A delay-specific differential outcomes effect in delayed matching to sample.

In delay-specific remembering, accuracy in delayed matching-to-sample tasks is enhanced after single delays or retention intervals relative to performance at other delays. In the differential-outcomes effect (DOE), accuracy is enhanced at all delays when the outcomes of correct choices are quantitatively or qualitatively different, compared to when outcomes are the same. In the present experiments, we aimed to demonstrate a delay-specific DOE by arranging differential outcomes for correct responses at some delays and same outcomes at other delays. In each of two experiments, four pigeons worked in delayed matching-to-sample tasks with delays of 0.5, 5, and 15 s, or 0 s, 3 s, and 12 s mixed within session. Correct choices produced different reward durations (differential outcomes) at one or two delays, or the same reward durations (same outcomes) at the other delays, on a within-session basis. There was evidence of improved accuracy at delays at which differential outcomes were arranged, compared to accuracy at delays at which same outcomes were arranged, that is, a delay-specific DOE. The more usual DOE was confirmed in a third experiment with same outcomes at all delays in one condition and differential outcomes at all delays in another. We discuss implications of a delay-specific DOE for theories of the DOE which attribute the effect to enhanced stimulus control by expectancies of reward outcomes generated at the time of sample presentation.

In search of consolidation of short-term memory in nonhuman animals.

Wixted (Annual Review of Psychology, 55, 235 ­ 269, 2004) has argued that forgetting is due to consolidation failure. Previous research with humans and nonhuman animals has reported evidence for consolidation in intermediate or long-term memory (LTM). The present study examines whether consolidation occurs in short-term memory in pigeons. Delayed matching-to-sample accuracy was reduced when retroactive interference (an extraneous task in Experiment 1 or houselight illumination in Experiment 2) was interpolated in the retention interval. Accuracy was not greater, however, when interference occurred at the end of the retention interval, as compared with when it occurred at the beginning. That is, there was no evidence for consolidation in short-term memory for pigeons. We did find, however, the beginning­end effect originally reported by Roberts and Grant (Journal of Experimental Psychology: Animal Behavior Processes, 4, 219­236, 1978) and the recovery from forgetting reported by White and Brown (Journal of the Experimental Analysis of Behavior, 96, 177­189, 2011). The results are discussed in relation to temporal distinctiveness theory as an alternative to consolidation.

Suboptimal choice in nonhuman animals: rats commit the sunk cost error.

The present experiments investigated the sunk cost error, an apparently irrational tendency to persist with an initial investment, in rats. This issue is of interest because some have argued that nonhuman animals do not commit this error. Two or three fixed-ratio (FR) response requirements were arranged on one lever, and an escape option was arranged on a second lever. The FRs were of different sizes, and escaping was the behavior of interest. Several variables that might influence the decision to persist versus escape were manipulated: the number of trials with different FR schedules in an experimental session (Exps. 1 and 2), effort to escape (Exp. 2), and the size of the larger FR (Exp. 3). The sunk cost error would result in never escaping, and the optimal strategy would be to escape from the larger FR. The main variable that determined persisting versus escaping was the size of the large FR. Rats that escaped from the large FR-apparently optimal behavior-did so at a suboptimal point, and hence committed the sunk cost error.

Reinforcer probability, reinforcer magnitude, and the reinforcement context for remembering.

Traditional theories of delayed matching-to-sample performance do not predict that accuracy will improve when absolute levels of reinforcement are increased. This prediction emerges only when reinforcement context is considered (J. A. Nevin, M. Davison, A. L. Odum, & T. A. Shahan, 2007). To provide quantitative data, the authors factorially manipulated between conditions the probability and duration of reinforcement for correct choices by pigeons. In Experiment 1, increasing the value of either variable improved initial discriminability of the forgetting functions, but did not affect the rate of forgetting. In Experiment 2, initial discriminability covaried with changes in choice immediacy and trial completion rate, suggesting a relationship with response strength consistent with Nevin et al.'s behavioral momentum model. Adding reinforcement context to K. G. White and J. T. Wixted's (1999) model also generates predictions consistent with the present experiments and with the effects of manipulating extraneous reinforcement. The inclusion of reinforcement context thus improves predictions of delayed matching-to-sample performance.

Timing, remembering, and discrimination.

Four pigeons were first trained in a timing procedure. In one condition, each trial began with the presentation of an X on the center key, followed by a delay (short or long), after which two side keys were lit. If the delay was short, pecks to the red side key were reinforced. If the delay was long, pecks to the green side key were reinforced. In a second condition, the opposite contingencies applied following presentation of a square on the center key. Choice responses were then tested at 10 time intervals ranging from short to long (1 to 4 s and 4 to 7 s in different conditions). The two timing conditions were combined to create a remembering condition in which correct responding depended upon discrimination of both the sample stimulus (X or square) and the delay interval (short or long). Choices varied systematically across delay in timing conditions, but in remembering conditions, accurate choice at the training delays did not initially generalize to intermediate delays. However, with prolonged training in the remembering task, the response pattern began to resemble that of the timing conditions. Generalization gradients were asymmetrical, in accordance with Weber's Law, in that greater generalization occurred with longer delays than with shorter delays.

Seasonal variation in pigeon body weight and delayed matching-to-sample performance.

The weights of 5 pigeons with free access to food, monitored over 3 calendar years in the laboratory, were found to fluctuate with season. All pigeons were at their heaviest in the winter and were lightest in the summer. Five different pigeons performed a standard delayed matching-to-sample task for 44 weeks from January to November. Their weights were held at 85% of their summer free-feeding weights, making their predicted deprivation level higher in the winter relative to predicted winter free-feeding weights. Slopes of forgetting functions fit to weekly response totals for each pigeon were shallower in winter, showing an improvement in accuracy with longer delays. Thus, delayed matching-to-sample performance may have been affected by the practice of maintaining the pigeons at a constant body weight throughout the calendar year.

On the effects of signaling reinforcer probability and magnitude in delayed matching to sample.

Two experiments examined whether postsample signals of reinforcer probability or magnitude affected the accuracy of delayed matching to sample in pigeons. On each trial, red or green choice responses that matched red or green stimuli seen shortly before a variable retention interval were reinforced with wheat access. In Experiment 1, the reinforcer probability was either 0.2 or 1.0 for both red and green responses. Reinforcer probability was signaled by line or cross symbols that appeared after the sample had been presented. In Experiment 2, all correct responses were reinforced, and the signaled reinforcer durations were 1.0 s and 4.5 s. Matching was more accurate when larger or more probable reinforcers were signaled, independently of retention interval duration. Because signals were presented postsample, the effects were not the result of differential attention to the sample.

The sunk cost effect in pigeons and people: a case of within-trials contrast?

Pigeons show the sunk cost effect in procedures in which their choice between two outcomes is biased by a prior investment. We review recent studies of the sunk cost effect in pigeons, in which choice procedures are analogous to studies with humans using hypothetical scenarios to make explicit a prior investment and the later choice. Zentall's (2010) theory of within-trial contrast can account for the sunk cost effect - an effortful prior investment contrasts with choice outcomes to increase the value of the outcome in which the prior investment was made. The account correctly predicts that in both pigeons and humans, increased prior investment increases the sunk cost effect. We present data from a study with humans using hypothetical scenarios in which delay was varied between the time of the prior investment and later choice. Extending the delay reduced the sunk cost effect, suggesting the need for a second process by which value is depreciated, in addition to the value-enhancing effect of contrast. This article is part of a Special Issue entitled: Tribute to Tom Zentall.

Local proactive interference in delayed matching to sample: the role of reinforcement.

Discriminability in delayed matching to sample was lower when the samples on consecutive trials differed compared with when samples on consecutive trials were the same. This local proactive interference occurred when correct choices on the previous trial were reinforced but not when correct choices on the previous trial were not reinforced. When the choice on the previous trial was incorrect, discriminability was higher on different consecutive trials than on same trials. These effects were amplified by varying the ratio of reinforcers for correct choices, as predicted by a model that attributes local proactive interference to an interaction between control by the sample on the current trial and the influence of reinforcers for correct choices on previous trials.

Memory deficits in Alzheimer's disease: the encoding hypothesis and cholinergic function.

Forgetting functions generated by delayed matching-to-sample procedures allow delay-dependent effects to be distinguished from delay-independent effects on working memory. Parameters of negative exponential functions estimate initial discriminability (intercept) and rate of forgetting (slope). Forgetting functions for patients with Alzheimer's disease indicate that they differ from normal controls in terms of reduced initial discriminability--that is, in the encoding component of memory performance--but not convincingly in rate of forgetting. Reanalyses of previous studies with different species suggest that pro- and anticholinergic drugs influence initial discriminability in delayed matching-to-sample performance, but not rate of forgetting. The results of our reanalyses are consistent with the conclusion that the cholinergic system plays a role in the encoding component of working memory and that this is the main characteristic of the memory deficit shown by patients with Alzheimer's disease.

Need probability effects in animal short-term memory.

Five pigeons performed in a delayed matching-to-sample (DMTS) procedure with five delay durations (0.5, 2.5, 5, 10 and 20 s) mixed within sessions. Contrary to the predictions of need probability theory, discriminability decreased when fewer short than long delays were included in each session. To test whether the decrease in discriminability was due to a decrease in obtained reinforcement at short delays, the number of trials at each delay was held constant and reinforcer probability was increased with increasing delay. This manipulation produced a similar decrease in discriminability as when the frequency of delays was manipulated. It was concluded that the effect of delay frequency on the forgetting function is mediated by the effect of the reinforcer distribution, which influences discriminability by weakening stimulus control.

Amount-dependent temporal discounting?

Amount-dependent temporal discounting has been demonstrated for human choice between outcomes differing in amount and delay. In the only study to date with non-humans, Grace reported no evidence for amount-dependent temporal discounting with pigeons in a concurrent-chains procedure. The present experiments repeated Grace's procedure but with modifications to enhance the discrimination between small and large magnitude outcomes. In Experiment 1, sensitivity of pigeons' initial-link choice to the terminal link delay ratio was greater with large reinforcer durations in the terminal links than with small reinforcer durations. This result is consistent with a greater rate of temporal discounting for larger reinforcers (the reverse of the result for humans), but can also be explained as enhanced discrimination of delay ratios with larger reinforcer durations. The results of a second experiment supported Grace's conclusion that amount-dependent temporal discounting does not characterize pigeons' choice in concurrent chains. Because reinforcer amount was held constant between choice alternatives in the present experiments and that of Grace, but varied in the human studies, our results question whether prior demonstrations of amount-dependent discounting reflect the effects of reinforcer delay or of reinforcer amount. Differences in the procedures used to study discounting in humans (titration procedures) and non-humans (concurrent chains) may contribute to the divergent results across species.

On the form of the forgetting function: the effects of arithmetic and logarithmic distributions of delays.

Forgetting functions with 18 delay intervals were generated for delayed matching-to-sample performance in pigeons. Delay interval variation was achieved by arranging five different sets of five delays across daily sessions. In different conditions, the delays were distributed in arithmetic or logarithmic series. There was no convincing evidence for different effects on discriminability of the distributions of different delays. The mean data were better fitted by some mathematical functions than by others, but the best-fitting functions depended on the distribution of delays. In further conditions with a fixed set of five delays, discriminability was higher with a logarithmic distribution of delays than with an arithmetic distribution. This result is consistent with the treatment of the forgetting function in terms of generalization decrement.

Social influence in pigeons (Columba livia): the role of differential reinforcement.

Socially-influenced learning was studied in observer pigeons that observed a demonstrator in an adjacent chamber performing a target response comprising standing on a box and pecking a key 10 times. In Experiment 1 there was no evidence for social learning in the absence of reinforcement of the observer's behavior. When the target response was already established in the observer's repertoire, but was not differentially reinforced in relation to the demonstrator's behavior, rates of extinction were not influenced by the demonstrator's behavior (Experiment 2). Reinforcement of the observer's target response in the presence of the modeled target response, and not in its absence, resulted in control of the observer's responding by the behavior of the demonstrator (Experiments 3 and 4). This control was extended in Experiment 5 to deferred responses that occurred following a delay since the demonstrator's target responses. The acquisition of social influence depended on differential reinforcement of the observer's target response, with the demonstrator's target behavior serving as the explicit discriminative stimulus.

The effect of reinforcer delays on the form of the forgetting function.

Pigeons were trained in a matching-to-sample procedure with retention intervals of 0, 2, 4, 6, and 8 s mixed within each session. In different conditions, reinforcement was delayed by 0, 1, 2, 4, 6, or 8 s from correct choice responses. Discriminability decreased with increasing retention-interval duration and with increasing reinforcer delay. Exponential forgetting functions were fitted to discriminability measures plotted as a function of retention interval. Initial discriminability (intercept of the fitted functions) decreased with increasing reinforcer delay. Rate of forgetting (slope of the fitted functions) increased with reinforcer delay, suggesting an interaction between the effects of reinforcer delay and retention interval. The data were well described by multiplying an exponential function describing the effects of retention interval by a hyperbolic function describing the effect of reinforcer delay. This description included an interaction term that allowed for a greater effect of reinforcer delay at longer retention intervals.

A good time to leave?: the sunk time effect in pigeons.

Persistence in a losing course of action due to prior investments of time, known as the sunk time effect, has seldom been studied in nonhuman animals. On every trial in the present study, pigeons were required to choose between two response keys. Responses on one key produced food after a short fixed interval (FI) of time on some trials, or on other trials, no food (Extinction) after a longer time. FI and Extinction trials were not differently signaled, were equiprobable, and alternated randomly. Responses on a second Escape key allowed the pigeon to terminate the current trial and start a new one. The optimal behavior was for pigeons to peck the escape key once the duration equivalent to the short FI had elapsed without reward. Durations of the short FI and the longer Extinction schedules were varied over conditions. In some conditions, the pigeons suboptimally responded through the Extinction interval, thus committing the sunk time effect. The absolute duration of the short FI had no effect on the choice between persisting and escaping. Instead, the ratio of FI and Extinction durations determined the likelihood of persistence during extinction.

Attentuation of the differential outcomes effect by extraneous reward.

The reinforcement context model for performance in delayed matching to sample tasks (White and Brown, 2014) predicts the course of forgetting based on the assumption that rewards for extraneous behavior compete with rewards for accurate matching and increase as a linear function of retention-interval duration. In the differential outcomes effect, greater matching accuracy occurs when correct choices produce different outcomes, which the model assumes have greater reward effectiveness than same outcomes. The model was tested in the present experiment with pigeons by arranging an additional task during the retention interval of a delayed matching to sample task, center-key pecking rewarded by food delivered at variable intervals. This additional source of extraneous reward resulted in attenuation of the differential outcomes effect as predicted by the model. The model was supported by satisfactory quantitative fits to the forgetting functions for same and different outcome conditions with and without additional extraneous reward.

Persistence in extinction: the sunk time effect.

We investigated the sunk time effect, persistence in a nonpreferred option owing to prior investment of time in that option. Pigeons chose between two concurrently available keys-2 fixed-interval (FI) food requirements were arranged on 1 key, and an escape option, which terminated the current trial and started a new one, on a second key. One FI was longer than the other, and the shorter FI was always more probable on any given trial. In most conditions, the different FI schedules were not signaled. In this situation, the optimal behavior would be for pigeons to escape from the long FI once the duration equivalent to the short FI had elapsed without reinforcement. Several variables that could influence persistence or escape behavior were manipulated: the presence and absence of cues signaling the type of trial in effect (Experiments 1 and 2), extinction in the long interval (Experiments 3, 4, and 5), the intertrial-interval duration (Experiments 5 and 6), and the duration of the FI schedules (Experiments 1-6). Overall, the results showed that pigeons tend to persist and finish the current trial, even with extinction in the long interval, a result consistent with the sunk time effect.

The effect of a prior investment on choice: the sunk cost effect.

We investigated the effect of prior investment on choice in pigeons, namely, the sunk cost effect, which is a tendency to continue an endeavor once a prior investment has been made, despite a better option being available. In a concurrent-chains procedure, pigeons chose between left and right keys in the choice phase leading to different work requirements in the outcome phase. Within each session, two components were signaled by red or green keys in the choice phase. Components were identical, except that in red components, the choice was preceded by a prior investment of 20 pecks on the left key, whereas in green components, the investment of 20 pecks was on the right key. Preference for the key with the prior investment was studied in a series of experiments in which we manipulated the absence or presence of the investment (Experiments 1a and 1b) and size of the investment (Experiments 2 and 3). We also investigated whether the bias observed was a result of carryover effects or of the sunk cost effect (Experiment 4). Overall, the results showed that choice was biased toward the alternative associated with the prior investment, consistent with the sunk cost effect, an effect that can be understood in terms of within-trial contrast and the delay-reduction hypothesis.

Sunk cost and work ethic effects reflect suboptimal choice between different work requirements.

We investigated suboptimal choice between different work requirements in pigeons (Columba livia), namely the sunk cost effect, an irrational tendency to persist with an initial investment, despite the availability of a better option. Pigeons chose between two keys, one with a fixed work requirement to food of 20 pecks (left key), and the other with a work requirement to food which varied across conditions (center key). On some trials within each session, such choices were preceded by an investment of 35 pecks on the center key, whereas on others they were not. On choice trials preceded by the investment, the pigeons tended to stay and complete the schedule associated with the center key, even when the number of pecks to obtain reward was greater than for the concurrently available left key. This result indicates that pigeons, like humans, commit the sunk cost effect. With higher work requirements, this preference was extended to trials where there was no initial investment, so an overall preference for the key associated with more work was evident, consistent with the work ethic effect. We conclude that a more general work ethic effect is amplified by the effect of the prior investment, that is, the sunk cost effect.