Modelling Lagged Associations in Environmental Time Series Data: A Simulation Study.
Modelling Lagged Associations in Environmental Time Series Data: A Simulation Study.
复制标题
DOI：
10.1097/ede.0000000000000533
复制
发表时间：
2016-11
期刊：
影响因子：
--
通讯作者：
Gasparrini A
中科院分区：
文献类型：
--
作者：
Gasparrini A
本研究评估了研究环境时间序列数据中线性和非线性滞后关联的两种替代方法，通过模拟将基于移动平均摘要的简单方法与更灵活的分布式滞后线性和非线性模型进行比较。结果表明，即使对于长期滞后的关联和存在强烈的季节性趋势，后者也提供了无偏差或低偏差的估计值以及接近名义置信区间的估计值。仅当存在相对较短的滞后期并且正确指定滞后间隔时，移动平均模型才是可行的替代方案。相反，使用移动平均线来粗略地近似长期且复杂的滞后模式，或者指定与实际滞后期不同的间隔，可能会导致严重的偏差。基于分布式滞后线性或非线性模型的更灵活的方法提供了显着的优势，特别是在假设复杂的滞后关联时。
This study assesses two alternative approaches for investigating linear and non-linear lagged associations in environmental time series data, comparing through simulations simple methods based on moving average summaries with more flexible distributed lag linear and non-linear models. Results indicate that the latter provide estimates with no or low bias and close-to-nominal confidence intervals, even for long-lagged associations and in the presence of strong seasonal trends. Moving average models represent a viable alternative only in the presence of relatively short lag periods, and when the lag interval is correctly specified. In contrast, the use of moving averages to roughly approximate long and complex lag patterns, or the specification of an interval different than the actual lag period, can result in substantial biases. More flexible approaches based on distributed lag linear or non-linear models provide noteworthy advantages, in particular when complex lagged associations are assumed.
影响因子：
10.4
作者：
Roberts S
通讯作者：
Roberts S