Organizing Groups

1
Organizing Groups

• Why, When and How to do it

2
Brief history of the need for groups

• In the old days of MARC records you could either
  be a machine readable file OR you could be a
  seriesbut not both
• We all knew that some data sets were more closely
  aligned than others but we werent allowed to say
  so
• DDI carried on this tradition by focusing on the
  codebook

3
Need for change

• Multiple files could be processed using the same
  codebook
• Some data collection processes were designed to
  be a series
• Longitudinal series
• Time series
• Waves
• Sometimes multiple files were integrated in
  order to make a single file described by a single
  code book

4
Why Group?

• Identifies study units that were intended to be
  analyzed together
• Provides comparative information for accurate
  linking where appropriate
• Allows for easier reuse of information and less
  redundancy
• Provides new discovery, analysis, and management
  options

5
When does grouping occur?

• Whenever it wants to
• Predetermined grouping patterns
• Occurs once the second study unit for the group
  is being prepared
• Standards can be moved into Group and used by
  subsequent modules
• Post-facto grouping
• Full instances exist already
• Comparative information is the primary concern

6
How to group

• Determine the criteria for the group
• Identify common items for standards
• Concepts note that comparability is based on the
  description NOT the label
• Questions
• Variable/categories
• Fill out the grid

7
The Grid

• Time
• Single or multiple
• Regular or Irregular occurrence
• Continuing or limited duration
• Instrument
• Single or multiple
• Panel
• Geography
• Data Sets produced (integrated etc.)

8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)