Step 1: Set the scope and choose an approach

There are many aspects to consider when designing observations, including defining a focus, the role of the observer and how structured the observation needs to be.

Define a focus

Start by defining a clear boundary around what is being observed, to keep things consistent and contained. Examples include:

  • unstructured play in the before-school playground
  • study habits in the library during recess and lunch
  • teacher interactions during staff meetings.

Decide on the role of the observer

The role the observer plays in the data collection process can vary from being a total observer (outside the situation) to a total participant (embedded in the situation) or somewhere in-between.

  • An example of a total observer is a teacher observing playground interactions between students without intervening. During this observation, the teacher might count how many times certain students are included in play after a program was put in place to address inclusion.
  • An example of a total participant is a teacher recording their own teaching and then watching the lesson later and assessing the implications of what was recorded and the impact on student learning.

Decide how structured the observation needs to be

As with other forms of data collection, there are several different approaches that can be taken in collecting observational data. It is important to make an informed decision when choosing an approach for conducting observations, as each has its strengths and weaknesses. The decision to use one over another will depend on the evaluation questions and the resources available.

StrengthsWeaknesses
Unstructured observation
  • provides maximum flexibility for recording details of actions and behaviours observed.
  • requires the observer to take extensive field notes
Semi-structured observation
  • provides guidance for behaviours to observe
  • allows observer to make notes against the expected behaviours (and unexpected behaviours)
  • can result in reduced focus and increased data analysis time when the observation is too open ended
Structured observation
  • usually uses a pre-tested, validated instrument
  • more efficient - check off actions rather than extensive notes
  • enables data to be compared with published norms and trends from other settings in which the tool has been used
  • requires training in the use of the instrument
  • requires detail of evidence to support assessment
  • limits the scope for the observer to deviate from the predetermined behaviours
Return to top of page