Call Us: US - +1 845 478 5244 | UK - +44 20 7193 7850 | AUS - +61 2 8005 4826

Hypothesis Testing Statistics

Hypothesis Testing Statistics

Hypothesis testing statistics is a statistical method which is applied when making statistical decisions using experimental data. It applies the sample data sampled from the population, to make conclusions regarding the population parameter or distribution. The hypothesis testing involves using statistics to determine the probability that a particular hypothesis to take place. Before analysing the process of hypothesis testing, lets look at the various terms used.

Terms and concepts used in Hypothesis Testing Statistics

Null hypothesis: this is a statistical hypothesis indicating that the observation could occur as a result chance factor.

Alternative hypothesis: it is the opposite of null hypothesis, which indicates that the observation comes from the real effect.

Level of significance: this is the degree of significance within which the null-hypothesis is rejected or accepted. It is not possible to have 100% accuracy; hence, we choose a level of significance.

Type I error:  this error occurs when the null hypothesis is rejected when it was true. It is denoted by symbol alpha.

Type II error: this error occurs when the null hypothesis is accepted when it is false. It is denoted by the symbol beta.

One tailed test: this applies when the statistical hypothesis assumes one value such as H0: μ1 = μ2

Two tailed test: this occurs when the given statistical hypothesis assumes less than or greater than value.

The process of hypothesis testing statistics

Step 1: formulation of the null hypothesis H0 (indicating that the results is from a pure chance) and alternative hypothesis Ha (indicating that the results is due to a real effect)

Step 2: specification of the test statistic that will be applied in the assessment of whether null hypothesis is true or not.

Step 3: computation of the p-value. This tests the probability that the test statistics would be significant or not, with an assumption that the null hypothesis is true.The smaller the p-value, the strong the evidence against the hull hypothesis.

Step 4: comparison of the p-value to the significance value of alpha . If the , this implies a statistically significant effect. In this case, the hull hypothesis is rejected and the alternative hypothesis is accepted; and vice versa is true. In this case, the following rule is used in hypothesis testing statistics.

  • If p-value > alpha; fail to reject the null hypothesis and conclude that the results are not significant
  • If p-value > alpha; we reject the null hypothesis and conclude that the results are significant