Introduction to the study

The authors of this study assert that mentally challenged humans have had preference assessment tests used on them, in conjunction with behavioral modification sessions, at a regular basis. Preference assessment-tests are therefore a regular component of all operant conditioning programs that have been applied to these mentally challenged people. It’s effectiveness in the improvement of the efficiency of behavioral training in such settings is exemplary, and the authors suggest it could be similarly effective in zoology and environmental research.

The study conducted was of the preferences of nine captive orangutans, and were investigated for different food items. The researchers used a pair wise presentation onto, in the recording of nonhuman, animal preferences as concerned five different foods. Over the course of six months, this was on six different occasions.  The orangutans, according to the Results of a Friedman’s 2-way, were found to show a clear overall preference for simple apple. For the five foods, though, there was a significant variability among different orangutans in the preference rankings. This assumption was derived by the readings shown by a Kendall’s tau. Additionally, there was a variability preference in the rankings observed across time for every orangutan. Basically, the preferences of the Orangutan’s change over time. The variations also occur according to individual and regular assessments that should identify items that were used as rewards in behavioral husbandry training. It could be used as a part of feeding enrichment strategies.

The researchers are inspired and motivated by the fact that over the past 15 years, very positive reinforcement in training has become popular in aspects of zoological institution and in non-human animal laboratory facilities. A difference exists on how individual animals respond to trainings and the speed of their progress. The researchers observe that a first reason for this variation is how positive reinforcement creates a difference in the response of individuals. It’s typical how preferences of individual primates for particular food rewards are assessed by means of no empirical uses by people who conduct training. However there are obvious imperfections to the assessment. An empirical means may be helpful to determine individual preferences.  The researchers acknowledge the extensiveness of the research done with human subjects.

A look at the human research as a model when initiating preference research studies on nonhuman primates is therefore necessary.

A description of the methods

The subjects

To obtain quality results for their study, the researchers had to choose the subjects and analyze their procedures carefully.

They chose nine Orangutans that were housed at Zoo Atlanta and these included 4 adult males 1 juvenile, as well as 4 adult female Orangutans. Among them were seven Sumatran Orangutans and 1 adult male Bornean Orangutan as well as an adult male hybrid.

In an indoor/outdoor holding area, they were safely housed. It was during the mornings that all test sessions were conducted, and this had to be before the animals were fed. It was done in the indoor holding area as well.

While Andrea W. Clay conducted the testing on 6 different occasions from September to December 2006, all the testing was complied with protocols that received approval by the Zoo Atlanta as well as by Georgia Institute of Technology. This is the institute of Animal Care and Use Committee.


Grapefruit, orange, apple, sweet potato, and carrot were the 5 items of fresh produce tested and each of the subjects were familiar with each of these foods. A standard PC method was used when presenting the items and it ensured a balance for left and right location and the randomization of the pairs presented during each session. The arrangement made on each possible pairing had item (a) on the left while (b) was maintained on right, then (b) on left and (a) on the right and was presented once per session in that it resulted to 20 pairings per session. 1 inch is the rough estimate for the food piece cuttings. A lexan tray was used to present the food pieces, and it had two equally sized recessed dishes at one end.  The animals selected one of the two items by pointing when this end of the tray was help up to the mesh of the Orangutans enclosure. Each of the animals easily pointed to one of the two items. But there were occasions where they pointed at once, to each of the two items. As indicated, 6 sessions for each of the animals were conducted for each animal over the period of the six months. To determine if there were individual biases, a Chi-square test was used and the total left versus total right selections for each animal throughout all six sessions.

If the animal’s selections were being determined by food preference, rather than side bias, then an approximate of 50% left and 50% right should have been the score. The preference of the group of subjects was analyzed using a Friedman’s two-way ANOVA and Kendall’s tau tests were used to analyze the preferences of the group of subjects. Therefore, Kendals tau tests were used to analyze the different orangutans and specifically for individual orangutans across test sessions.

 A summary of the results

Of the female orangutans, one actually showed a significant side bias in her responding as it indicated a significant Chi-square value; hence the researchers dropped her from the analysis.

A score was obtained for each of the five food items, and each of the test session by using this formula; number of times selected/ number of times the item was presented. The 100D% score was for each of the particular items. The calculation was done for each of the individual, and hence there was a score for each food item as obtained in each test session for every individual animal.

The computation of a mean score was done for each item and the rating across all six sessions for all the orangutans. An indication by A Friendman’s two-way test was that the Orangutans, in all, had a preference for some of the food items over others. Apple was the highest in the overall score, and what followed were the orange, sweet potato, grape fruit, and finally the carrot. The tabulation was tabled.

Individual preferences were compared, and mean percentage score were obtained for each food item throughout all six sessions for each individual animal. A comparison of the scores was done too.

Individual rating scores were graphed for the foods across six sessions and this was necessary to access the stability of each individual orangutan.  Some animals, though, had relatively stable score while others did not.


The importance of these applications could not be emphasized more, according to the researchers. And when the goal of the entire study is to use items from the animal’s daily diet for training and enrichment purposes, it is even more beneficial. The use of preference-assessment tests in conjunction with enrichment or training procedures to determine effects of using preferred items could provide a logical progression from the findings of the study and can use preferred items as rewards. This is opposed to the use of randomly selected items. It has shown that there is an increased efficiency in training behaviors in humans when this is used and should therefore be considered assessing as a method for increasing efficiency of training behaviors in non-human primates as well.