Deleting brvaia A common stratepy for handling a case in which an obser vation i

Question

Deleting brvaia A common stratepy for handling a case in which an obser vation is missing data for one or more variables is to fill those missing variables with Os and add a variable to the model that takes the value 1 for that one observation and 0 for all other observations. Show that this "strategy" is equivalent to discard- ing the observation as regards the computation of b but i do have an cffect on R, Consider the special case in which X contains onlv a constant and one variable Show that replacing missing values of x with the mean of the complete observations has the same effect as adding the new variable.

Explanation / Answer

Using an OUTPUT, REPLACE, or REMOVE statement overrides the default write action at the end of a DATA step. (OUTPUT is the default action; REPLACE becomes the default action when a MODIFY statement is used.) If you use any of these statements in a DATA step, you must explicitly program all output for new observations. The OUTPUT, REPLACE, and REMOVE statements are independent of each other. More than one statement can apply to the same observation, as long as the sequence is logical. If both an OUTPUT and a REPLACE or REMOVE statement execute on a given observation, perform the OUTPUT action last to keep the position of the observation pointer correct. Because the REMOVE statement can perform a physical or a logical deletion, REMOVE is available with the MODIFY statement for all SAS data set engines. Both the DELETE and subsetting IF statements perform only physical deletions; therefore, they are not available with the MODIFY statement for certain engines.

Related Questions

Navigate

• Browse (All)
• Browse D
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.