might be more representative than simple random sampling Disadvantages need names of all population members difficulty in reaching all selected in the sample o Oversampling : You want to do cluster sampling, but some of your clusters are really small (e.g., Native American) A study on some characteristic or mechanism in the population But you think that race makes a difference But there are very few Native American and African American people in community Do simple random sampling from each race but NA and AA Ex: You set a percentage of how many African American and Native American people you want in the study (say, 10%), and from their group, you take more people – but randomly o Cluster Sampling : Will lead to same result as random sampling Population is divided into groups, usually geographic or organizational Some of the groups are randomly chosen Whole cluster is sampled 22
Ex: You want to randomly sample high school students in Florida Start with a list of high schools (that’s your cluster variable) Take a random sample of, say, 20 of those HS Interview each student in each selected HS o Multi-Stage Sampling : Start with same steps as cluster sampling, but then add simple random sampling to each cluster Ex: You want to randomly sample high school students in Florida Start with a list of high schools (that’s your cluster variable) Take a random sample of, say, 20 of those HS Then, from each school, randomly select students from entire student population Advantages Efficient researcher doesn’t need names of all population members Disadvantages fewer sampling points make it less like that the sample is representative o Systematic Sampling : Selecting individuals within the defined population from a list by taking every K th name. For example, imagine I asked you to name yourselves from numbers 1-10, in order I randomly pick two numbers between 1 -10 (say, 3 and 7) and sample every person who got a 3 or 7 Advantages Sample selection is simple Disadvantages need names of all population members difficulty in reaching all selected in the sample Sampling Error: amount of inaccuracy in estimating some value that is caused by only a portion of a population (i.e. a sample ) rather than the whole population. Chance or luck observations. - sampling error is to be expected - to avoid sampling error entirely, a census of the entire population must be taken 23
You've reached the end of your free preview.
Want to read all 24 pages?
- Fall '08
- external validity, internal validity