Validity and reliability in observation

It is important that observation is as objective as possible and observers need to be aware of their own interests and biases so that the information gained is as valid and reliable as possible and not selected to prove a point.

• Observers usually keep on-the-spot notes to complement an observation schedule so that any additional or unusual events can be noted. For example ‘X entered the room and spoke to tutor, this interrupted class for 4 minutes’; ‘Firebell rang at 3.34pm so the class was shorter today than usual’.
• ‘Field notes’ are usually written after an observation and the advice is to do this at the earliest possible moment so that memories of what happened are as fresh as possible. Robson suggests that you should ‘never embark on a second observation session until you are sure you have sorted out your notes for the first one’ (p324)
• Practice in using a schedule before the ‘real’ observation will help the observer to iron out uncertainties in how to use it or how to log unexpected or ambiguous information.
• Reliability of observation will come from observer consistency. Observers must ensure that they make similar decisions about similar events on different occasions. They must also make the same decisions about the same events if they see or hear them again, say on video or audio tape.
• Ideally, more than one observer should be involved in looking at the same events, at least in initial practice sessions, so that there can be agreement on what is going on and how it is to be coded. Reports of observational research should always include information on how reliability was tested. Robson (p340) refers to ways in which ‘inter-observer agreement’ can be calculated.