# 这个作业来自美国的关于用R语言对实验案例进行建模、方差分析以及回归分析的R语言代写，作业详情可扫码咨询客服

STA 466/566: Experimental Design Methods

1.1实验方法和范例（蒙哥马利1.1-1.2）

（C）混合时间（60秒； 80秒）； （D）批次重量（1500磅； 2000磅）； （E）之间的延迟

Giles-Sims，由Oehlert描述。 1979年开始对14至21岁的女性进行大规模调查，

1988年和1990年。在其他情况下，他们被问及孩子的反社会行为，以及

1988年通常在1990年有更多的反社会行为。打屁股会引起反社会行为吗？

Treatments: the different levels of a factor or different combinations of levels of several factors.
Nuisance factors: Sometimes there are factors in the experiment that explain some of the variation in the
response but are not a part of the treatment. They should be included in the analysis because they reduce
the amount of unexplained variability in the response. If these factors are discrete, they are called blocks. If
they are continuous, they are called covariates.
Experimental Unit: the object on which the treatments are applied; the Measurement Unit is the object
on which the response is actually measured. Often experimental and measurement units are the same, but
not always.
1.2 Some Basic Principles of Design (Montgomery 1.3)
Replication: independent repetition of a treatment.
Independent replication vs. repeated or nonindependent measurement.
1. Four subjects are assigned to a drug and four are assigned to a placebo. For each, three different blood
draws are taken and a measurement is obtained from each.
2. Four subjects are assigned to a drug and four are assigned to a placebo and a measurement is taken on
each.
3. Retinal cell example. The cells are grown in containers. Consider a version of the experiment with
four treatments, where each treatment is made by mixing up a set of compounds. Suppose four
mixtures are made, each consisting of a different treatment. Then, three samples are taken from the
mixes and administered to cells in separate containers. So a total of 12 containers are used (three for
each treatment).