American Intercontinental University
Statistics
June 26, 2008
After reading an article regarding how sample size can affect clinical research, I decided the composition by Mr. David Streiner would be an interesting and informative topic to write about. In his composition “When is Enough Enough?” Streiner (2006) dissects the ways that sample size can have an effect on research, studies, and analysis.
According to Streiner, there are three apparent ways in which smaller sample sizes can be inherently problematic for a researcher, which include findings too generalized, the power of analyses, and assurance of results (2006). Streiner’s first example has to do with the power of analyses, involving a study by Lind regarding the effects of lemon on individuals ailing from the affliction known as scurvy (vitamin C deficiency). During the study, there were basically six groups of two subjects each. According to Streiner (2006), not only can this small of a sample may lead to an inaccurately impressive finding, it can also lead to small effect going relatively unnoticed altogether.
According to Streiner (2006), sample size may have an effect on the width of a confidence interval that cannot be disregarded. Because small sample groups have a more significant error margin, the confidence interval is henceforth wider. Streiner goes on to reiterate that the overall goals of studies are to document the numerical significance of differences, not just prove that sample groups are different (Streiner, 2006).
Since errors occur at random intervals, larger numerical samples may be required in order to cut down standard error percentages. But Streiner does caution against completely rejecting or discounting studies based on ...