Hypothesis Testing in the Applied Decision Making uk
Hypothesis testing is the utilization of gathered statistics to establish the probability that the said hypothesis is indeed true (Good, 2000). It may also be referred to as confirmatory data analysis. A null hypothesis test is usually used to come up with decisions. The common baseline is if the null hypothesis is deemed to be true, what is the probability of establishing a value that is more or less close to the one initially seen? Further, hypothesis testing is used to determine if results of experiments have substantial data suggesting trends contrary to the normal culture or routine.
Hypothesis testing entails the following distinct steps, namely formulation of a null and alternative hypothesis, idetifying the test size, computation of the test statistic and comparison between the p-value, decision making and finally conclusion about the data (Hoel, Port and Stone, 1971).
The first step in hypothesis testing is the formulation of the null hypothesis and alternative hypothesis. So far, the truth is unknown and a conclusive result may only be achieved if a good null hypothesis has been stated. Here, accuracy is important. An accurate null hypothesis will ensure clarity in the rest of the subsequent processes. If the null hypothesis is established to be untrue, this is where the alternative hypothesis comes in (Good, 2000).
After the null hypothesis, the next step is to determine the test sizze. This is basically a quantification process of the whole paradigm. The researcher identifies the rejection area, i.e. the possibilities that fall beyond the limits of the test, and then narrows down on all the probable outcomes.
Thirdly, there is computation of test statistic and probability value, also known as p. Here, the researcher constructs the confidence intervals. In decision-making, we either accept or reject the null hypothesis (Hoel, Port and Stone, 1971). Finally a conclusion is made regarding the data and results from data are interpreted.
Hypothesis testing therefore is a key element of statistical analysis and it aids in generation of informed decisions and conclusions.