Data Collection Methods

Observation

Observation is one of the easiest and least invasive data collection techniques available to the analyst. During the observation, the analyst should do his or her's best to avoid direct involvement in the process or workflow being observed. If interaction with the study subject is necessary, an analyst may wish to supplement the observation with interviewing, which will be discussed later. Observation works best when the process or workflow is performed in an unadulterated state, meaning the what is observed matches exactly what happens during regular, unobserved operations. Only then can the analyst really evaluate the process or workflow honestly and without bias.

Observations can also be either direct or indirect. Direct observation is when you watch interactions, processes, or behaviors as they occur; for example, observing a teacher teaching a lesson from a written curriculum to determine whether they are delivering it with fidelity. Indirect observations are when you watch the results of interactions, processes, or behaviors; for example, measuring the amount of plate waste left by students in a school cafeteria to determine whether a new food is acceptable to them.

(2008) Data Collection Methods for Program Evaluation: Observation, Department of Health and Human Services, No. 16, Page 1

It's fair to mention pure observation is difficult to achieve, more likely the analyst will be involved in the process or workflow and the technique is more work study than observation. Under these circumstances, the analyst can still get a mostly fair and unbiased view of the process or work. However, the contribution or hinderance of the analyst should be noted and accounted for in the data.