It is important to note that Model (1.1) and Model (1.2) take different approaches
to estimating whether the effect of PM10 has changed over time. In Model (1.1) the
effect of PM10 changes at a single discrete time, while in Model (1.2) the effect of
PM10 changes continuously over time. In particular, the two models need not reach
the same conclusion when fitted to data from a particular city. For example, if the
effect of PM10 changed relatively abruptly (due to, for example, a sudden change in
emission standards) then it is possible that Model (1.1) would detect this change but
not Model (1.2). Conversely, it is possible to hypothesize scenarios where Model (1.2)
detects a change in the effect of PM10 but not Model (1.1).