On Incomplete Data
|•||Avoid Incomplete Data|
|•||Address Missing Data|
A study is conducted on a random sample of 14 individuals. Scores are obtained on one dependent and two independent variables. These scores are recorded in a matrix where rows represent individuals and columns represent variables, as shown in Exhibit 16.1.
The researcher is eager to perform statistical analyses. But, wait! Not all cells in the matrix have values.
Entire rows are without values. Individuals 8 and 9 did not respond to the study—they are nonrespondents. They may be individuals who fail to return questionnaires. Or, they may be individuals chosen for an experiment who fail to participate.
Exhibit 16.1 also shows that Individuals 3, 5, 7, and 14 are missing data for Y or X1. They participated, but incompletely. Survey participants may fill out only a portion of a questionnaire; experimental participants may participate for a time, but drop out before the study is complete.