Communication Research
Asking Questions, Finding Answers
Joann Keyton, Ph.D.
Chapter
7
Sampling,
Significance Levels, and Hypothesis Testing
Ready to Review?
These websites, from
Generating Random
Numbers
Several different web sites offer the opportunity to download random number generators. The Research Randomizer can be used online and the results saved to your computer. It is a free service designed to assist students and professors in generating random numbers for research purposes. The Web site is complete with a tutorial and help functions to assist the user in setting up a number generating program.
Sampling the
Web
Because the World Wide Web is so vast and changes so rapidly, it can be a fascinating context for generating and answering research questions. But how could a researcher sample from the entire population of Web pages? Probability sampling would not be feasible due to the dynamic capacity of the Web. Using nonprobability sampling is the answer. Herbeck and Hunter (1998) demonstrate that constructing careful procedures for determining the sample helps to ensure that the sample represents the population. Using a list of university student web servers, the research team selected four universities to represent public and private universities, small and large institutions, and four geographical regions. Next, over a two-day period, they divided the total number of student-authored web pages on each server by 100 to obtain a random sample. While their sampling procedure would not qualify as a probability sample, Herbeck and Hunter devised a procedure that resulted in a diverse set of web pages to examine.
Back to Web listings by chapter