Manuscript Psy02-B@2 · Psy02-B-v2.tex · section · cfd06e2e

Survival and independent viability conditions

Track A6–A7: forward instrument and robust viability › Survival and independent viability conditions · lines 1190–1213

For $R > 0$ , $τ_{OR} = 1/ R$ , the Markov survival test is

D_{R} (τ_{OR}) \geq D_{m i n} ⟺ \frac{κ + γ}{R} \leq lo g (D_{0} / D_{m i n}) - 1.

For a separately specified colored endpoint factor it becomes

κ / R + χ (1/ R) \leq lo g (D_{0} / D_{m i n}) - 1.

Neither is an identification theorem.

Define a viability set by the independent conditions

T \geq 1/ R, D_{R} (1/ R) \geq D_{m i n}, Var (F_{k}) \leq V_{k, m a x}, R_{-} \leq R \leq R_{+}, g_{robust} > 0.

Displayed formulas in this passage

unlabeled · line 1193
$D_{R} (τ_{OR}) \geq D_{m i n} ⟺ \frac{κ + γ}{R} \leq lo g (D_{0} / D_{m i n}) - 1.$
unlabeled · line 1200
$κ / R + χ (1/ R) \leq lo g (D_{0} / D_{m i n}) - 1.$
unlabeled · line 1206
$T \geq 1/ R, D_{R} (1/ R) \geq D_{m i n}, Var (F_{k}) \leq V_{k, m a x}, R_{-} \leq R \leq R_{+}, g_{robust} > 0.$