In a research context, the term construct refers to the characteristic or trait that is being evaluated. Constructs are classified as direct or indirect. Some examples of constructs are intelligence levels, measures of satisfaction, weight and height.
Indirect constructs are those that can not be directly measured. Instead, they are calculated by taking into account reported indicators or behaviors. For example, happiness may be measured by having individuals report their reactions to certain instances. The reactions can then be evaluated to provide a happiness level on a predetermined scale.
A direct construct is one that is easily measured. Height, weight and blood pressure are calculated using verifiable methods and stated units of measurement.