Observation is one of the most basic research methods in economics. It entails collecting data through scientific observations of objects by humans or machines. For example, you can track how people use a certain commercial product, such as a detergent. An economist will then be able to recommend that business managers include a list of possible uses of this detergent on the label.
An experiment is used when observation is not possible due to the scale of the task. For example, it may be physically difficult to observe how every household uses a detergent. Instead, this detergent can be given to volunteers who will then record how they use this product. An experiment, is essentially, a test that allows researchers to gather data from a sample of the population. Assuming that the sample reflects the structure of the population, researchers can reasonably believe that the outcome of the experiment will be the same for the whole of this population.
According to economist Klaus G. Troitzsch, measurement is the attempt to map properties of real-world entities. Measurement research often includes surveys, conducted with interviews or questionnaires. Measurement as a research method in economics is used when the data is easily quantifiable -- for example, GDP or corporate sales figures.
Humans can react differently when they are observed. For example, households may use the detergent in an experiment in significantly different ways than they would had they not been watched by researchers. This poses a dilemma that researchers resolve in different ways. Accordingly, researchers need to do everything possible to minimize the effect of reactivity. In the case of the detergent, researchers could tell the households that they would be testing the amount of detergent households utilize in a week rather than the ways in which they use the product. Thus the subjects would not focus on the variable being measured and would demonstrate typical behavior in terms of how they use the detergent.