联系我们： 手动添加方式: 微信>添加朋友>企业微信联系人>13262280223 或者 QQ: 1483266981
这是一篇美国作业是一个R统计作业的Project
Components
Motivation
Short section discussing topic with motivation, questions/issues of interest, etc.
Design
The design should include details such as…

experimental units, measurement units, outcomes, etc.
target of inference: conducting inference for just units in experiment or is there some
population you are generalizing to
which factors are random and which are fixed
the actual levels of factors
the number of observations
how randomization will be done (e.g., a single randomization, randomization separately
by block, randomization separately by split plots, etc.)

It is not necessary to include a power analysis.
It is important that you design an experiment, not an observational study; make sure that you are assigning experimental units to treatments and not simply observing them.
Analysis Plan
Please include an analysis plan. The level of detail I’m looking for may include, for example, a partial ANOVA table including degrees of freedom, expected mean squares, and hypotheses you wish to test with statement of appropriate test statistics. An R code fragment for essential parts of the analysis could be included.
You will not be doing an actual experiment or otherwise collecting data; the assignment is simply to design a plausible experiment. There should be no actual data or actual analysis in your paper, just a discussion of how analysis would proceed.
Providing thoughtful details will help enrich your project. You should consider problems you may realistically run into (e.g., violation of assumptions, possibility of units dropping out of study, sources of unwanted variation) and explain (at least briefly) how you would deal with them.
How complex should my design/analysis plan be: You should demonstrate that you have a thorough understanding of the course material and enough mastery to design your own experiment. I also want you to include what I call a “level up.” This means at least one piece of your design or analysis is not something we explicitly covered in class. You only need to do one level up. For instance, you might…
…explore a new design and the appropriate analysis method for such a design
Examples: partially balanced incomplete block designs, rerandomization, conjoint design
…or discuss a new analysis method or component
Examples: multiple testing procedure not covered in class, different model for data, new way to assess assumptions, tests for quantities other than means
…or consider a problem you may encounter that we have not covered in class and discuss one strategy for dealing with it in the design and/or analysis stage
Examples: methods to deal with unbalanced data not covered in class, non compliance, censored data, missing data