Review and discuss type I and Type II errors associated with hypothesis testing.

Review and discuss the difference between statistical significance and practical significance..

1. Describe the common elements present in all hypothesis tests.

 

Discussion 5

Type I And Type II Errors Associated with Hypothesis Testing

            Type I error (α) comes about when a person rejects a true null hypothesis. It is also referred to as a false-positive. In such situations, people conclude that treatments are different and that one is more effective when that is not the case (Akobeng, 2016). Several issues increase the probability of occurrence of type I error, and most of them are due to conducting multiple statistical tests on given data. To minimize the occurrence of type I error, researchers should avoid doing numerous tests on the same data. Also, using conservative α levels like 0.001 and 0.01 can reduce the occurrence of type I error by 0.1% and 1%, respectively (Oliveira, 2021). However, doing so increases the possibility of type II error occurring. On the other hand, Type II error (β), also known as a false negative, occurs when a researcher concludes that different treatments are not different when they are (Akobeng, 2016). Conventionally, β is set at 20%, meaning researchers allow a 20% probability of inaccurately concluding that groups are not significantly different. Researchers can reduce type II errors by increasing the sample size and using a higher level of significance.

Difference Between Statistical Significance and Practical Significance

            Statistical significance is the probability that population variables are not related, or two groups are not different. Using inferential statistics in the evaluation of data of a sample from a population helps in the calculation of probability (ρ). Αlpha (α), ρ_critical, is set at 0.01 or 0.05 levels and is used to determine if inferential results are significantly different (Rosen & DeMaria, 2012). While statistical significance is an estimated mathematical probability of the occurrence of a sample statistic, practical significance uses a researcher’s judgment from prior research. Also, p -values denote statistical significance while effect sizes represent practical significance. The effect sizes help measure the divergence of sample statistics from the null hypothesis, and they are divided into variance accounted-for, standardized difference, and corrected. Unlike statistical significance, the sample size does not affect practical significance values because they do not rely on the sample size (Rosen & DeMaria, 2012).

 Common Elements in All Hypothesis Tests

            Hypothesis tests have seven common elements. One element is the null hypothesis (H₀) which represents the status quo that people believe. Another element is the Alternative (research) hypothesis (H_a), which contradicts the null hypothesis. It represents what will be accepted when the evidence establishes the truth. A test statistic is also common in hypothesis tests and helps decide if the null hypothesis should be rejected. The rejection region represents the values of the statistic where the null hypothesis is rejected. Assumptions about the sampled population are also usually part of hypothesis tests. Next is experimentation and calculation of the test statistic to determine the numerical value. The last element is the conclusion. If the test statistic value is within the rejection region, researchers reject the null hypothesis, concluding that the alternative hypothesis is true. If it does not fall within the rejection region, the H₀ is accepted, but researchers reserve their judgement on the true hypothesis.