Statistical Sciences 1023A/B Chapter Notes - Chapter 4: Dependent And Independent Variables, Sampling Bias, Observational Study
3 Feb 2017
School
Department
Professor
Document Summary
Sample survey: subgroup of a large population is questioned on a set of topics, should be representative of whole population if chosen correctly. Experiment: manipulating the environment in some way, and measuring the effects. Randomized: manipulation is assigned to participants randomly (makes two groups equal in all respects expect explanatory variable) Case-control study: observational study with an attempt to include an appropriate control group. Results of observational studies more readily extended to real world. Meta-analysis: quantitative review of a collection of studies all done on a similar topic, combining information from various researchers may result in the emergence of patterns or effects that weren"t conclusively available before. Census not feasible when trials destroy units (e. g. fireworks) Sample advantages: census isn"t feasible, speed, and accuracy. Easier to train small group of interviewers than a large one. Probability sampling plans: when everyone in the population must have a specified chance of making it into the sample.
