Guidelines for collecting and coding data
 
  • Please consult us before starting data collection.
  • Keep in mind the recent HIPAA regulations on preserving patient privacy. Many times, collecting patient names, addresses, and personal information is not needed for the project. In this case use both patient numbers and initials to identify patients (redundancy is needed to guard against mistakes); keep your own list with sensitive data. If including patient personal data in the database is beneficial, then we will take special precautions to ensure patient privacy (separate access rights).
  • Excel spreasheet may be acceptable for a small one-time project, where there is no need for periodic updates and the data is collected by one investigator. However, entering and coding data in Excel is particularly prone to errors of typing, consistency, etc. Cleaning such errors puts additional burden on the statistician performing the project and delays the analysis. To minimize errors make sure to not mix numeric and text data (a common error is putting comments or question marks in a column of numbers; please designate an additional column for comments). There should be a description of the variables, and the first raw should have variable names or symbols. If you code qualitative categories, please list the categories in each column in the rows beneath data, for example: "M=Male, F=Female" in a "gender" column.
  • If you are familiar with Microsoft Access software, then a a preferable way of storing data for a small-project is using Access tables.
  • We recommend putting your database on our server, "biostat". This is particularly beneficial for a longer-term project or one which involves updating the records. We will create the database for you, when you prepare an Excel sheet with the data to be read-in, or a detailed description of variables. The database residing on the server will be regularly backed-up and professionally supported, with secure access rights. An additional benefit is the possbility to enter or review data from any Partners computer, and by several people simultanously.
  • Depending on particular needs, databases on the server can be accessed in several ways, from an Excel-like Access spreadsheet to a specialized program with screens, menus, etc.