- Evaluate whether each of the following statements are true or
false.
- It is appropriate to compute the correlation coefficient when
one of the variables is ordinal as long as at least one variable is
in interval. (True/False)
False (Reason is mentioned in the last
question)
- R (correlation coefficient) between X and Y represents the
percentage of variability of variable X explained by variability of
variable Y (True/False)
false: this is called the R2 or coefficient
of determination
- When correlation coefficient is negative, standard error of
prediction is also negative (True/False)
false: standard error is mainly depending on the
standard deviation and n, and both are positive, and correlation
sign depends on the covariance.
- When correlation coefficient is positive, beta (B) value in
regression equation is always positive (True/False)
true; both variables sign depends on the
covariance.
- b) Explain the reason why you chose the answer for i in
question 17 above (whether it is appropriate to compute the
correlation coefficient when one of the variables is ordinal).
reason: correlation data is the paired data. if the one
variable is ordinal and its corresponding variable must be a single
value, it should not be interval.
Thnaks