<<

Producing Data, Randomization, and Experimental Design Goals

 Identify observational studies versus

 Design experiments to test hypotheses using appropriate randomization

 Use the random number tables to assign subjects correctly to experimental groups

 Define, use, and know the concepts behind all the new vocabulary words

Starting with a research question

 Often can’t simply study the whole population

 If you want to know the life expectancy for cancer patients you simply can’t identify all patients and then wait for them to die

 Use a to draw conclusions about the whole vs.

 Observational study observes and measures variables of interest.

 Experiment imposes a treatment in order to observe

New Terminology

 Population Sample

 Voluntary response sample

 Convenience

 Bias

(SRS) Use Random Numbers to Generate SRS

 Label all the individuals in a population with numerical labels.

 Use random number table (or statistical package) to choose individuals randomly.

 Example: To divide 100 students into two groups of 50 label them 00 to 99 and go through the table starting at a random line until the first 50 have been chosen for a group. Other Sampling Designs

 Probability sample

 Individuals chosen with some given probability

 Stratified random sample

 Population divided into strata and individuals chosen at random from each strata

 Multistage random sample

 Sample chosen in a number of stages Problems

 Undercoverage: some groups have no chance of being sampled; as in phone polling

 Nonresponse: individuals chosen cannot or will not participate

 Response bias: people may lie about illegal or embarrassing behavior; may respond to the questioner

 Wording of questions may effect the outcome Designing Experiments

 Experimental units or subjects

 Treatment: experimental condition imposed

 Factors: explanatory variables

 Level: value of a given factor Example: Television ads Comparative Experiments

 Compare 2 or more groups

 Use a control group to eliminate and placebo effect

 A randomized comparative experiment uses comparisons between two (or more) groups and randomization of subjects into treatment groups. Design of a randomized comparative experiment Principles of Experimental Design

 Control effects of lurking variables via comparison of several treatments

 Randomization to assign units to treatment

of experiment on many units to reduce chance variation