Concept of Hypothesis Testing

Based On the article, circulation-2006-Davis-1078 Roger B. Davis and Kenneth J. Mukamal focus on the concept of hypothesis testing, a process used to conclude on the basis of the statistical testing of the data gathered in addition to a suitable approach to test means. The authors assert that in biomedical research, researchers often hypothesize the relationship of diverse factors, collect data to examine an existing relationship and lastly, attempts to conclude as related to the connection from the collected data. From the article, it is clear that researchers often test the correlation through making the comparison between average levels of the factors between two groups or even between one group and standard references. From the article, I have learned various standard methods used in testing hypothesis on the means of the observed measurement including; a transformation of research question into a null hypothesis and the alternative hypothesis. These related to what we learned from the book elementary statics that focused on the methods for testing hypothesis where the two seem to agree on the aspect of changing the research question in null and alternative hypothesis noted as H0 and HA respectively. In both instances, a researcher will be equally interested in whether the predictors result into higher or even lower outcome levels. 

The methods discussed in this article and in the classroom textbook have been established to be suitable for all the measures that are made on the interval or even ration scales including t-tests for comparing 1 sample to the reference group and also, in comparing two paired or 2 independent samples. Based on the book analysis, I have realized these distinct methods in most instances will yield excellent power compared to the nonparametric options, however, are commonly robust to the distribution of the measurements that are being tested, mainly when the sample sizes are large.