Skip to main content Accessibility Help
Open menu

Observation is an evaluation methodology where you record people’s behaviours. For more about using observation in your evaluation see How to conduct observations

When you analyse observational data it is important to remember the limitations of this approach.

  • People change their behaviours when they realise they are being observed
  • Unless you were able to interview the people being observed during or afterwards, you can only infer the reasons why they behaved in the way they did

Despite these limitations observation can provide immensely valuable data – illustrating what people actually do, rather than what the remember, or are willing to admit to doing.

A five-step guide to analysing observational data

Before you start – define the types of behaviour you will be analysing

Before you start your observational study, identify the behaviours you will be looking to record. Your observation sheet should be designed so that it is quick and easy for the observer to record the presence or absence of each behaviour and, where appropriate, its durations and frequency of occurrence. And it should provide space for the observer to record unanticipated behaviours that may be relevant to the evaluation objectives.  

Your observation sheet should also record some basic demographic information about the people being observed e.g. age category they fall into and the type of group they are in - family; school; adult; alone. This information will be judge ‘by eye’, not by asking the people being observed. NB. we strongly recommend not trying to assign people to a gender or ethnicity category in this manner.  

  

After you have collected the data – number the observation sheets

Assign a number to each observation sheet. That way if there are any queries, you can go back and check the original data. This also means you can – if you wish – assess whether certain types of behaviour tend to coincide e.g. subjects 1, 8, 17, 33 and 45 were all observed to do A and B but not D.  

Create a spreadsheet to collate the data

Use a programme such as Excel to create a spreadsheet where you can record: 

  • The observation sheet number 
  • The age category of each person  
  • The type of group they were part of - family, school, adult or on their own 
  • The start time, end time, and duration of the observation 

In Excel you can create drop-down menus for things like age category and type of group. This makes data entry easier and more reliable – the person entering the data just selects the appropriate item from the drop-down menu rather than typing it each time.  

You also need to create a series of columns in the spreadsheet to record the occurrence or absence of the behaviours you are looking for. Such behaviours might include: 

  • Did they complete the workshop activity? 
  • Did they leave the event before the end? 
  • Did they work in collaboration with others? 
  • Did they appear to be attentive / happy / participating etc? 

We recommend creating drop-down menus with the options 0 for No and 1 for Yes. Using 1 or 0 like this means that all you need do to collate the data, is sum the total for each of these columns. This shows you the number of people who have displayed each of the behaviours.  

You may want to add additional unformatted columns at the end of the spreadsheet to record behaviours that do not fit into simple ‘it happened / it didn’t happen’ categories. For example: 

  • The number of interactive exhibits was used 
  • The number of questions asked 
  • How long people spent … 

Finally you should add columns where the person entering the data can transcribe the observers comments. 

Enter data into the spreadsheet

Using the spreadsheet described above enter the data captured on each of the observation sheets.  

Remember the data will consist of a mixture of: 

  • 0 – when a behaviour did not occur 
  • 1 – when a behaviour did occur 
  • Numerical data e.g. time spent on activity; number of questions asked etc 
  • Category data – age, type of group 

Data analysis

Once you have entered the data into the spreadsheet, you can analyse it.  

When examining the data you will want to answer questions such as: 

  • What percentage of people demonstrate each behaviour?  
  • Does the frequency of a behaviour vary between different ages or types of group? 
  • Are some behaves frequently occurring together? 
  • How much time are people spending … ? How many … ? 
  • What is hardly ever / never happening? 
  • How does this compare to what was expected to happen / what you wanted to happen?