STAT 110 1nd Edition Lecture 4 Outline of Previous Lecture I. Lurking/Confounding Variables II. Placebos and the Placebo Effect III. Randomized Comparative Experiment (RCE)IV. Principles of Experimental Design V. Comparing OutcomesOutline of Current LectureI. Experiments In the Real World a. GoalsII. 3 Experiments to Meet Our Goalsa. Completely Randomized Experimental Designb. Block Designc. Matched Pairs Design Current LectureI. Experiments In the Real Worlda. The goal of experiments is to generalize our results to the population of interest and defend those results. In order to meet these goals, we must have the following three things:i. Statistically significant results ii. Realistic settingsiii. Repeatability: that is, we must be able to replicate the results II. 3 Experiments to Meet Our Goals a. Completely Randomized Experimental (CRE) Design i. In a CRE Design, all subjects are randomly assigned to the different treatments. You can compare the CRE Design to the Simple Random Sample of a survey designs method. b. Block Designi. In a Block Design experiment, we separate subjects, which are similar in some way, into groups called blocks. These blocks are then randomly assigned to different treatments. You can compare the Block Design to the Stratified Random Sample of a survey designs method. These notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.1. Block designs allow us to draw specific conclusions about each block, and helps to reduce confounding by accounting for blockingvariables. Blocking variables are lurking variables within each block, such as gender, age, ethnicity, or income bracket. c. Matched Pairs Designi. A Matched Pairs design is a type of Block Design in which two treatments are compared by pairing subjects together that are as similar as possible. For example, two Caucasian, twenty year old females might be paired together. Then, each member is randomly assigned to a treatment. 1. Sometimes, a matched “pair” is only one person, in which case, that person receives both treatments, in a random
View Full Document