ANT333 Lecture #5 – Methodological Issues in Studying Primate Positional
Behavior & Adaptations
Major goal of primate evolutionary biology is relate limb anatomy to aspects of locomotor or
Data can be used to reconstruct behavior of extinct primates.
Despite importance of above, relatively few quantitative studies of primate in wild.
Most studies of primate positional behavior focused on platyrrhines and few cattarhine
Data should be relevant to question
Methods in Studies of
Study should be representative of question
Many questions about what sort of data to collect, how to collect data, & how to analyze data?
Answers to these questions depend of what we want to ask of our data.
Most studies are very descriptive, and seek only to provide data useful for explaining anatomical
Ethics need to be considered
Four Important Steps in any Scientific Study
1. ID exact question(s) to be answered and clearly ID relevant population, null hypothesis &
2. Develop plan for collecting data. Plan should describe in detail conduct of an observational
study or an experiment. It should be carefully designed so that collected data are
representative of population in question.
3. Collect the data. Be careful to minimize errors leading to biased data.
4. Analyze the data and draw conclusions. ID all possible sources of error.
Observer veer – difference in observation between beginning and end of research
Morphological vs. Behavioral Data Sets
Morphological and behavioral data sets are very different from each other.
Most aspects of morphology are static over life of primate; easily quantified.
Positional behavior generally sampled for brief time period, and by studies that tend to be
limited in space (1 site).
“Event” versus “State” Behaviors
Event: A behavioral unit of short duration (e.g.. < 5 s) more suitable for recording frequencies
(e.g., a leap by a sifaka).
State: A behavioral unit of long duration (e.g.. > 10 s), which could be timed using a stopwatch
(e.g, walking lemur).
1 Data Collection
Typically, studies of primate positional behavior follow one of two methods:
1. Continuous (bout) sampling
2. Instantaneous time sampling.
Continuous (bout) vs. Instantaneous Sampling
Continuous (bout): Frequency of occurrence of a specified event is recorded during sample
period, or duration of states (start and stop times) are recorded.
Instantaneous: At specified point in time, action of focal animal(s) is recorded; this is repeated at
specified intervals (10-60 sec.) during observation period.
Usually assumed that results--in terms of proportions of behaviors--will be comparable.
In other words, behaviors that occur more frequently than others should take up more of a
However, the above statement depends on duration of behavior.
Let’s look at simple example of how data collection can affect results using data collected on
two species of Malagasy strepsirhines:
Eulemur fulvus and Propithecus diadema.
Why Difference in Sifakas?
Propithecus diadema (diademed sifakas) leap more often than they sit (leaping = 88.2% of all
bouts; sitting=15% of all bouts), but bouts of leaping last much shorter than bouts of sitting
(sifakas spend only 3.5% of their time leaping and 64% of their time sitting).
Continuous vs. Instantaneous Sampling
These data collection techniques designed to answer different questions.
Continuous (bout) recording is designed to measure frequency of occurrence (i.e., how often
does this event occur?).
Instantaneous sampling designed to measure % of scans devoted to states (i.e., what is
proportion of scans devoted to t