Probability & Statistics: Week 10 Learning Objectives for Chapters 7 & 8 - Prof. Deborah S, Study notes of Statistics

Download Probability & Statistics: Week 10 Learning Objectives for Chapters 7 & 8 - Prof. Deborah S and more Study notes Statistics in PDF only on Docsity! MATH-1530- 05/07/15/17 Probability & Statistics Fall 2004 / Week 10 LEARNING OBJECTIVES (Chapters 7 & 8) SOME DEFINITIONS:  Taking a census is the name of a process that attempts to survey or obtain measurements (on one or more variables) from a population of “individuals.” Sampling is the name of the process where only a subset of the population is selected for measurement.  Research often falls into one of these two categories : observational studies versus experiments. The design of a study is biased if it systematically favors certain outcomes. Anecdotal evidence is based on a haphazardly selected individual case and is rarely trustworthy.  A variable is described as lurking if it has an important effect on the relationship among the variables in a study but is not included among the variables actually measured in the study. Two variables (either explanatory or lurking variables) are confounding when their effects on a response variable cannot be distinguished from each other. SAMPLES AND SURVEYS: Review chapter 7 for definitions of different types of random samples such as simple random sample (SRS), stratified random sample, multi-stage random sample, systematic random sample, and cluster sample. A. Recognize when a simple random sample (SRS) of size n has been chosen from a population and what implication that has for the probability that any set of n individuals in the population is the sample chosen. Understand the importance of random selection of the individuals for samples. Be able to use the table of random digits to select an SRS. B. Recognize when stratification is a suitable sampling technique, and use the table of random digits to choose a stratified random sample. C. Identify (name) other popular sampling methods after reading a detailed description of that sampling process as used in a given study. D. Recognize the presence of undercoverage and nonresponse as sources of sampling error in a survey. E. Recognize the bias in voluntary response samples and other haphazard sampling methods that are inferior to random sampling. F. Identify the population, sample, or sample frame in a given sampling situation; understand the distinction between census and sample. EXPERIMENTAL DESIGN: Review chapter 8 for definitions of terms and different types of experiments. An experiment is a study for which we actually do something to people, animals, or objects in order to observe the response. The objects on which the experiment is performed are the experimental units (often called subjects or participants if they are human beings). A specific experimental condition applied to or imposed on the units is called a treatment. The explanatory variables in an experiment are called factors. A factor will often have two or more levels. A. Identify the experimental units, response variable, the factors, levels, and treatments in a single- or multi-factor experiment. The outline for an experiment (using a line and arrow flowchart) should include specific descriptions about the experimental units, the randomization scheme, the sizes of the treatment groups, the specific treatments, and the response variable being measured. Draw an outline of a completely randomized, one-factor experimental design. Do the same for an experiment with 2 or 3 factors. Be able to compute the necessary number of treatments and experimental units for experiments having more than one factor. Use the table of random digits to carry out the random assignment of exp. units to groups in a completely randomized experiment. B. Recognize when the use of blocks is an appropriate method of control in an experimental design. Outline a randomized block design, with a single factor and use the table of random digits to carry out the randomization in the separate blocks. C. Recognize when matched pairs design may be appropriate in an experimental design. D. Recognize and understand the concept of control groups, the placebo effect and the use of blind or double-blind techniques. PRINCIPLES OF EXPERIMENTAL DESIGN: CONTROL of the effects of lurking variables on the response, most simply by comparing several treatments. RANDOMIZATION is the use of chance to assign exp. units to treatments in order to eliminate the bias that may be caused by systematic differences among treatment groups. REPLICATION of the experiment is carried out on enough exp. units or subjects to reduce chance variation in the results. WARM_UP QUESTIONS 1. The process of collecting data from an entire population is called what?