曝光臺 注意防騙 網曝天貓店富美金盛家居專營店坑蒙拐騙欺詐消費者

trained observers are not available, or for getting estimates of behaviour levels.
3.9 Choosing the Sampling Period and Focusing the Observations
It is useful to divide observation periods into equal length sample periods. This
assists check-sheet design, and data analysis. A basic observation period includes at
least one complete replication of data collection: that is, each subject is observed a
minimum of once in random order. In general, an observation period will last for ½ -
1 hour during faster tempo operations, and 2-3 hours when things are occurring at a
slower tempo. As a rule of thumb, the duration should be shorter than the fatigue
threshold of the observer. This is no longer than 2 hours in high tempo situations,
and increases as the tempo decreases.
In addition, in a field/operational setting, the observer cannot pay attention to
everything at the same time. Thus, during a sampling period, the observer restricts
their focus. The most common focus is on a single individual (focal subject) such that
all behaviours initiated by that subject are recorded. This may also include
interactions that the focal subject has with other individuals. The focus, which can be
any of the following, needs to be predetermined in the methodology:
1. Focal subject: Individual selected from total group, for example, Duty Officer
(most effective if used with continuous sampling).
DSTO-TR-1034
12
2. Focal subgroup: Related sub-set selected from total group, for example, the Duty
Clerk and Duty Officer within the Operations Cell (most effective if used with
continuous sampling).
3. Group or subgroup, one individual at a time: Duty Clerk and Duty Officer or
entire Operations Cell (most effective if used with instantaneous/scan sampling).
4. All occurrences of certain behaviours: This is equivalent to focusing on the total
group, while restricting attention to certain behaviours, such as communication.
5. Sequences of behaviour: Sequenced behaviours, such as those that occur during
interactions, or while performing complex tasks. Individual identities and rates
of occurrence may have to be sacrificed unless video or audio recorders are used.
6. Location: Focusing of attention on a particular location, such as the Operations
Cell, and recording what events occur there.
To avoid observer bias, the order in which focal subjects or locations are sampled
during an observation period needs to be randomised. Likewise, observation periods
need to be balanced.
3.10 Collecting the Data
There are several ways of recording data, and they vary in the reliability, ease of use,
cost and time required for transcription and analyses. A major problem with audio
and videotape is that they usually require at least three times as much time to
transcribe as to record. However, if rapid, unpredictable behaviours are occurring,
recording them is the most successful way of ensuring that data are captured. One
way to make transcription from audio or video less time-consuming is for the
observer to narrate during the observations (as long as it is not disruptive to the
subjects). This saves re-coding behaviour, and allows a 1:1 transcription ratio. Small
laptop computers are also useful, and can be used with commercial data collection
products such as The Noldus Observer Software. However, they can limit sampling
to single method. As was mentioned, the most common method with experienced
observers is for mixed-sampling methods to be used. In addition, transporting
computer equipment around for data collection in a field setting can be cumbersome
and extremely impractical. In a military setting, when movement around the field is
reliant on Army transport, it is usually impossible to carry more than the basic
necessities. Consequently, in most observation data collection situations, the “oldfashioned
” pencil and paper method is usually the most appropriate.
3.10.1 Codes
Codes are useful for recording behaviour in many sampling scenes. Depending on
how many behaviours are going to be scored, one may simply code each behaviour
with one to three letters or numbers. If there are many behaviours, higher reliability
will occur if mnemonic abbreviations are used (eg. AP = Approach). In addition, a
dimensionalised coding system can be used where the first letter denotes a general
category, while the second a specific behaviour. A teamwork example is “CR =
Communicate-Request”. Codes can also be used to identify individuals and locations
and to discriminate between actors and recipients.
DSTO-TR-1034
13
3.10.2 Data Sheets
As was mentioned, for many field projects, a data sheet is a simple and inexpensive
　
中國航空網 www.k6050.com
航空翻譯 www.aviation.cn
本文鏈接地址：航空資料29(66)