Degree of Structure

Observational methods vary greatly in terms of the amount of structure provided for the observer.  They range from much unstructured observations, which attempt to provide as complete and nonselective a description as possible, to very structured methods, which provide a complete list of expected behaviors and require only that the observer check which ones occurred.

It should be realized, however, that complete recording of an event is virtually impossible.  Even with videotaping, exact replication will not be obtained because of biases.  This fact should be recognized if you use unstructured observation.  As soon as you begin your observations they will be given structure by your editing process.  Even so, a rich depth of material can be gathered from unstructured observation a depth that will never result from the use of structured methods can be used to great advantage in transportation research.

Structured observation can take one of several forms, but perhaps the most common is the checklist.  A checklist allows the researcher to record whether or not a given behavior occurs.  The desired behaviors must be explicitly defined so that there is no question in the mind of the observer as to whether or not they occur. 

Structured observation, when appropriate, is an excellent method of collecting data.  Many more subjects can be observed, in less time, than with unstructured observation, and the data analysis is much simpler. Taking results from a checklist merely involves counting how many times a particular behavior occurred.  The results of unstructured observation, on the other hand, consist of quantities of descriptive data, since the observer was trying to record everything that happened.  These data must be sorted out to see if there are any patterns to the observed behavior-a very time consuming process.

The level of the study will affect the type of observational technique you choose to measure your variables.  Unstructured observation is a method for use in Level I studies where flexible exploration is needed.  This type of observation is not appropriate for Level II and Level III studies where precision in measurement of the variables is required.  The concept that each level of research builds on the preceding level has been emphasized.  To carry this idea through to choosing data collection methods, you look at the ability to provide structure in measurement as dependent on knowledge about the topic and/or variables.  It is not possible to develop a structured tool to observe behavior unless you can base the tool on previous descriptions of behavior under the same circumstances.  Therefore, developing a tool for a Level II study requires the results of a Level I study on the same topic.

In the same light, the tool used in a Level III study must be precise enough to differentiate between experimental and control groups.  The tool, therefore, must be highly structured and based on considerable knowledge of the variable.  Remember the degree of structure increases with the level of the study and the extent of research and theory available to explain the action of the variables.