Table 2. Adjusted ARIMA model δ .
| Regressive Parameters, by month € | Coefficient | 95% | CI | p |
|---|---|---|---|---|
| 1 | -0.41 | -0.91 | 0.11 | 0.124* |
| 2 | -0.33 | -0.85 | 0.18 | 0.201* |
| 3 | -0.56 | -0.88 | -0.24 | 0.001 |
| 4 | -0.55 | -0.93 | -0.15 | 0.007 |
| 5 | -0.46 | -0.88 | -0.04 | 0.033 |
| Random disturbances, by month ¥ | ||||
| 1 | -1.28 | -1.89 | -0.67 | 0.000 |
| 2 | 0.61 | 0.02 | 1.21 | 0.042 |
| Seasonal regressive parameters, by month β | ||||
| 12 | -0.64 | -1.00 | -0.28 | 0.001 |
δ The parameters correspond to the observed effects for a differentiated series, i.e., a series where the trend and seasonal deterministic patterns were removed first, in order to improve the stochastic estimation in the stochastic model.
* Non-significant at a level of 0.05
€ Previous months identified as important for predicting the variation of the current month
¥ Disturbances occurred during a previous month, which affect the current realization of the phenomenon
β Seasonal effect (past year value) that affects the current value of the time series