| As almost any researcher can attest, missing data are | | | | completed because they have moved out of the area, |
| a widespread problem. Data from surveys, | | | | died, no longer see personal benefit to participating, or |
| experiments, and secondary sources are often missing | | | | do not like the effects of the treatment. Surveys |
| some data. The impact of the missing data on the | | | | suffer missing data when participants refuse, or do not |
| results of statistical analysis depends on the | | | | know the answer to or accidentally skip an item. Some |
| mechanism which caused the data to be missing and | | | | survey researchers even design the study so that |
| the way in which the data analyst deals with it. This is | | | | some questions are asked of only a subset of |
| the first in a series of three articles that discusses | | | | participants. Experimental studies have missing data |
| issues surrounding missing data. This article outlines the | | | | when a researcher is simply unable to collect an |
| mechanisms of missing data and some of their | | | | observation. Bad weather conditions may render |
| impacts. Subsequent articles will explain common but | | | | observation impossible in field experiments. A |
| problematic solutions to missing data, new and better | | | | researcher becomes sick or equipment fails. Data may |
| solutions, and the software available for implementing | | | | be missing in any type of study due to accidental or |
| these solutions. | | | | data entry error. A researcher drops a tray of test |
| Data are missing for many reasons. Subjects in | | | | tubes. A data file becomes corrupt. Most researchers |
| longitudinal studies often drop out before the study is | | | | are very familiar with one (or more) of these situations. |