Dear Luna Maria Terra

the absence of correlations between independent variables and the dependent variable raises concerns about the validity and interpretability of a multivariable linear regression analysis. It's important to carefully evaluate your data and consider alternative approaches to ensure robust and meaningful results.

If you're unsure about the appropriateness of your analysis or how to proceed, consider consulting with a statistician or data analyst. we can provide expert guidance on the best approach for analyzing your data and interpreting the results.

Dear Luna Maria Terra

consider alternative modeling approaches Depending on the nature of your data and research question, other statistical techniques such as logistic regression, survival analysis, or machine learning algorithms may be more suitable.

If your current dataset does not show correlations between the IVs and DV, consider collecting additional data or including different variables that may have a stronger relationship with the DV. Increasing the sample size or including more relevant variables could help improve the predictive power of your model.

If none of your independent variables (IVs) correlate with the dependent variable (DV), it raises questions about the appropriateness of conducting a multivariable linear regression analysis. In linear regression, the independent variables are expected to have some level of correlation with the dependent variable to explain variation in the DV.

Problem with multivariate analysis has to do with correlation of independent variables among themselves and determining which variables may be causative. No correlation between dependent and independent variables would indicate little explanative value of independent variables. High correlation of independent variables does not necessarily mean causation, but can still be useful for prediction. I am not sure what you mean by no correlation. Are the correlation values all zero? Or are they just low?

first, to be correct with the language, you want to perform a multiple linear analysis, if you were multivariate, you should additionally have more than one dependent variable, is that right?

Check highly influential values, use cook's distance for that.

Then, it is impossible that 15 independent variables do not present collinearity.

this is one of the main drawbacks in the r2 in high MSE.

check for heteroscedasticity, this can also be a problem.