In: Statistics and Probability
1. What are the two key pieces of information you are expected to get out of running a statistical test?
2. What are the 3 potential outcomes of a correlation analysis?
1. Statistical test : statistical tests are conducted to check the significance of a pre-defined null hypothesis for a given level of significance. in order to test the null hypothesis key information is :
- The value of the attribute or the confidence interval with some level of precision, where the true value lies. For example, in testing the value of mean of a data ( say mu_0 ) one will get the confidence interval in which the actual value of mean lie for a pre-defined alpha or the probability of type 1 error.
- The significance level of the attribute. Sometimes, researcher do not wants to know the confidence interval for alpha level of significance but wants to know the accuracy of the decision made. That is also obtained in statistical tests known as the p value of the parameter.
2. Correlation analysis is the analysis of significance of the correlation value in order to obtain the best possible clear picture of the data using correlation. the range of correlation coefficient is from -1 to +1. It has three possible outcomes outcomes
- Positive correlation : the two variable under study are positively correlated if they move in the same direction. That means if X increases Y also increases or if X decreases Y also decreases.
- Negative correlation : the two variable under study are negatively correlated if they move in the opposite direction. That means if X increases Y decreases or if X decreases Y increases.
- No correlation : the two variable under study are completely uncorrelated, there is no pattern visible in the directions of two variables. Although the correlated value is not exactly zero. It is very close to zero, the hypothesis of correlation equals to zero is accepted after testing of hypothesis.