Selection of Subjects
Quasi experiment needs non-probabilistic sampling.
Choice of Design Type
1st Classification
Randomized design 
Randomized block design (there is a blocking variable (independent variable))
Matched pairs design
Summary
randomized design: 随机分成2组人，每一组根据independent variable给不同的treatment randomized block design: 先把人按照想要研究的block variable分成2组，然后每一组再根据independent variable给不同的treatment matched pairs design: 给同一组人不同的treatment（如果无法实现，就找同一组条件差不多的2个人，2个人分别给不同的treatment）
2nd Classification
Between-subjects vs. within-subjects design
Choice of design type
Instrumentation
Validity Evaluation
Commit Participants
