Observations
Apart from the basic
principles, the following issues should be considered in observations (adapted from evaluationsnetz, 2005):
- Observation schedule and coding system: Set up an observation schedule with a coding system for the data collection. This makes it easier to enter and analyse data and ensures a standardised procedure if there are multiple observers.
|
o
|
Observation schedule: Bbefore the observation, define which behaviour, actions, statements and interactions, etc. you want to observe (this depends on the question). Then, set up a list or a table of all the phenomena that you wish to observe with aspects that are particularly relevant. If you need a strictly structured observation (which is primarily to provide quantitative data) you should prepare a very detailed observation schedule. This can be, for example, a table that is filled in during the data collection process (examples of such schedules can be found in: Bortz & Döring, 2003, p. 247 - Tafel 26; Hoang, 2000, Appendix 1). For less structured observations, the observation schedule is less detailed; although the basic questions still have to be precisely specified (cf. Sapsford & Jupp, 1996).
|
|
o
|
Coding system: A coding system allows record “complex behaviour” to be recorded as codes (e.g. the action “the student executes a double-click” can be coded as 'D' or '1'). This makes recording during the observation easier.
|
- Documentation of observational data: In some observational methods (e.g. in the analysis of log files), observational data is to some extent documented automatically. However, when you observe, for example, a group of students during a direct interaction, you should decide whether to record the interaction on video-tapes (or audio-tapes) and subsequently to analyse the video (advantage: a scene can be observed several times), or whether to code 'live', i.e. during the events.
- The following rules should be considered in a live-coding (direct observation):
|
o
|
Observe the situation and the person/people as they influence each other.
|
|
o
|
Avoid interpretations during the recording.
|
|
o
|
Use codes for complex or heterogenous statements.
|
|
o
|
Also record the times (what is observed when and for how long).
|
|
o
|
Replace the observer in long observations.
|
|
o
|
Try to clarify ambiguities by a subsequent interview after the observation.
|
