preview

Consequences Of Statistical Inferences Providing A Result

Decent Essays
Consequences of statistical inferences providing a result rather than an answer could include researchers incorrectly applying related management implications to their studies, when in reality they are applying a treatment that was not correctly interpreted. This could result in the new study producing a non-significant result when a biologically significant result should have been detected. Researchers desperate for a significant P value may claim an observation that was not present in the study, or perform researchers degrees of freedom, which is defined as decision making as researchers conduct the study, and not beforehand.
b. “Truth” cannot be derived from a single study; it must be repeated to increase the precision of the study and therefore the strength of inference made about the results. Often, using the same sample and similar methods will not procure similar results, and therefore a consensus of studies should be provided to make a general claim. Studies that truly follow the scientific method and include randomization and repeatability will be closer to discovering the ‘truth’ than other studies.
c. In the vitamin E study, the first studies conducted on the effects on vitamin E continued to be cited and defended even when new evidence showed their results were incorrect because they ‘felt true.’ This still happens in natural science; for example, McCullough’s study on the George Reserve in Michigan was based on 1 population of deer consisting of a small
Get Access