, medical/insurance claims 
  • Metadata creation and documentation
  • Representative projects: Baker (MathIQ), Vogler (GOB)

Data sharing strategies

HeadStart
  • Archiving, dissemination, data sharing plans for NSF/NIH grant proposals
  • Anonymizing strategies for human subjects datasets
  • Representative projects: SodaPop, HeadStartREDI
Sodapop
 

Pre/Post-Award services

  • Preliminary analysis for grant proposals, NIH minority inclusion tables
  • Feasibility studies
  • Followup analysis and graph production for publications
  • Representative: Greenberg hand-held device software evalution, Mpofo dataset exploration

Mentoring/training grad students

Pic 1
  • Hire to supervise, mentor research assistants
  • Training on-demand
  • Online and walk-in consultation
 

Hiring Scenarios -

Scheduling arrangements are very flexible and are dictated by the specific needs of your project. However, there are several common scenarios:

Data Manager

(12-32hr/wk continuing)
Coordination of all data related tasks including: gathering files, data conversion and standardization of file structures, inventory, data cleaning. Maintaining appropriate security. Design of data entry and web-based data collection systems for primary data collection. Generating technical documentation & codebooks for in-house and public-release data. Web-site design. Extracting, merging & re-shaping of data prior to analysis. Conducting univariate & multivariate statistical analyses & generation of statistical graphics. 

Programmer/Analyst

(task-by-task basis 8 to 32 hrs/wk)
Prepare data files, construct high-level computer programs to combine & modify data sets. Build data entry systems. Writing high-level computer programs and macros for statistical analysis. Graphical user interfaces for applications. Recoding & constructing new variables. Imputation of missing values. Producing statistical graphics. Web-page construction.

Consultant

(4-12 hrs/wk for 2 weeks or more)
Advising on data management strategies and software requirements for grant proposals. Providing dedicated assistance to faculty who write their own code. Training & monitoring computer work of graduate assistants assigned to the project.

Statistical Consultant

DPS faculty director David Johnson is available to provide advice on the appropriate use of statistical software packages. If you have methodological questions, we will match you up with an expert.

Funding sources...

  • Infrastructure supports ~20% of most programmers.
  • Funded research by NIH, NSF, NIE, USDA, others
  • Internal seed money from PRI, CYFC, others

Collaboration with other units...

  • Data/project management -- Survey Research Center  (Prosper, PARTS)
  • Administrative systems -- SSRI/PRI Administration Core (RATS)
  • Data sharing -- PRI Information Core, University Libraries (SodaPop)
  • Training -- Dual-degree in Demography Program, PSU ITS
  • Web programming development --  Weblion project
  • Statistical programming referrals -- Methodology Consulting Center, QuaSSI

Why this works

  • For the staff: varied and challenging projects, job security, collegial atmosphere
  • For researchers: level of expertise and experience of individuals, and benefits of group 
"When you hire an SSRI/PRI programmer, you will in a sense be acquiring the skills and experience of the entire team"

 

Document Actions

Copyright ©2014, The Pennsylvania State University | Privacy and Legal Statements
Contact the Help Site Administrator | Last modified Jan 28, 2013 | Weblion Partner