Coding in The Observer XT
03/10/2017

The coding scheme or ethogram is a key step in measuring observed behaviours and will determine what you can do with your data later.  Often it can be tempting to log everything, however a more targeted approach towards answering your research question(s) can mean a smaller amount of behaviours to score giving a more reliable outcome and making data collection much quicker.

Behavioural Groups

• Each group of behaviours should be clearly defined, so that anyone could code the observation and get the same result.
• Behaviours within groups should be mutually exclusive and exhaustive, meaning that 2 behaviours within a group cannot occur at the same time, and all the behaviours within that group are included.  If there are a numerous possible behaviours within a group but you are only interested in 2, then label the rest as ‘other’ or equivalent.  For example if you are interested in how much time students spend looking at their phone in a study session, you may have a ‘Looking’ behaviour group that includes: ‘phone’, ‘lecturer’, ‘board’ and ‘other’.  Despite what the student says you cannot be looking at the phone and lecturer!

The Observer Process

• Start with a template or draft your own coding scheme. 
• Add subjects – people or animals that you intend to score.  Note that this only applies if you are scoring interactions between individuals during an observation – e.g. a parent and child interacting.  Do not confuse this with independent variables such as the name of a subject, their age or gender etc. The Observer allows you to include independent variables and use these to select sets of data.
• Add behaviour groups and behaviours.  Each group should be a vital part of answering your research question(s).
• Add modifiers – these add more detail to specific behaviours.  For example, in an infant study, the behaviour shaking a rattle may be in a behaviour group named ‘interaction with toys’.  A modifier might include the intensity of the shake, such as ‘vigorously’ or ‘delicate’.  Modifiers can be nominal or numerical. 
• Define behaviours as ‘state’ or ‘point events’.  Some of your behaviours will have a duration, such as talking/not talking and will be defined as a state.  Others will not, such as a hand clap or blink and these are defined as point events.  Most people find that they need to know the duration and frequency of behaviours and score states most of the time.  This is vital for analysis since it not possible to use point events to find out how much time other state behaviours are also co-occurring. 
• Pilot your scheme – at this point you may spot behaviours that aren’t covered by your coding scheme or want to simplify others.  Make sure you are able to create data profiles for each of your questions and look at the results.  This is the single most ignored part of learning The Observer!
• Scoring can be done live, from video or from both.  If your coding scheme is not too complex, then live scoring is practical; you are also able to go back and check scoring on video later and make any appropriate modifications.
• Up to 4 videos can be used in Observer for scoring behaviours from different angles, particularly useful if there is are many individuals/animals to score.
• One of The Observer’s greatest strengths is that you can import and analyse time-stamped data from many other sources such as eyetrackers, physiological measurement equipment, FaceReader etc. If you think your project would benefit from these other measures we can help!

With a well-defined coding scheme, The Observer XT allows you to score your behaviours easily, enabling you to answer your research questions.  We are here to help you every step of the way – and it’s a free service so please do make use of it!