Table 1.

Comparative age-period-cohort analysis^a

^a Set-up for $G\ge 2$ stratum each defined on the same Lexis diagram with $A$ age groups and $P$ calendar periods

^b Model $\mathit{df}$ for N-PH is $2G\left(A,+,P,-,2\right)$ 

^c Intercept terms for log-scale estimable functions are: ${\mu }_{\mathit{FTT}}^{(k)}=logCrossAg{e}^{(k)}\left(a_{*}\right)$, ${\mu }_{\mathit{CAC}}^{(k)}=\log {\mathit{FTT}}^{(k)}\left(p_{\ast },\left(p_{\mathit{ref}}\right)\right)$, ${\mu }_{\mathit{LAC}}^{(k)}=log{\mathit{FCP}}^{(k)}\left(c_{*},\left(c_{\mathit{ref}}\right)\right)$, ${\mu }_{\mathit{FCP}}^{(k)}=log{\mathit{LongAge}}^{(k)}\left(a_{*},\left(a_{\mathit{ref}}\right)\right)$