Structured Observation

Structured observation entails the collection of data according to a set of predefined rules and procedures. The structure of the observation and the predefined variables and their values are derived from the purpose of the study. Structured observation is a type of nonparticipant observation in that it involves direct observation of a setting without interaction with participants. More quantitative in nature, it is also referred to as systematic observation.

Structured observation uses observation schedules or checklists in which data are recorded according to predefined criteria as values of variables that have been explicitly defined to ensure consistency in data collection. It emphasizes factual measures (e.g., whether or not a particular behavior has occurred) over those requiring judgment or interpretation (e.g., application of a scale related to the intensity of a particular behavior). As observations are collected in real time, they have a temporal dimension and yield information about the duration, frequency, and sequence of events. Tests of interobserver agreement are conducted to ensure the reliability of data. Structured observation typically involves a strong investment in the development and the pretesting of variables as well as the training and monitoring of observers.

Because of the rigorous quantitative nature of the variables and data collection processes, structured observation is regarded as having the potential to yield results with high validity, making replication and generalization possible. If the same observation schedules and coding schemes are used, it enables comparison across two or mores sites.

Structured observation allows for the collection of data (e.g., frequency of particular types of behavior, actions of very young children) that are not readily captured by other methods such as surveys or interviews. Structured observation is appropriate for data collection in both laboratory and natural settings. When combined with interviews, structured observation allows researchers to compare what people say with what they actually do. It is very useful for understanding processes and frequencies of behaviors within a given time period.

Structured observation is not appropriate for a research topic about which little is known, as it requires prior knowledge of the setting for the development of observation schedules. Unlike more flexible [Page 839]forms of observation such as participant or unstructured observation, structured observation, with its predefined variables and processes, is not effective in capturing behaviors or other features of a research context that are unexpected. Reactivity can be a major problem and should be addressed through purposeful habituation and monitoring. Structured observation is difficult to implement in busy, complex settings where it may be hard for observers to focus on particular behaviors. This challenge can be overcome somewhat through the use of multiple observers, who may increase reactivity, and the strategy of focusing on one participant at a time, which may result in the loss of other data. Structured observation requires highly skilled, well-trained observers who are familiar with and can understand what they are observing.