Time series valued experimental designs: A review.

Area of Science:

• Statistics
• Experimental Design
• Time Series Analysis

Background:

• Literature review on time-series valued experimental designs.
• Designs categorized by time variable status: experimental factor vs. replication context.

Purpose of the Study:

• To review existing literature on time-series experimental designs.
• To survey analysis methods, including time and frequency domain approaches.
• To discuss applications and potential pitfalls in data analysis.

Main Methods:

• Literature synthesis of experimental designs involving time-series data.
• Review of time-domain and frequency-domain analysis techniques.
• Survey of signal detection models, Bayesian methods, and optimal design strategies.

Main Results:

• Categorization of designs based on the role of time.
• Identification of analysis methods applicable to time-series experiments.
• Highlighting the risk of misleading results using standard F-tests on correlated data.

Conclusions:

• Time-series experimental designs require specific analytical considerations.
• Standard statistical tests may be inappropriate for highly correlated time-series data.
• Further research and application in fields like medical experiments and field trials are supported by this review.