Approaches to the analysis of behavioral medicine data with few individuals but many repeated observations are discussed. Difficulties with common analytic strategies are noted and applications from econometrics using cross-sectional pooled time series designs are recommended. Additional complications due to outliers, missing data, measurement error, and aggregation bias are discussed.

中文翻译：

---

具有许多观察结果的时态数据的统计分析：行为医学数据的问题

讨论了对少数个体但多次重复观察的行为医学数据进行分析的方法。指出了常见分析策略的困难，并建议使用横截面汇总时间序列设计进行计量经济学应用。讨论了由于异常值、缺失数据、测量误差和聚合偏差而导致的其他复杂情况。