Measures of Tameness

The scientific approach is to measure the progress on our way to our goal.

If we find ways of measuring progress between the wild and tame state, we can begin to optimize our efforts towards making wild horses tractable to handling. In 2009 and 2010 we examined the measurement of one standard of tameness, flight distance. This was a preliminary investigation to expose the limits, properties, and virtues of using this measure in the adoption facility environment. Our hope was to provide a way to objectively compare taming methods, horses, and the effects of pen-mates..

Oxen taming provides a model we can use.

The taming of oxen, as studied by Dr. Drew Conroy, provides the best understood model of animal taming. Conroy breaks taming into two basic processes: initial acclimation to humans and capture/restraint training. During the initial acclimation the goals are: 1) to calm the animals; 2) to reduce their flight distance; 3) to make them realize they need not fear humans; and 4) to let them become familiar with their trainer. During the second process the animals learn to accept touch, handling, and to accept restraint. When mustangs are given the same type of training, the need for special facilities at the adoptive home is diminished. The wild horse quickly becomes as tame as a domestically raised horse, ready to be trained to ride or drive.
Using the oxen model, we can contrast the states of wild and tame horses.

Table 1. Comparison of Wild and Tame Animals.
Low threshold for flight
Tense in the presence of humans
Intense fear
Resists restraint
 No or high threshold for flight
Calm in the presence of humans
No or little fear
Accepts restraint

Flight Distance is the variable of interest.

Flight distance (FD) is the distance the human can approach the animal before the animal moves away. It also known as escape distance, flight initiation distance, approach distance, reaction distance or flush distance. Flight distance has been studied in relation to optimum escape theory, the design of wildlife refuges, and as a measure of the human-animal relationship . Flight distance is both species specific and subject to decrease by habituation. It has been found to be consistent over time in individual goats, heifers, and poultry. A study in Japan showed that FD decreased in Thoroughbred horses who were more extensively handled by their caretakers. When it is used for the assessment of animal welfare, measuring flight distance as a reaction to a moving human resembles the everyday handling of the animals. The human approaches the animal in a slow standard way and the test is completed when the animal steps away.

What properties does a measure need to be useful?

Disciplined scientific thinking requires that we quantify the confidence we have in our conclusions. The elements of animal psychology are affected by many causal factors, uncontrollable and often unidentifiable. Statistics is needed to measure such variable phenomena with a degree of precision and ascertain the reality of observed differences. Scientists working on quantifying the human-animal relationship suggest that the variables and analysis used to study human–animal relationships should ideally be established as reliable and valid prior to their use. The purpose of our study was to examine the statistical validity and reliability of flight distance measures.

What other questions can we ask about the problem?

The primary objective of this study is to examine the value of measuring flight distance in reaction to a moving human during the initial gentling of wild horses as a measurement of tameness. In addition to examining the statistical properties of FD, I wanted to pose the following questions:
  1. Is flight distance a characteristic of an individual animal at a given stage of tameness?
  2. Does flight distance decrease over time with and without training?
  3. What kinds of problems exist in measuring flight distance within a wild horse holding facility?
  4. Is it feasible to use flight distance as a criterion to evaluate the welfare of captive wild horses?
  5. Does flight distance differ between animals trained by different methods?
  6. Does flight distance differ by gender or age of animal?
  7. Does the calculation of the daily change in FD provide a more useful variable than FD?


The results of this exploratory analysis suggest that horses have somewhat stable flight distances which are open to be modified by training. There is no apparent age limit to modification of the flight response. We found no evidence that age or gender was a major factor, but rather training and especially the type of training were the identifiable factors responsible for variation in the flight distance measure. Although training modifies flight distance, it does not seem to do it immediately. We saw some suggestion that stress associated with moving may temporarily affect FD, but this was confounded with other changes and activities of adoptions. We saw that changes in the composition of subjects strongly impacts the distribution of FD scores observed in the herd. The unbalanced aspect of using an intact “herd” was found to confound almost all of our attributes thus creating problems for statistical interpretations about groups. Error variation and statistical regression created by extreme scores overwhelmed our ability to detect true flight distance changes in the within-subject Matched Pairs analysis. This suggests that true daily progress towards lower FD happens in smaller increments than error variation.

The complete flight distance paper is attached to this page on the link below.
Patricia Barlow-Irick,
Feb 21, 2011, 7:06 PM