Do you still remember the term hypothesis? The hypothesis is a temporary conclusion from the research results that will conduct. You can also state that the hypothesis is a preliminary answer from the research results based on our presumptions. In this article, I will discuss the Hypothesis Test for Regression and Correlation Analysis on this occasion.
Determining the hypothesis must justify based on theoretical references and empirical study. This justification can be arranged by reading more reference books related to the research topic. Suppose in economic theory; there is a law of demand. When the price increases, the quantity of demand will decrease; vice versa, demand will increase when the price decreases.
Then in formulating the hypothesis, it can be based on empirical study. The empirical study can be obtained from previous related studies that have been tested empirically and have been tested using statistics. Regression and correlation are very popular in the world of research, yes. If you open Google Scholar, you will find many research publications that use regression and correlation analysis.
“Why do many researchers choose regression and correlation as to their analytical tools?” the answer is quite simple, “Because it fits the purpose of our research!” So, regression analysis is used to analyze the effect of one variable on other variables. Then correlation analysis is used to analyze the relationship between variables. Both regression and correlation, some assumptions must be passed. In testing the hypothesis for regression and correlation can used two ways, namely:
Hypothesis Test with P-Value
In testing the hypothesis, it can be seen from the p-value. In formulating a hypothesis, we need to determine an alpha value that we will use. Experimental studies generally use 5% and 1%. In socio-economic, the alpha value limit between 5% and 10% with justification for the research environment is not fully controllable. The meaning of the p-value of alpha 5% is that when the experiment is carried out 100 times and the failure is five times; the research has succeeded. We can also say that the confidence level is greater than 95%.
Suppose there is a study that aims to find out how the effect of price on sales. In this study, alpha was set at 5%. You can formulate research hypotheses in the null hypothesis and alternative hypothesis as follows:
Ho = Price has no significant effect on sales
Ha = price has a significant effect on sales
Next, we can test using simple linear regression. Based on the analysis results, you will get the calculated F value, T count, and p-value (sig.)
Hypothesis testing criteria can follow these rules:
1. p-value (sig.) > 0.05, the null hypothesis is accepted
2. p-value (sig.) ≤ 0.05, the null hypothesis is rejected (accepted alternative hypothesis).
If the regression analysis results, for example, the p-value of the T-test is less than 0.05, the null hypothesis (Ho) is rejected (accepting the alternative hypothesis). Thus, it can be concluded that the price has a significant effect on sales.
Hypothesis Testing by comparing Statistical Tables
Then you can use alternative hypothesis testing by comparing the t value with the t table. Using the p-value criterion alone is sufficient, but it is also important to know the alternative criteria. This criterion is very important if you do manual regression analysis calculations using a calculator.
Using this alternative is the same as the first method. You need to compare the t value with the t table. Suppose we use the same case study example, you can formulate research hypotheses in the null hypothesis and alternative hypothesis as follows:
Ho = Price has no significant effect on sales Ha = price has a significant effect on sales
Ha = price has a significant effect on sales
Next, for hypothesis testing, you can follow these rules:
1. T-value < T table then the null hypothesis is accepted
2. T-value ≥ T table means the null hypothesis is rejected (accepting alternative hypothesis).
If the T value is greater than T-table, the null hypothesis (Ho) is rejected, or the alternative hypothesis (Ha) is accepted. In conclusion, the price has a significant effect on sales. Well, I hope this article is useful for you. See you in the next article. Thank you, bye.
Pingback: Differences Between the Null Hypothesis and the Alternative Hypothesis in Statistical Analysis - KANDA DATA