# L5 — Family density matrix (recovery concepts)

Research-only. No scoring.

## Matrix legend

- **D** depth (years usable) · **F** frequency · **M** missingness risk · **B** methodology-break severity · **P** PIT · **S** shock detectability · **R** recovery observability  
- Scale: H/M/L qualitative metadata judgments

| Family | D | F | M | B | P | S | R | Recovery-concept fitness |
|--------|---|---|---|---|---|---|---|--------------------------|
| Consumer sentiment SCA | H | H | M early / L modern | M | H | H | H | **Primary societal monthly lead** |
| Conference Board CCI | H | H | L (if access) | M | H | H | H | Parallel monthly lead |
| SCA expectations components | H | H | M | M | H | H | H | Primary expectations lead |
| SCE | L | H | L | L | H | M | M | Short-panel mechanism |
| L1 trust individual | H span / L density | L | H | H | H when present | M | L | **Not for half-life** |
| Gallup approval | H | M (era-dep) | M | **H** | H | H | H within era | Era-stratified primary political |
| Approval aggregators | M | H | M | M | M (revisions) | H | H | Recent high-freq political |
| Generic ballot | M | M | M | M | M | M | M | Seasonal confound |
| ACLED protests | M | H | M | M | H | H | H event | Event damping lead |
| GDELT | H | H | M | H | H | H | M validity | Needs construct lock |
| Media tone | M | H | M | M | H | H | M cross-layer | Narrative half-life |
| VIX/markets | H | H | L | L | H | H | H | External benchmark only |
| Google Trends | M | H | M | M | M | H | M | Attention ≠ attitude |
| ANES | H span / L N | L | H | M | L | L | L | Insufficient grid |

## Recommended next research pulls (still no scoring)

1. Pin SCA ICS + components as recovery-concept **density reference** (already partially landed).  
2. Build Gallup **era-stratified** coverage ledger (no splice).  
3. Document access path for Conference Board + ACLED full US.  
4. Keep Exp002/L1 trust on the **anti-pattern** shelf for recovery half-life.  
5. Optional: aggregator snapshot protocol for PIT honesty.