Here is a simple thought exercise. Imagine measuring a group of 100 people on any construct that might interest you, using one of the many measurement tools communication researchers have at their disposal for doing so. You might ask these 100 people to evaluate a political figure on a feeling thermometer, or request them to indicate on a 1 to 10 scale how likely they would be to buy a product after showing them an advertisement for the product, or have them respond to the Communication Anxiety Inventory, or provide their attitudes toward birth control by responding to a series of questions that you can use to generate a quantification of their attitude. At the end of this process, you’d have a measurement of each of these 100 participants-a set of observed scores or observed measurements quantifying each person on the construct you have measured. Now imagine repeating this a week later using these same 100 people using exactly the same procedure, resulting in another set of 100 observed scores quantifying each person on the same construct you had previously measured them on. With measurements of each person at two different times, you could determine how many of the participants responded consistently, providing you with the same observed score at both occasions. About how many of the 100 people do you think would provide a consistent response?