SDCGLHM/SDCDFOLS: pre-2015 explicit source variable mappings and Master coverage gaps

[DougManuel]

Background

Identified during L0–L7 review of language variables for PR #171 (ethnicity branch). Both variables have coverage that could be extended but no mappings exist in the current PR or v3 worksheets. Deferring from PR #171 to keep scope manageable.

See ceps/cep-010-ethnicity/language-variables-review.md (§L1) for full analysis.

SDCGLHM gaps

Current coverage: PUMF 2007–2018; share files 2009–2012

• 2007–2014 PUMF relies on [SDCGLHM] default — pre-2015 explicit source variable mappings absent
• No Master database coverage

Master source variable: SDCDVLHM — uses a 7-category structure identical to SDCDLNG:

SDCDVLHM codes	→ cchsflow recEnd
1 (English only), 5 (English & other)	→ 1
2 (French only), 6 (French & other)	→ 2
3 (English & French only), 4 (English, French & other)	→ 3
7 (Neither)	→ 4

Any Master extension will require 7→4 category recoding (same pattern as SDCDLNG for SDCGLNG).

SDCDFOLS gaps

Current coverage: PUMF + Master 2011–2014; databaseStart extends to 2017–2018 using SDCDVFLS

• Pre-2015 explicit source variable mappings absent (relies on [SDCDFOLS] default for 2011–2014)
• The 2015–2018 mapping to SDCDVFLS needs verification — MCP shows SDCDFOLS only exists in 2011–2014 PUMF and Master

Open question: Does SDCDFOLS / SDCDVFLS exist in 2015–2018 PUMF/Master? If not, the current databaseStart range overstates coverage.

Tasks

• Verify SDCDFOLS/SDCDVFLS availability in 2015–2018 PUMF and Master via StatCan documentation
• Add pre-2015 explicit source variable mappings to SDCGLHM (PUMF 2007–2014)
• Determine Master coverage available for SDCGLHM (SDCDVLHM cycles)
• Add SDCGLHM Master extension with 7→4 recoding rows (pattern from SDCDLNG in current PR)
• Resolve SDCDFOLS 2015–2018 overstated coverage if confirmed absent

Labels / priority

P2 — no active analysis blocked by these gaps; PUMF coverage is complete for both variables.