Hypothesis testing and its application in Business. We often need to make a decision based on outcome data from the process. That means we need to make a “yes” or “no” decision on the process results. In order to satisfy the need, we must perform a hypothesis test on the results. Based on a given confidence level, can we draw a conclusion on the process results? Some examples of hypothesis tests can be a marketing manager’s proposal on cutting the contract price in order to attract more customers to sign up for the contract services.
The marketing manager has a gut feeling that if they can cut a 5% contract service price, they will get a significant amount of customers to sign up for the service contract, such as 10%. To prove the marketing manager’s gut feeling, we can use hypothesis tests by designing an experiment, taking the data and reaching for a definite conclusion. This is a scientific approach on statistical thinking.
What is the relationship between confidence interval and hypothesis tests? Obviously, they are both apart of statistical inference tools and are closely related tools. Confidence intervals provide a range for the estimated value so that you can decide if you want to reject the hypothesis or not. We noted that when we cannot reject the hypothesis, it does not mean that we have accepted the hypothesis; it simply means that we do not have sufficient evidence to reject the hypothesis. We may have to collect further evidence (more data) to verify the hypothesis again. We also note that the confidence interval is closely related to the confidence level. Increasing the confidence level also increases the confidence interval.
Step 1: State the Hypothesis. State the null and alternative hypothesis. Based on the convention, the null hypothesis always contains an equal sign. For example, taking cold medicine and taking a placebo will get you better after catching a cold in the same amount of time. That is the null hypothesis. The alternative hypothesis becomes: Taking medicine will allow you to recover faster.