Types of error common in entry or translation into analysis files

• Clerical error: transposition of #s, missing a line, overlooking an observation, typos

• Incorrect assignment of data to variables

• Incorrect recodes

• Incorrect labeling of non-quantitative variables

• Proofreading

• Double entry of data

• Hire professional keypunchers

• Range checks

Type of Error

Approach to control

• Use data management software

• Inspect results of recodes

• Label all recodes

Previous slide

Next slide

Back to first slide

View graphic version