What Are Hypotheses?
Statistics allows us to use mathematical models to test the significance of judgments. In statistics, we call such judgments hypotheses because they always require testing and can be accepted or rejected. For example, a hypothesis might be formulated: “People who are tall earn more money” or “there is a relationship between excellent academic performance and time spent studying.” Both of these hypotheses are not unproven truths, so we must use statistical methods to test them.
How Do We Test Them?
Strictly speaking, we always formulate two hypotheses in statistics: the null and the alternative hypotheses. The alternative hypothesis builds on and essentially replicates the research question, while the null hypothesis is the opposite of the alternative hypothesis (IET, 2021). There are special rules that force us to create two hypotheses that correspond to each other. For example, if the alternative hypothesis postulates the superiority of one value over another (>), then the null hypothesis must cover the rest of the value space, that is ≤. We use parametric or nonparametric tests to handle the data set. For example, to compare the equality (or inequality) of the mean values of two groups, we use the t-test; if there are more than two groups, we use ANOVA (Dallanoce, 2021). Regardless of the particular type of test, the output is a criterion that allows us to judge the statistical validity of previously obtained results. If this criterion, usually called the p-value, is below the critical threshold, then the null hypothesis is rejected, and the alternative hypothesis is accepted as proven. Otherwise, the null hypothesis must be accepted. In other words, the main criterion we operate in hypothesis testing is to study the value of the test parameter, the p-value. We must strive to find it during data processing and then compare it to a critical threshold value. Such a value determines the level that divides the significance of the test into two sides.
References
Dallanoce, F. (2021). ANOVA, T-test and other statistical tests with Python. TDS. Web.
IET. (2021). Alternative hypothesis: Definition and when to use it. Indeed. Web.