SOP for Time Budget

The observed time is split into time slots (10 minutes long in this case), and the data sheet provides a row for each timeslot in a 24h period, and a column for each major behavioral category. For ease of reference, alternate hours are shaded white or grey. The sheet contains columns for key behavior categories. These categories each contain many individual behaviors, but are designed to allow the investigator to rapidly pick up functional differences between animals. The categories are as follows:

1. The I column is for recording inactive behaviors
2. The GA column is for recording general activity
3. The M column is for recording maintenance behaviors
4. The NB column is for recording nest building behaviors
5. The AI column is for recording agonistic interactions
6. The SB column is for recording sexual behavior
7. The AB column is for recording abnormal behavior

Time budgets can be collected either using interval or 1/0 recording methods. Which method is used really depends on the research question.

1. 1/0 recording is effective for recording the amount of one or two target behaviors (e.g. agonistic interactions, or stereotypies) that occur briefly, and will be analyzed separately.
2. 1/0 recording is not suitable, primarily for analytical reasons, if the overall budget (i.e. how does the animal spend its time?) is the primary question of interest.
3. If the overall budget is of interest, then interval recording should be used instead, not least because with interval recording only one behavior is scored per timeslot, and thus the total budget always adds up to 100%.
4. 1/0 recording or interval recording can be used for reconnaissance observations - i.e. when simply looking for when different behaviors occur. The choice of method is dictated by whether the key behaviors of interest are brief (such as agonistic interactions or stereotypies), and typically occur in bouts less than the length of the timeslot. If so, then 1/0 recording should be used.

Note that General activity is the exhaustive category in the ethogram. Thus at every timeslot at least one column can be scored. Note however, that 1/0 methods are not necessarily exclusive, in that multiple behaviors include General activity can be scored in the same timeslot.

Proceed with interval recording as follows

1. Move the video to the exact time of the timeslot, and simply record the behavior being performed at this time.
2. It may be necessary to match a few seconds before and after the time point to be certain which behavior to record.
3. Move to the next timeslot.
4. The calculations performed on the raw data will depend on the question at hand. If circadian rhythm or time of day is of interest, the 24h  may be split up into smaller sections, with suitable sample size this may be an hour long, though in practice 3h long sub-blocks are often a good compromise between time resolution, sample size, and smoothing of variability. For each time block (or for the whole 24h period, if a simple budget is required). Simply total the number of timeslots in which each behavior was observed.Simply calculate the total % of time spent in each behavior. Check your work by making sure that the data total to 100%.
5. We generally analyze these data using a multivariate approach. Examples can be found in Some like it hot: Mouse temperature preferences in laboratory housing.

Proceed with 1/0 recording as follows

1. Watch through each timeslot in turn (an experienced observer can easily do this at 16x normal playback speed).
2. Only record a behavior if it STARTS within the timeslot. Or, if it PERSISTS throughout the entire timeslot. Record such behaviors with a slash (/) in the appropriate column. Record behaviors which END, but do not begin in a timeslot with a circle (o).
3. When learning this method it may be easier to pause playback whenever a behavior is observed (or more accurately a new behavior, as each behavior is only scored once per timeslot). However, experienced observers can keep mental note of the categories observed, and good video playback software can be set to automatically pause at the end of each timeslot.
4. Remember, a timeslot may have several different behaviors happen within it, in which case you will want to mark all the appropriate columns.
5. As before, data can be summarized in sub-blocks, or for the whole 24h.
6. For some research questions it may be useful to divide these numbers by the ‘time active’ (or simply, the number of timeslots in which anything other than 'inactive' was observed). This is particularly true for agonistic interactions, or stereotypies which only occur when the animal is active (and hence false positives are easily generated by animals with different levels of inactivity).

Singly housed vs. group housed animals

1. The methods above are straightforward for singly housed animals.
2. For group housed animals, you may be interested in encoding different individuals separately (e.g. if you have treatment and control animals in the same cage), in which case simply keep separate data sheets for each animal.
3. Alternatively, you may be interested in the behavior of the group as a whole. In which case the procedures above are modified differently for 1/0 and interval recording.
   • - For interval recording, record the number of animals performing each category of behavior at each timeslot. Thus the total for each timeslot will equal the number of animals in the cage.
   • - For 1/0 recording, simply treat all animals as one unit - if any animal performs a behavior mark it as occurring. It is relatively rare that a research question will lead you to choose this method.
   • - Calculations are then performed for each method as detailed above.