Using the Observation Technique for Requirements Elicitation

Most people are used to observing events in their personal and professional lives. In some cases, we come to conclusions based on the events we observe. The observation technique is an effective means of deciphering how a user does their job by conducting an assessment of their work environment. It increases the analyst's familiarity with the culture and working style of a group of people. This technique can also be used to verify requirements and deliver instant requirements worthy of consideration.

Every observation must be guided by clearly stated objectives. The analyst should know what data is to be collected, how observation will be done, when and where to observe, how the data will be collected and what the data will be used for after analysis.

Observation technique is best applied where:

• A current process is to be monitored
• The objective is to improve a process
• Stakeholders find it hard to explain what they do or what their requirements are
• Processes are highly repeatable e.g. manufacturing
• The validity of data collected through other means is in question

Observation should be planned to ensure that all the required data elements are pre-determined beforehand. This will reduce uncertainty during the observation session and ensure that the analyst can focus on the task of observing without wondering which event should be recorded and which should not. The more the information collected during the observation session, the more time the analyst has to set aside afterwards to make sense of all the information. Consequently, the level of detail and relevant events to look out for should be determined before observation begins. 

The analyst will also find it useful to determine which period (peak, normal or low) to conduct the observation. Observation may be done first during normal periods and then repeated at peak periods to get a more detailed view of events. The analyst should also request for any document that will be used by the worker during the observation session.

Observation should be done with a conscious effort to eliminate bias. Observations based on the analyst’s opinions are subjective and prone to misinterpretation since different people have different perceptions of reality. To attain a reasonable degree of objectivity, observation is best applied to collecting data for quantitative or statistical analysis.

Observation can be performed with multiple aids or tools; this will ensure that all the important details are captured. A checklist or a tally sheet may be used during observation to record the frequency of events. Some analysts choose to use tools such as video recorders, tape recorders, still photography or note-taking tools to record the details of their observation sessions.

Observation may be done passively or actively. Passive/Invisible observation happens where the analyst has no interaction with the worker while the observation is going on, but takes notes. The analyst would typically ask follow-up questions after the observation session by using a prepared list of questions or asking questions on the fly.

With active/visible observation, the analyst can interrupt the worker to ask questions during the observation session. In some cases, the observer may participate in the activity as an apprentice.

Pros & Cons of the Observation Technique

There are certain advantages to using the observation technique: 

• The data gathered during observation sessions are quite reliable; it is often used to confirm the data extracted using other techniques
• Observation can be used to extract information on the physical environment where the task is performed. For example, the analyst can get information on level of noise, physical layout, traffic and so on
• It is relatively inexpensive
• It allows the analyst to perform work measurements

There are however, some downsides to using the observation technique that you should be aware of. One is that exceptions are difficult to capture in one session; repeated observation sessions and interviews may be needed to supplement the facts gathered. Observers are also prone to bias reflected in the form of seeing what they expect to see and what they want to see, which can affect the results of the observation. 

Also, where work is guided by a high level of intellectual analysis, the observation technique is least effective. Lastly, stakeholders are prone to interruptions during observation sessions and can respond differently when being studied as demonstrated by the Hawthorne Effect.