Anim Cogn (2014) 17:1019–1022 DOI 10.1007/s10071-014-0735-2
SHORT COMMUNICATION
Suboptimal choice by dogs: when less is better than more Kristina F. Pattison • Thomas R. Zentall
Received: 6 November 2013 / Revised: 27 December 2013 / Accepted: 10 February 2014 / Published online: 18 March 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract The less is more effect, an example of an affect heuristic, can be shown in humans when they give greater value to a set of six baseball cards in perfect condition, than to the same set of six perfect cards together with three additional cards each with some value but in fair condition. A similar effect has been reported in monkeys which will eat both grapes and cucumbers but prefer grapes, when they prefer a single grape over a single grape plus a slice of cucumber. In the present experiment, we tested the less is more effect with a nonprimate but social species, dogs. We used dogs that would eat a slice of carrot and a slice of cheese but preferred the cheese. When we then gave them a choice between a slice of cheese and a slice of cheese plus a slice of carrot, most dogs preferred the single slice of cheese. Thus, the less is more effect appears to occur in several species. Keywords Affect heuristic Less is more Incentive motivation Value Choice Dogs
Introduction We have been told by economists that humans are rational (e.g., Persky 1995). This view led to rational choice theory (Becker 1976) which suggests that humans act rationally unless they do not have adequate information. However, this theory has been challenged by research showing that humans use various heuristics in making decisions and that those decisions are not necessarily rational or optimal K. F. Pattison T. R. Zentall (&) Department of Psychology, University of Kentucky, Lexington, KY 40506, USA e-mail: [email protected]
(Kahneman and Tversky 1979). Humans often base decisions on affective heuristics (their subjective feeling). An example of an affective heuristic known as the less is more effect comes from an experiment in which subjects were asked to evaluate a set of dishes (Hsee 1998). One set consisted of 24 pieces, all in good condition, whereas the other set consisted of 40 pieces, 31 in good condition but nine of them broken. Paradoxically, the smaller number of pieces all in good condition were given greater value (almost 41 % more) than the larger number, some of which were broken, even though the larger number had seven more pieces in good condition. According to Hsee (1996), the affect heuristic will normally be applied to attributes that are readily evaluable, such as item quality, rather than by those that may be objectively more important, such as total worth of a set of items of mixed quality. The overall quality of the set appears to detract from its objective value. Thus, humans may average the quality of a set rather than judge its overall value. Would animals show a similar suboptimal choice effect when it comes to the choice between two quantities of food? According to optimal foraging theory (Stephens and Krebs 1986), they should not, because nature should select against any tendency to prefer an alternative that provides less food over one that provides more food. Thus, animals should always optimize their rewards. However, there is evidence that a similar less is more effect does occur in monkeys (Kralik et al. 2012). In that study, it was found that the monkeys would eat grapes and sliced cucumbers, but when offered a choice between them, they preferred grapes. On critical choice trials, when the monkeys were offered a grape or a grape plus a slice of cucumber, they showed a strong (78 %) preference for the grape alone. Thus, the less is more phenomenon does not
123
1020
appear to depend on culture or biases resulting from human language or specific experiences, since monkeys, too, base their choice on the overall quality of the foods rather than their quantity. Both of the species previously studied were primates, and it is possible that the effect occurs in this order but is not generally true. To test the possibility that the less is more effect is characteristic of primates but not of other species, in the present experiment we asked whether a similar less is more effect would be found in dogs, a nonprimate but social mammalian species.
Methods Subjects Ten dogs (Canis familiaris) that would consume a piece of carrot and a piece of cheese when offered, and had not exhibited a side bias in pretesting, were recruited for the experiment. These four females and six males, ranging in age from 6 to 144 months (M = 43 months), were tested in two sessions each. Of the dogs that participated in the experiment, one was a Belgian Tervuren, one was a Golden Retriever, one was a Siberian Husky, one was a Cavalier King Charles Spaniel, and five were of mixed breed. All the dogs were owned by members of the public. Materials Pilot testing of several possible food items revealed that among less-appealing food options, dogs would readily consume carrots. Thus, the two food choices selected were baby carrots (Bunny-Luv Baby Carrots, Grimmway Farms) and string cheese (Cow Pals String Cheese, Kroger Company). Both items were cut into 1-cmdiameter, 5-mm-thick rounds. Containers of food for testing were hidden behind the experimenter during both sessions. Procedure Dogs were allowed to explore the testing room for 5 min. Dogs were then screened by offering them the following: a piece of carrot, a piece of cheese, and lastly a piece of cheese and a piece of carrot together. Dogs that did not readily eat all the offered treats were dropped. Dogs were then given 5–10 standard carrot versus cheese trials to test for side biases. Dogs that chose one hand exclusively were also dropped. All testing and screening took place inside a white-painted room (3.96 m 9 3.81 m). The experimenter sat on one seat that was positioned along the
123
Anim Cogn (2014) 17:1019–1022
middle of one wall. The handler was seated on another seat facing the experimenter. The dog was seated approximately 1.0 meter from the experimenter directly in front of the handler facing the experimenter. The handler held the dog’s collar so that the dog could initially see and smell, but not reach, the choice being offered by the experimenter. Following the screening trials, the dogs experienced standard trials in which the experimenter held up first her left open hand, then her right hand, each time announcing the type of food available, and then lowering both hands, moving them up and down slightly to attract the dog’s attention, while being careful that each choice was presented in the middle of each hand. The experimenter then fixed her gaze on the floor between the two food presentations, brought her hands close together and low enough for the dog to easily reach the displayed food, paused for three seconds without moving, and then gave the verbal cue ‘‘Choose’’ at which time the handler (eyes fixed on the ground between the handler’s feet) released her hold on the dog’s collar. As soon as any part of the dog touched one of the experimenter’s hands, the experimenter closed her other hand over the unchosen display item(s). Dogs were allowed to consume their selection before being returned to the handler, and the next trial began. Side of item presentation was pseudo randomized across trial types with no more than three trials in a row with the items on the same sides. In each session of 24 standard trials, six probe trials involving a choice between two items and one (counterbalanced for side presentation) were inserted at regular intervals. Because the dogs had not been given such a choice involving two items before, we gave them a second testing session, at least 6 h later, in which there were six probe trials involving a choice between one and two pieces of cheese. If the dogs were merely avoiding two items, they should also tend to avoid two pieces of cheese.
Results All the dogs showed a preference for the cheese over the carrot. To determine whether that preference was stable during testing, there were 24 choice trials during the critical first test session and six less is more probe trials. Mean choice of the single piece of cheese over the single piece of carrot was 86.7 % (SEM = 2.0 %) with a range of 75.0–95.8 % on the first test session and 84.2 % (SEM = 1.9 %) with a range of 75.0–95.8 % on the second test session. Interestingly, the carrot was chosen at least once by each of the dogs, suggesting that it did have some value. The critical trials were the six probe trials on which the dogs were given a choice between one piece of cheese and
Anim Cogn (2014) 17:1019–1022
1021
a piece of cheese together with a piece of carrot. On those probe trials, most of the dogs preferred the single piece of cheese over the cheese plus carrot. The data from the probe trials appear in Fig. 1. As can be seen in the figure, all of the dogs except one, Dakotah, showed a clear less is more effect, and when the data were pooled over all of the subjects, the dogs chose the cheese plus carrot (mean = 26.7 %) significantly less than the single piece of cheese, as indicated by a two-tailed, single-sample t test, t(9) = 3.50, P = 0.007. On the probe trials involving a choice between two pieces of cheese and one piece of cheese, the dogs showed a strong preference for two pieces of cheese (mean = 95.0 %, SEM = 3.54 %). A two-tailed, single-sample t test indicated that the preference for two pieces of cheese was statistically significant, t(9) = 12.65, P \ 0.0001.
Discussion In the present experiment, we found evidence that dogs too show the less is more effect. Although they would eat both carrots and cheese, they showed a strong preference for the cheese. Yet when offered a choice between a piece of cheese plus a piece of carrot and a piece of cheese alone, they preferred the piece of cheese alone. The less is more effect, first demonstrated in humans, is an affect heuristic that results in a preference for the qualitative over the quantitative evaluation of options. Its function appears to have been the rapid evaluation of alternatives. It is likely that in many cases it is relatively easy to judge the qualitative value of alternatives but perhaps more difficult to judge their quantitative value, and when rapid decisions are necessary, such heuristics
Percentage Choice Optimal
100 90 80 70 60 50 40 30 20 10 0 Dakotah Carla
Boon
Maria
Finn Sprinkles Winter
Leo
Lilly
Simon
Subject
Fig. 1 Percentage preference for the optimal alternative (a piece of cheese plus a piece of carrot) over the suboptimal alternative (a piece of cheese alone) for each dog in the study
may be quite functional. For example, within-species competition may favor rapid decisions because hesitation may result in losing food to faster competitor. Furthermore, the risk of predation may provide an advantage to those animals that make rapid decisions even if those decisions are not always optimal when considered in isolation. The present research indicates that the less is more effect is not unique to primates but can occur in other mammalian species, at least those that are socially organized such as primates and canids. It would be of interest to know whether less socially organized species such as rats would show the less is more effect and whether nonmammalian species such as birds would show it as well (see Freidin and Kacelnik 2011). That one of the ten dogs failed to show the suboptimal less is more effect suggests that there may be conditions under which the effect will not occur. The one dog for which the effect did not occur was Dakotah, a dog that had been adopted from the local animal shelter. This dog had been picked up as a stray as an adult and had been on the street for an unknown time. Although all of the dogs in the sample were well cared for pets, Dakotah had had a history of fending for himself. This difference suggests the possibility that level of motivation may play a role in the less is more effect. For example, in all of the cases in which the less is more effect has been found, the level of motivation of the subjects was relatively low. The monkeys in the Kralik et al.’s (2012) experiments were either well provisioned and the choices were between added fruit and vegetables (in the field study), or they were minimally food deprived (in the laboratory study). Similarly, in the present experiment, the carrots and cheese were dog treats that supplemented the normally ample diet of the dogs. For this reason, it might be informative to examine the effect of level of food motivation on the likelihood of finding a less is more effect. Although the less is more effect was first reported in humans, it is interesting that when it occurs in humans it has been found primarily when the alternatives were evaluated independently by different subjects (Hsee 1996). When the same evaluations have taken place in the context of a simultaneous discrimination, subjects generally evaluate the options optimally. The difference between humans and other animals may be that humans typically have had extensive experience in making decisions between alternatives presented simultaneously. It is likely that humans have found a way to overcome the suboptimal consequences of the affect heuristic when making simultaneous comparisons and the suboptimal remnants of the heuristic show up primarily when humans are asked to evaluate alternatives individually.
123
1022 Acknowledgments We would like to thank Jordan Wade for her considerable help in conducting the present experiment. Correspondence can be sent to Thomas R. Zentall, Department of Psychology, University of Kentucky, Lexington, KY 40506.
References Becker GS (1976) The economic approach to human behavior. University of Chicago Press, Chicago Freidin E, Kacelnik A (2011) Rational choice, context dependence, and the value of information in European starlings (Sturnus vulgaris). Science 334:1000–1002
123
Anim Cogn (2014) 17:1019–1022 Hsee CK (1996) The evaluability hypothesis: an explanation for preference reversals between joint and separate evaluations of alternatives. Organ Behav Hum Decis Process 67:247–257 Hsee CK (1998) Less is better: when low-value options are valued more highly than high-value options. J Behav Decis Making 11:107–121 Kahneman D, Tversky A (1979) Prospect theory: an analysis of decision under risk. Econometrica 47:263–291 Kralik JD, Xu ER, Knight EJ, Khan SA, Levine JW (2012) When less is more: evolutionary origins of the affect heuristic. PLoS ONE 7:e46240. doi:10.1371/journal.pone.0046240 Persky J (1995) Retrospectives: the ethology of homo economicus. J Econ Perspect 9:221–231 Stephens DW, Krebs JR (1986) Foraging theory. Princeton University Press, Princeton
SHORT COMMUNICATION
Suboptimal choice by dogs: when less is better than more Kristina F. Pattison • Thomas R. Zentall
Received: 6 November 2013 / Revised: 27 December 2013 / Accepted: 10 February 2014 / Published online: 18 March 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract The less is more effect, an example of an affect heuristic, can be shown in humans when they give greater value to a set of six baseball cards in perfect condition, than to the same set of six perfect cards together with three additional cards each with some value but in fair condition. A similar effect has been reported in monkeys which will eat both grapes and cucumbers but prefer grapes, when they prefer a single grape over a single grape plus a slice of cucumber. In the present experiment, we tested the less is more effect with a nonprimate but social species, dogs. We used dogs that would eat a slice of carrot and a slice of cheese but preferred the cheese. When we then gave them a choice between a slice of cheese and a slice of cheese plus a slice of carrot, most dogs preferred the single slice of cheese. Thus, the less is more effect appears to occur in several species. Keywords Affect heuristic Less is more Incentive motivation Value Choice Dogs
Introduction We have been told by economists that humans are rational (e.g., Persky 1995). This view led to rational choice theory (Becker 1976) which suggests that humans act rationally unless they do not have adequate information. However, this theory has been challenged by research showing that humans use various heuristics in making decisions and that those decisions are not necessarily rational or optimal K. F. Pattison T. R. Zentall (&) Department of Psychology, University of Kentucky, Lexington, KY 40506, USA e-mail: [email protected]
(Kahneman and Tversky 1979). Humans often base decisions on affective heuristics (their subjective feeling). An example of an affective heuristic known as the less is more effect comes from an experiment in which subjects were asked to evaluate a set of dishes (Hsee 1998). One set consisted of 24 pieces, all in good condition, whereas the other set consisted of 40 pieces, 31 in good condition but nine of them broken. Paradoxically, the smaller number of pieces all in good condition were given greater value (almost 41 % more) than the larger number, some of which were broken, even though the larger number had seven more pieces in good condition. According to Hsee (1996), the affect heuristic will normally be applied to attributes that are readily evaluable, such as item quality, rather than by those that may be objectively more important, such as total worth of a set of items of mixed quality. The overall quality of the set appears to detract from its objective value. Thus, humans may average the quality of a set rather than judge its overall value. Would animals show a similar suboptimal choice effect when it comes to the choice between two quantities of food? According to optimal foraging theory (Stephens and Krebs 1986), they should not, because nature should select against any tendency to prefer an alternative that provides less food over one that provides more food. Thus, animals should always optimize their rewards. However, there is evidence that a similar less is more effect does occur in monkeys (Kralik et al. 2012). In that study, it was found that the monkeys would eat grapes and sliced cucumbers, but when offered a choice between them, they preferred grapes. On critical choice trials, when the monkeys were offered a grape or a grape plus a slice of cucumber, they showed a strong (78 %) preference for the grape alone. Thus, the less is more phenomenon does not
123
1020
appear to depend on culture or biases resulting from human language or specific experiences, since monkeys, too, base their choice on the overall quality of the foods rather than their quantity. Both of the species previously studied were primates, and it is possible that the effect occurs in this order but is not generally true. To test the possibility that the less is more effect is characteristic of primates but not of other species, in the present experiment we asked whether a similar less is more effect would be found in dogs, a nonprimate but social mammalian species.
Methods Subjects Ten dogs (Canis familiaris) that would consume a piece of carrot and a piece of cheese when offered, and had not exhibited a side bias in pretesting, were recruited for the experiment. These four females and six males, ranging in age from 6 to 144 months (M = 43 months), were tested in two sessions each. Of the dogs that participated in the experiment, one was a Belgian Tervuren, one was a Golden Retriever, one was a Siberian Husky, one was a Cavalier King Charles Spaniel, and five were of mixed breed. All the dogs were owned by members of the public. Materials Pilot testing of several possible food items revealed that among less-appealing food options, dogs would readily consume carrots. Thus, the two food choices selected were baby carrots (Bunny-Luv Baby Carrots, Grimmway Farms) and string cheese (Cow Pals String Cheese, Kroger Company). Both items were cut into 1-cmdiameter, 5-mm-thick rounds. Containers of food for testing were hidden behind the experimenter during both sessions. Procedure Dogs were allowed to explore the testing room for 5 min. Dogs were then screened by offering them the following: a piece of carrot, a piece of cheese, and lastly a piece of cheese and a piece of carrot together. Dogs that did not readily eat all the offered treats were dropped. Dogs were then given 5–10 standard carrot versus cheese trials to test for side biases. Dogs that chose one hand exclusively were also dropped. All testing and screening took place inside a white-painted room (3.96 m 9 3.81 m). The experimenter sat on one seat that was positioned along the
123
Anim Cogn (2014) 17:1019–1022
middle of one wall. The handler was seated on another seat facing the experimenter. The dog was seated approximately 1.0 meter from the experimenter directly in front of the handler facing the experimenter. The handler held the dog’s collar so that the dog could initially see and smell, but not reach, the choice being offered by the experimenter. Following the screening trials, the dogs experienced standard trials in which the experimenter held up first her left open hand, then her right hand, each time announcing the type of food available, and then lowering both hands, moving them up and down slightly to attract the dog’s attention, while being careful that each choice was presented in the middle of each hand. The experimenter then fixed her gaze on the floor between the two food presentations, brought her hands close together and low enough for the dog to easily reach the displayed food, paused for three seconds without moving, and then gave the verbal cue ‘‘Choose’’ at which time the handler (eyes fixed on the ground between the handler’s feet) released her hold on the dog’s collar. As soon as any part of the dog touched one of the experimenter’s hands, the experimenter closed her other hand over the unchosen display item(s). Dogs were allowed to consume their selection before being returned to the handler, and the next trial began. Side of item presentation was pseudo randomized across trial types with no more than three trials in a row with the items on the same sides. In each session of 24 standard trials, six probe trials involving a choice between two items and one (counterbalanced for side presentation) were inserted at regular intervals. Because the dogs had not been given such a choice involving two items before, we gave them a second testing session, at least 6 h later, in which there were six probe trials involving a choice between one and two pieces of cheese. If the dogs were merely avoiding two items, they should also tend to avoid two pieces of cheese.
Results All the dogs showed a preference for the cheese over the carrot. To determine whether that preference was stable during testing, there were 24 choice trials during the critical first test session and six less is more probe trials. Mean choice of the single piece of cheese over the single piece of carrot was 86.7 % (SEM = 2.0 %) with a range of 75.0–95.8 % on the first test session and 84.2 % (SEM = 1.9 %) with a range of 75.0–95.8 % on the second test session. Interestingly, the carrot was chosen at least once by each of the dogs, suggesting that it did have some value. The critical trials were the six probe trials on which the dogs were given a choice between one piece of cheese and
Anim Cogn (2014) 17:1019–1022
1021
a piece of cheese together with a piece of carrot. On those probe trials, most of the dogs preferred the single piece of cheese over the cheese plus carrot. The data from the probe trials appear in Fig. 1. As can be seen in the figure, all of the dogs except one, Dakotah, showed a clear less is more effect, and when the data were pooled over all of the subjects, the dogs chose the cheese plus carrot (mean = 26.7 %) significantly less than the single piece of cheese, as indicated by a two-tailed, single-sample t test, t(9) = 3.50, P = 0.007. On the probe trials involving a choice between two pieces of cheese and one piece of cheese, the dogs showed a strong preference for two pieces of cheese (mean = 95.0 %, SEM = 3.54 %). A two-tailed, single-sample t test indicated that the preference for two pieces of cheese was statistically significant, t(9) = 12.65, P \ 0.0001.
Discussion In the present experiment, we found evidence that dogs too show the less is more effect. Although they would eat both carrots and cheese, they showed a strong preference for the cheese. Yet when offered a choice between a piece of cheese plus a piece of carrot and a piece of cheese alone, they preferred the piece of cheese alone. The less is more effect, first demonstrated in humans, is an affect heuristic that results in a preference for the qualitative over the quantitative evaluation of options. Its function appears to have been the rapid evaluation of alternatives. It is likely that in many cases it is relatively easy to judge the qualitative value of alternatives but perhaps more difficult to judge their quantitative value, and when rapid decisions are necessary, such heuristics
Percentage Choice Optimal
100 90 80 70 60 50 40 30 20 10 0 Dakotah Carla
Boon
Maria
Finn Sprinkles Winter
Leo
Lilly
Simon
Subject
Fig. 1 Percentage preference for the optimal alternative (a piece of cheese plus a piece of carrot) over the suboptimal alternative (a piece of cheese alone) for each dog in the study
may be quite functional. For example, within-species competition may favor rapid decisions because hesitation may result in losing food to faster competitor. Furthermore, the risk of predation may provide an advantage to those animals that make rapid decisions even if those decisions are not always optimal when considered in isolation. The present research indicates that the less is more effect is not unique to primates but can occur in other mammalian species, at least those that are socially organized such as primates and canids. It would be of interest to know whether less socially organized species such as rats would show the less is more effect and whether nonmammalian species such as birds would show it as well (see Freidin and Kacelnik 2011). That one of the ten dogs failed to show the suboptimal less is more effect suggests that there may be conditions under which the effect will not occur. The one dog for which the effect did not occur was Dakotah, a dog that had been adopted from the local animal shelter. This dog had been picked up as a stray as an adult and had been on the street for an unknown time. Although all of the dogs in the sample were well cared for pets, Dakotah had had a history of fending for himself. This difference suggests the possibility that level of motivation may play a role in the less is more effect. For example, in all of the cases in which the less is more effect has been found, the level of motivation of the subjects was relatively low. The monkeys in the Kralik et al.’s (2012) experiments were either well provisioned and the choices were between added fruit and vegetables (in the field study), or they were minimally food deprived (in the laboratory study). Similarly, in the present experiment, the carrots and cheese were dog treats that supplemented the normally ample diet of the dogs. For this reason, it might be informative to examine the effect of level of food motivation on the likelihood of finding a less is more effect. Although the less is more effect was first reported in humans, it is interesting that when it occurs in humans it has been found primarily when the alternatives were evaluated independently by different subjects (Hsee 1996). When the same evaluations have taken place in the context of a simultaneous discrimination, subjects generally evaluate the options optimally. The difference between humans and other animals may be that humans typically have had extensive experience in making decisions between alternatives presented simultaneously. It is likely that humans have found a way to overcome the suboptimal consequences of the affect heuristic when making simultaneous comparisons and the suboptimal remnants of the heuristic show up primarily when humans are asked to evaluate alternatives individually.
123
1022 Acknowledgments We would like to thank Jordan Wade for her considerable help in conducting the present experiment. Correspondence can be sent to Thomas R. Zentall, Department of Psychology, University of Kentucky, Lexington, KY 40506.
References Becker GS (1976) The economic approach to human behavior. University of Chicago Press, Chicago Freidin E, Kacelnik A (2011) Rational choice, context dependence, and the value of information in European starlings (Sturnus vulgaris). Science 334:1000–1002
123
Anim Cogn (2014) 17:1019–1022 Hsee CK (1996) The evaluability hypothesis: an explanation for preference reversals between joint and separate evaluations of alternatives. Organ Behav Hum Decis Process 67:247–257 Hsee CK (1998) Less is better: when low-value options are valued more highly than high-value options. J Behav Decis Making 11:107–121 Kahneman D, Tversky A (1979) Prospect theory: an analysis of decision under risk. Econometrica 47:263–291 Kralik JD, Xu ER, Knight EJ, Khan SA, Levine JW (2012) When less is more: evolutionary origins of the affect heuristic. PLoS ONE 7:e46240. doi:10.1371/journal.pone.0046240 Persky J (1995) Retrospectives: the ethology of homo economicus. J Econ Perspect 9:221–231 Stephens DW, Krebs JR (1986) Foraging theory. Princeton University Press, Princeton
