MKC2500 Lecture Notes - Lecture 7: Categorical Variable, Univariate, Codebook
18 Jun 2018
School
Department
Course
Professor
ETC2500 notes
The Data Preparation Process
1. Preliminary plan of data analysis
2. Questionnaire checking
3. Editing
4. Coding
5. Transcribing
6. Data cleaning
7. Variable respecication and recoding
8. Selecting a data-analysis strategy
Questionnaire checking
• Look for missing or unexpected responses, may indicate missing information
• Questionnaire is modified if needed
Editing
• Review all completed questionnaires
• Keep track of the number or proportion of unusable questionnaires
Coding
Use numbers to assign categories to categorical data
• Even if a string variable could have been used
Include any information about the sampling process you can
• Order of respondent
• Interviewer (if more than one used)
• Date, time and location of interview
Use fixed field codes wherever possible, i.e.
• Every respondent answers all questions
• All questions coded using same number of variables
Document Summary
The data preparation process: preliminary plan of data analysis, questionnaire checking, editing, coding, transcribing, data cleaning, variable respecication and recoding, selecting a data-analysis strategy. Questionnaire checking: look for missing or unexpected responses, may indicate missing information, questionnaire is modified if needed. Editing: review all completed questionnaires, keep track of the number or proportion of unusable questionnaires. Use numbers to assign categories to categorical data: even if a string variable could have been used. Include any information about the sampling process you can: order of respondent. Interviewer (if more than one used: date, time and location of interview. Use fixed field codes wherever possible, i. e. every respondent answers all questions: all questions coded using same number of variables. Try not to ask many open ended questions: however, if you do, assign each response to its own column, code with a 1i f response is indicated, 0 if otherwise.
