Internal Validity

History

History, or events outside of a research study, can influence subjects’ responses and can act as unexpected and unplanned independent variables. For example, a campaign to reduce cigarette smoking among teenagers may coincide with a decrease in cigarette purchases by teens, but an increased tax on cigarettes during the same time period may also be associated with a reduction in cigarette purchases. The “event” of increasing taxes on cigarettes may be an alternative explanation for the reduction in cigarette purchases.

A researcher may attempt to control for history threats by making all participants’ experiences identical except for the independent variables, but this is often difficult or impossible to do. Other ways to reduce the threat to internal validity due to history is to engage in random assignment and to include a control group in the research design. This ideally results in a research study in which the effects of history will be similar across groups.

Longitudinal studies or studies involving repeated measures on subjects over time are more susceptible to threats to internal validity due to history than studies in which data are collected at a single point in time.

Maturation

Maturation is another common threat to internal validity in longitudinal studies. Maturation refers to the natural changes, psychological or physiological, that occur as participants age or time passes. Studies that involve children are often most vulnerable to maturation effects. For example, perhaps you have implemented a behavioral therapy among a group of 3-year-olds. When one measures their behavior at age 4 years, these children demonstrate remarkable improvement. The question a researcher must ask is whether the program led to improvements in behavior, or whether the children are more well-behaved because they have naturally become more mature. Maturation threats to internal validity can also be [Page 771]related to physiological changes. For example, participants may become hungrier, thirstier, or more tired as a study drags on, and hunger, thirst, or fatigue (but not your independent variable) may lead participants to respond differently on measures of the dependent variable.

...