The timing of history effects

Despite being called history effects, the events that happen in the environment that change the conditions of a study do not only occur in the past. These events can occur prior to the study taking place and during the study. Each is discussed in turn:

The length of a study

In principle, the longer a study takes place, the more likely that history effects may become a threat to internal validity. This is based on the assumption that the longer a study last, the more likely an unpredictable event may take place that threatens the internal validity of your study.

The magnitude of history effects

When we talk about the magnitude of history effects, we are interested in how critical an event is; that is, how likely an event is to change the outcome of your study.

In Study #1, mentioned above, the fire alarm was a threat to internal validity because it affected the sleep patterns, and arguably, the quality of sleep that students received before taking part the next day in an experiment on sleep quality. In other words, the event (i.e., the fire alarm in the middle of the night) was critical / very likely to change the outcome of the study (i.e., change the outcome of the pre-test, where students levels of tiredness were recorded). The same event (i.e., the fire alarm) could be considered critical in other experiments that were not about sleep; for example, studies where the pre-test involved (a) students taking an exam, (b) performing memory tests, (c) taking part in a sports competition, or (d) some other pre-test that could be influenced by a lack of quality sleep. However, if the study was unlikely to be affected by sleep quality (e.g., studies on (a) attitudes towards African American prejudice, (b) factors that make Apple a popular brand, etc.), then it is probably unnecessary to mention such an event (i.e., the fire alarm) as a possible history effect; that is, as a possible threat to internal validity. In other words, the event was not critical / not likely to change the outcome of your study.

Events that lead to history effects are often unpredictable, which makes them very difficult to plan for. When an event does take place, it can still be difficult to reduce the threat to internal validity that such an event creates. The important point is to understand the impact of the history effect and its potential magnitude, and explain what effect it may have had on your results.

1 2 3 4 5 6 7