Info | ||
---|---|---|
| ||
Expand | ||
---|---|---|
| ||
Minutes of Simple Codebook group, Tuesday June 21, 2016 Attendees: Steve McEachern, Larry Hoyle, Oliver Hopt Review of the output from Norway continued. The group focussed on how to resolve the outstanding fields identified in the DDI-C profile (from Larry's spreadsheet). The proposal was for the remaining content to be addressed through three mechanisms: 2. Variable metadata: 3. ResponseUnit and AnalysisUnit (and Unit of Measurement) Variable content characteristics may be best addressed now: - VariableInterval, Recommended for deferral - All fields within Methodology section of DDI-C - Also includes Imputation This requires output of Methodology team 2. PhysicalLayout
|
Expand | ||
---|---|---|
| ||
Codebook meeting 2016-05-10 Attending: Dan Gillman, Gillian Kerr, Oliver Hopt, Larry Hoyle We reviewed the spreadsheet https://docs.google.com/spreadsheets/d/1VDbVz2KRRSX_KEf0IfuE-QqMyTDupftCZfBdBM6VPT8/edit#gid=1652443366 , sheet NewStartingPointCdbk_4 Which now has xpaths to ddi2.5 elements (column F) and description of the DDI4 classes (column E) which correspond as well as descriptions of needed DDI4 classes We discussed the creation of a view. We need two new classes one for the whole activity producing data and one to describe each wave, or phase. Issues are associated with the top level (e.g. design) but then there are specifics at each repeated instance producing different data. The general and specific shouldn’t be duplicative for one time activities. An example of the top level information would be the purpose for the whole set of activities. Another would be the funding source for the whole, or authorizing legislation. What terms could we use? --- Activity? Data capture activity? Need anchor class and specific class “anchor class” and “concrete anchor instance class”
In stats agencies ongoing activities – designs change, the overall is known by a name and has a funding source. (e.g. CPS or American community survey). The specific might be a monthly collection e.g. monthly CPS as input to calculation of unemployment rate. Another example would be the Christmas bird count which has annual data collections but can also be considered to be an overall series. Decision: “StudySeries” as overall “Study” for the specific – the user community is familiar with this term, even if developers don’t like it Conceptual would be the best current package for these classes Oliver will create classes then the rest of us can work on descriptions. Larry will add other classes to the view
Goodbye “TOFKAS” (The Object Formerly Known As Study). Even Prince went back to “Prince” |
...
Expand | ||
---|---|---|
| ||
Codebook meeting 2 February 2016 Attending: Dan, Michelle, Steve, Oliver, Jon, Larry, Jared There’s some lack of clarity about where this group is at. Discussed what to include in simple codebooks. One idea is to review the spreadsheet of common elements (summary of CESSDA) and build on that. Essentials seem to include: enough information to read the data into statistical package, label values, understand universe, understand what measure means so you can interpret the data, attribution information. Another idea is to look at examples of simple codebooks, identify what they use, and then map to a model. We need to be careful to keep things simple. Even older versions of DDI 2 weren’t exactly simple. If we nail down definitions, then do we make instances of previous versions incompatible? As we define what information elements we want in DDI 4.0, we can specify which element you want in 2 if you’re going backwards. Next steps:
|
...