Figure 1
From: How much can we trust life tables? Sensitivity of mortality measures to right-censoring treatment
The effect of random fluctuation on the identification of the modal age at death (identified by a dashed line). The red curve corresponds to the dx column for the Swedish male population in 1970 (Source: HMD, 2015). Without smoothing, the modal age at death is misspecified by about 3 years because of random fluctuation. The ΓGM fit (blue curve) is almost identical to the non-parametrically smoothed (by LOESS) dx (green curve).
