question about correlation and regression

Hi, I have not posted here before but could use some help!

I am analysing survey data with nominal data (demographic information such as gender and race) and ordinal data (responses to 5 point likert scales)
to test hypothesis of scale responses varying by demographic data, I feel I should do spearman correlation? is this correct?

I also plan to do multiple linear regressions, to test hypothesis of demographic characteristics predicting the measures scores , but have been advised to do scatterplots - which will not work with the ordinal data? any suggestions would be very appreciated thank you