When I try to create a scatterplot with my continuous variables, I get a grid of equidistant dots instead of scattered dots. I'm not sure why this happens.
Without seeing it, I can't tell you why that happens/is happening. The only thing I can think of off the top of my head is that you might be plotting integers and the plotting engine in R is choosing the scale based on that? Feel free to email me or catch me on my discord and post some screenshots.
What do I do if my variables dont show a linearity, am I still allowed to do pearson or spearman for the correlationmatrix since linearity is a condition for the computation of the correlation?
If you have a clear nonlinear/curvilinear pattern in your data, you cannot use linear regression to analyze the data/determine patterns. You'll need to use polynomials and other methods to determine the actual shape.
@@AlexanderSwan I dont understand what you mean, can you explain it to me again please? 🙏🏼 am I allowed to still do a spearman or pearson correlation analysis? I‘m NOT doing multiple regression, just the normal correlations via correlation matrix
When I try to create a scatterplot with my continuous variables, I get a grid of equidistant dots instead of scattered dots. I'm not sure why this happens.
Without seeing it, I can't tell you why that happens/is happening. The only thing I can think of off the top of my head is that you might be plotting integers and the plotting engine in R is choosing the scale based on that?
Feel free to email me or catch me on my discord and post some screenshots.
@@AlexanderSwan I think you're right. All of my numbers are whole integers on a relatively small scale no matter which variables I'm using.
What do I do if my variables dont show a linearity, am I still allowed to do pearson or spearman for the correlationmatrix since linearity is a condition for the computation of the correlation?
If you have a clear nonlinear/curvilinear pattern in your data, you cannot use linear regression to analyze the data/determine patterns. You'll need to use polynomials and other methods to determine the actual shape.
@@AlexanderSwan I dont understand what you mean, can you explain it to me again please? 🙏🏼 am I allowed to still do a spearman or pearson correlation analysis? I‘m NOT doing multiple regression, just the normal correlations via correlation matrix
@@AlexanderSwan is the pearson correlation coefficient a synonym for a linear regression?