Table 2

Conditional least squares ARIMA and transfer function estimates for the intervention model explaining the ratio of Tasmanian restaurant to retail turnover

ParametersThe model of abrupt and permanent impact*The model of gradual and permanent impact†The model of abrupt and temporary impact‡
Coefficientp ValueCoefficientp ValueCoefficientp Value
*(1−B)Yt = θ0+(1−θ1B)at+ωIt, where Yt is the value of the ratio at time t, and It is intervention dummy variable coded “0 = before intervention” and “1 = thereafter”.
†(1−B)Yt = θ0+(1−θ1B)at+[ω/(1−δB)]It, where Yt is the value of the ratio at time t, and It is intervention dummy variable coded “0 = before intervention” and “1 = thereafter”.
‡(1−B)Yt = θ0+(1−θ1B)at+[ω/(1−δB)]It, where Yt is the value of the ratio at time t, and It is intervention dummy variable coded “0 = before and after intervention” and “1 = at the moment of intervention”.
θ0−0.0000.875−0.0000.885−0.0000.809
θ10.382<0.0010.382<0.0010.375<0.001
ω−0.0010.641−0.0010.6400.0000.672
δ0.3440.832−0.993<0.001
Q60.810.9760.800.9770.630.986
Q122.240.9982.240.9982.290.997
Q186.570.9886.630.9886.630.988
Q2425.780.31225.710.31426.620.272