Based on the factorization approach, we show that the CLEO data for the ratio Γ(B → ψK * )/Γ(B → ψK) and the CDF measurement of the fraction of longitudinal polarization in B → ψK * can be accounted for by the heavy-flavor-symmetry approach for heavy-light form factors provided that the form factor F 0 behaves as a constant, while the q 2 dependence is of the monopole form for F 1 , A 0 , A 1 , and of the dipole behavior for A 2 and V . This q 2 extrapolation for form factors is further supported by B → K * γ data and by a recent QCD-sum-rule analysis. We then apply this method to B → D(D * )π(ρ) decays to extract the parameters a 1 and a 2 . It is found that a 1 (B → D ( * ) π(ρ)) = 1.01±0.06 and a 2 (B → D ( * ) π(ρ)) = 0.23±0.06 . Our result a 2 /a 1 = 0.22±0.06 thus significantly improves the previous analysis that leads to a 2 /a 1 = 0.23 ± 0.11. We argue that, contrary to what anticipated from the leading 1/N c expansion, the sign of a 2 (B → ψK ( * ) ) should be positive and a 2 (B → ψK ( * ) ) > ∼ a 2 (B → D ( * ) π(ρ)).