Learning Goal: I'm working on a r exercise and need support to help me learn.If we want to avoid corruption of the correlation calculation due to missing data, what is the best additional argument to supply into the Cor function (i.e which approach incurs the least amount of data loss?)Use=”p”
Na.rm =TRUE
Method=”spearman”
Correct answer not given
Many textbooks will express the assumption of linearity in modeling enterprise as “Must be linear”. Which is silly, since we are often trying to test for linearity, and therefore could not credibly assert it. What is the best way to phrase this assumption?Must be linear when sampled in triplicate
Must not be heteroscedastic
Must not be systematically non-linear
Must not have an amorphous scatter
Consider the following circumstance in a study of two groups: Group A has 100 participants; 90 of them scored 45 points on a test, and 10% scored 95 points (Average score is thus 50%). In Group B (also 100 participants), all participants scored 48%. What is the circumstance that we want to avoid when drawing conclusions about the groups in the study? Benford’s Law
Berkson’s Bias
Collinearity
Ecological Fallacy
What is true about the p-value associated with the correlation coefficient?Correlation’s p-value is viewed as less meaningful than the correlation value itself
It is technically impossible to obtain p<0.05 for the correlation coefficient
You can obtain the p-value via the additional argument use=”p”
There is no p-value associated with the correlation coefficient
I will tell you the data scatter of Height-versus Weight among Yale student athletes in heteroscedastic. What does this mean for our ability to perform correlation?R’s cor Function will yield a correlation value, but the value could be misleading
R will produce a warning about the scedacity but will yield an r-value anyway
R will yield an error, acknowledging the scedacity and will not yield an R value
R’s cor will yield NA, for the User to realize that this means heteroscedasticity
Requirements: 1-2 sentences