Survey Design and Procedures II and Data Cleaning Dr' Travor C' Brown

1
Survey Design and Procedures- II and Data
CleaningDr. Travor C. Brown
2
Objectives

• Examine key issues in cover letters
• Examine issues in data editing, coding and
  cleaning issues

3
Cover letters

• Keep to one page
• Present purpose and linkage
• Discuss confidentiality
• Provide contact person and contact info
• Discuss incentives
• Encourage a quick response
• Discuss due dates
• Discuss any id numbers or funders

4
Data Cleaning

• Coding
• Changing response values to numbers
• Must be exhaustive
• Mutually exclusive
• Appropriate to research question
• Use of precoding (see survey) vs a code book (see
  p. 427)
• Use of a coding log/decision rules log
• Do not know issues.
• Editing
• Detects errors, omissions, and corrects them when
  possible
• Designed to guarantee
• Data accuracy
• Completeness
• Arranged to improve coding

5
Data Entry

• Optical and voice available
• Usually manual input into a data base
• Hintcan you enter in excel 2.1 and then open in
  SPSSfaster

6
Preliminary Analysis

• Run distributions (I.e, frequencies, etc)
• Check for missing data and errors
• 6 when the scale is 1-5
• Check for measures of central tendency
• Mean, median and mode (as appropriate)
• Check for measures of spread
• Range, standard deviations, variance
• Measures of shape
• Skewness should be between 1 and 1
• Negative skewed to the left
• Positive- skewed to the right
• Kurtosis-measure of flatness
• Between 1 and 1 is desired