Mayur,

You can have contradictory results on correlation studies (one that favors the domain knowledge and the other does not) due to the way the experiment was set up. It is a common to assume underlying factors as not important when they are and the other way around. This builds the foundation of models which are idealizations of complex situations and are simplified to help understanding. While they help in our understanding, they can give contradictory results if one deviates from the initial assumptions and add additional variables to the experiment.

Note: I am also assuming that the experiments were done correctly

Hi

In order to not get any contradiction in interpreting the results of correlation coefficients, the following remarks/principles need to be satisfied:

1. This assumptions of the sample size, both variables should be normally distributed (normally distributed variables have a bell-shaped curve). Other assumptions include linearity and homoscedasticity. Linearity assumes a straight line relationship between each of the two variables and homoscedasticity assumes that data is equally distributed about the regression line have to be satisfied for all measures of correlation coefficients and hypotheses testing.

2. Correlation analysis required a linear relationship between the independent random variables and the dependent random variable.

3. If the sample size of your random variables >30 you can use only the parametric correlation coefficient.

4. If the sample size of your random variables

Depicting contradictory result of correlation on domain understanding may be due to bad quality of data.In such situation right & adequate data collection for analysis is suggested in lieu of thinking otherwise.