American Time Use Survey: Report on Dress Rehearsal

1
American Time Use SurveyReport on Dress
Rehearsal

• Diane Herz
• American Time Use Survey
• IATUR meetings in Lisbon, Portugal
• October 16-18. 2002

2
The ATUS Dress Rehearsal

• Mimicked full-production conditions
• 2 months sample from March and April household
  surveys (3,240)
• CATI interview
• 8-week fielding period
• 1 designated person/household
• Each assigned designated day
• Average 18-minute interview
• x-day training provided

3
Telephone householdsResponse and refusal rates
4
Main reasons for refusal

• Reluctance to do survey because they just did the
  household survey
• Too busy/not enough time

5
Telephone households Contact rates
6
Reasons for high noncontact rate

• Calling protocol was not followed at call center
• Case management system sent cases into hold
  queues
• Many cases were left in callback status
• Process improved in week 4 of April sample

7
No-telephone-number households additional
protocol

• Sent 40 incentive with advance materials
• Requested respondents call in to do interview

8
Response rates were low for no-telephone-number
households
9
Demographics of respondents and refusals
10
Number of activities reported in diary
11
Representativeness of days
12
Overview of coding results

• Average number of codeable activities per case
  19
• Average number pre-coded during interview X
• Average time to code case declined each month
• Most errors were interviewing errors, not coding
  errors
• Some problem areas some easy areas

13
Coding efficiency increased during the field
period
Minutes to code case
14
Debriefing results

• Operations
• Fix the system
• Add operations training
• Training
• Develop FAQs for specified populations
• Provdie document on applications in other
  countries

15
Debriefing results

• Questionnaire
• Enjoyed conversational interviewing
• Explain paid work questions
• Provide all-day child care button
• Respondents
• Work to avert labor force survey refusals
• Coding
• Liked coding
• Instrument worked well
• Add some activities
• Clarify some rules

16
Estimates by activity and sex
17
Secondary child care estimates
18
Changes resulting from dress rehearsal

• Operations
• Went to prefielding early to correct case
  management operations
• Training
• Added flow charts desk aids to coding training
• Clarified paid work questions
• Provided FAQs and applications job aids

19
Changes resulting from dress rehearsal

• Coding
• Added some categories examples
• See Tina Shelleys presentation!
• Data collection software
• Added reason for refusal question
• Added interviewer data quality question
• Added all-day child care button
• Showed awake time for child care questions

20
Next Steps

• Early prefielding (Sept/Oct)
• Prefielding (Nov/Dec)
• Full production (January 2003)
• Data release (mid-2004)