Video Recording

Video Recording

Description

Video recording is a technique rather than a method. It is often used as an intermediary where direct observation is not practical.

Discussion/Reference

Video Recording is a technique, rather than a method, providing what seems at first to be a means of preserving the activities recorded for more leisurely analysis. There is a temptation, when planning a study, to include video-recording of the entire process. This is often a waste of time and resources, leading only to racks of poorly identified tapes in storage.

The definition available on standard domestic video-cameras and camcorders is not sufficient to provide readable copies of video displays. Recordings may also be affected by synchronisation problems, producing dark bands on the display image.

Professional video experts are aware of these problems, and have ways of avoiding them. As a minimum, S-VHS rather than VHS recording should be used.

It is increasingly easy to use direct digital recording. Although the amount of storage required appears considerable, sophisticated data reduction and storage techniques are readily available.

Some additional techniques are available which may simplify the analysis of video recordings. Multiple cameras may be recorded on a single record, either as split screen images, where each camera frame contains images from several cameras (with reduced precision), or by storing successive images on the same tape. The latter technique provides images which are as good as separate recordings, but at wider intervals. Some systems weight the sampling frequency according to the changes taking place in the images, so that a camera observing a vacant work position would take very few images, while one where intense activity was taking place would be sampled with maximum frequency.

Categories
Generics
Type of method Subjective
Target of method Stress, Strain
Video recordings may be used as raw material for a variety of techniques, some aimed at workload (stress), others at strain.
Time Scale of method Seconds
Video recordings are usually made at 25 frames per second, but distinguishable activities usually take seconds to identify.
Portability of method No
Practically, video cameras and recorders need to be set up and aligned with care before observation starts.
Observer Effect No
Operators rapidly ignore the presence of video cameras.
Context of studies
Laboratory studies Use
 
Simulation studies Use
 
Field studies Use
 
Potential problems with the method
Failure risk Moderate
Modern video equipment can operate in lower light than unaided human vision, but unless care is taken in positioning cameras, it may be difficult to identify critical activities from records.
Bias risk None
Recording, in itself, should not be subject to bias.
Ethical problems Privacy
Video recordings may show activities that reflect on safety or morality. The experimenter should agree with the operator(s) beforehand how the recordings may be used, and ensure that this agreement is observed.
Costs of the method
Staff Cost High
Video recording usually requires skilled or experienced operators.
Set-up Cost High
Video cameras and recorders are still expensive. Although digital Webcams are cheap, they lack definition, and may also lack the ability to transmit sufficiently frequent images.
Running Cost Moderate
Video tapes or discs are now reasonably priced.
Analysis Cost High
The analysis of video recordings is an extremely tedious performance, taking far more time and effort than most observers expect.
Analysis data
Analysis Speed Slow
Conclusions are available immediately.
Data Automation Difficult
Normal video recordings do not provide automatable data. However, where (expensive) special equipment and software is available, data collection can be automated.
Analysis Automation Difficult
Although most analysis requires extensive human effort, some forms of video analysis can be automated.
Status Established
Routinely used, although most video recordings are filed without being analysed.

SKYbrary Partners:

Safety knowledge contributed by: