Delayed Measurement

Reasons for Delay in Measurement

In research experiments, including those dealing with communications, one consideration is whether one expects a dynamic result as an outcome of an intervention. Consider the issue of teaching a simple mathematical skill like addition: once a person has mastered the skill, it is expected that the skill is retained over time. Similarly, learning that George Washington was the first president of the United States is information that does not change over time: Once the information becomes a part of a person’s long-term information storage, it is not subject to additional [Page 369]influence. Comparing information acquisition to a behavioral routine like brushing teeth or refraining from smoking involves a different long-term dynamic. The behavioral actions are subject to both formation and change, and are capable of becoming replaced or modified.

Measurement delay occurs in some experiments in order to permit a comparison and evaluation of a trend over time. The reason for delay may be to assess decay in the impact of an intervention, since a person may forget or revert to a prior behavior that he or she sought to extinguish. For example, if a set of public health messages seeks to change diet and weight gain scores, the person’s immediate post-test may show a change that declines over time as the impact of the intervention diminishes and has a less direct effect on the behavior of a person. In this case, the impact of the intervention is the maximum value at the immediate post-test and continues to decline over time. A delay in the examination of the change compared to the person’s pre-test would demonstrate less overall long-term change in the case of delayed measurement. Essentially, a person practices the new diet but over time may slowly go back to the old dietary eating patterns and lose all the advantage gained by the original shift. As the measurements are delayed, the impact of the delay indicates a decline in the size of the observed impact.

The impact of a decline contrasts with a curve that reflects a skill that improves with practice over time. Instead of a decline, over time as the skill is practiced, performance or adherence to the desired outcome improves. Under this condition, the least amount of change or impact occurs at the time of the immediate post-test. One example of this pattern can be found in efforts to train a person to improve his or her public speaking skills. If a class or other intervention aimed at improving the quality of a person’s public presentation is subject to measurement, there is a high chance that the speaking quality measured at the end of the course would show less improvement than a delayed measurement taken a year later. Theoretically, a person practicing the improvement advice should show more improvement over time, so later measurements would demonstrate larger effects when compared to the pre-test.

...