Data Management Lab: Session 3 Data Entry Best Practices
1 like622 views
The document outlines best practices for data entry management, emphasizing the importance of dataset integrity, the separation of coding and data entry tasks, and employing robust data entry programs to detect errors. It suggests careful quality control measures, including double entry and consistency checks, as well as systematic variable naming and labeling. Additionally, it recommends organizing variables into groups for easier analysis and ensuring long-term data storage in consistent formats.
Data Management Lab: Session 3 Data Entry Best Practices
- 1. IUPUI University Library Center for Digital Scholarship Data Management Lab: Spring 2014 Data Entry Best Practices Data Entry 1. Dataset creation and integrity a. Separate the coding and data entry tasks as much as possible b. Coding should be performed so that distractions to coding tasks are minimized c. Arrange for particularly complex tasks to be carried out by people specially trained for the task d. Use a data-entry program that is designed to catch typing errors (i.e., one that's pre- programmed to detect out of range values) e. Perform double entry of data f. Carefully check the first 5-10 percent of the data records created, then choose random records to quality-control checks throughout the process g. Let the computer do complex coding and recoding, if possible 2. Things to check a. Wild codes and out-of-range values b. Consistency checks - comparisons across variables c. Record matches and counts - relevant in longitudinal studies where subjects may have more than one record and varying numbers of records 3. Variable names a. Prefix, root, suffix systems is a systematic approach (compared to one-up numbers, question numbers, and mnemonic names) 4. Variable labels a. Should provide three pieces of information i. The item or question number in the original data collection instrument ii. A clear indication of the variable's content iii. An indication of whether the variable is constructed from other items 5. Variable groups a. Groups are recommended if a dataset contains a large number of variables b. Can effectively organize a dataset an enable secondary analysts get an overview of a dataset quickly 6. Over the long-term, store data in a consistent format References 1. ICPSR. (2012). Guide to Social Science Data Preparation and Archiving, University of Michigan, Ann Arbor, MI. From http://www.icpsr.umich.edu/files/deposit/dataprep.pdf. 2. Scott, T. 2012. Guidelines for data collection and entry. From http://www.mc.vanderbilt.edu/gcrc/workshop_files/2012-09-07.pdf 3. DataONE Education Module: Data Entry and Manipulation. DataONE. From http://www.dataone.org/sites/all/documents/L04_DataEntryManipulation.pptx Heather Coates, 2013