Effects of behavior-correlated reinforcement magnitude and cocaine using an automaintenance procedure

MISSING IMAGE

Material Information

Title:
Effects of behavior-correlated reinforcement magnitude and cocaine using an automaintenance procedure
Physical Description:
x, 90 leaves : ill. ; 29 cm.
Language:
English
Creator:
Miller, Michelle L
Publication Date:

Subjects

Subjects / Keywords:
Psychology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Michelle L. Miller.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003163567
System ID:
AA00002037:00001

Full Text
EFFECTS OF BEHAVIOR-CORRELATED REINFORCEMENT MAGNITUDE
AND COCAINE USING AN AUTOMAINTENANCE PROCEDURE
By
MICHELLE L. MILLER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004


Copyright 2004
by
Michelle L. Miller


ACKNOWLEDGEMENTS This research was supported by USPHS Grants DA04074 and DA 16009 from the National Institute on Drug Abuse. I would like to thank Margaret Gratton, Julie Marusich, Jonathan Pinkston, Matthew Weaver, and Jin Ho Yoon for assistance in experimentation.

111


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS
111
LIST OF TABLES......................................................vi
LIST OF FIGURES....................................................vii
ABSTRACT...........................................................ix
CHAPTER
1 INTRODUCTION
2 GENERAL EXPERIMENTAL METHOD
Subjects....................
Apparatus...................
General Procedures...........
Training....................
Drug-Administration Procedures
3 EXPERIMENT 1
1
Automaintenance (Baseline)...........................................
Behavior-Correlated Schedule: Direct Relation between Responding
and Grain Availability..............................................
Behavior-Correlated Schedule: Inverse Relation between Responding
and Grain Availability.............................................'
4 RESULTS: EXPERIMENT 1.....................
Auto shaping (Acquisition)........................
Behavior-Correlated Schedules....................
Behavior-Correlated Schedule: Direct Relation between Responding and Grain Availability................
iv


Behavior-Correlated Schedule: Inverse Relation between Responding
and Grain Availability..............................................39
5 EXPERIMENT 2...................................................50
6 RESULTS: EXPERIMENT 2..........................................52
Automaintenance (Baseline)...........................................52
Behavior-Correlated Schedule with a Direct Relation.......................61
Behavior-Correlated Schedule with an Inverse Relation.....................62
7 DISCUSSION......................................................71
Experiment 1.......................................................71
Experiment 2.......................................................79
LIST OF REFERENCES.................................................84
BIOGRAPHICAL SKETCH..............................................90
v


Table
LIST OF TABLES
page
1 Response requirements, grain-access times, and number of sessions in each
condition for all pigeons during automaintenance and behavior-correlated schedules.......................................................12
2 Mean number of responses per trial as a function of preceding
intertrial interval (seconds) during automaintenance......................20
3 Mean overall responses per minute across conditions of automaintenance,
rate-increasing, and rate-decreasing schedules..........................23
4 Mean overall latency in seconds to first key peck on each trial during
the initial automaintenance condition, return to automaintenance, combined automaintenance data, rate-increasing and rate-decreasing correlated contingencies....................................................43
5 Mean overall time in seconds from last key peck during 8-s key illumination
to grain delivery for automaintenance, return to automaintenance, combined automaintenance data, rate-increasing and rate-decreasing schedules
for Pigeons 4986, 11, 446, 922, and 198..............................44
6 Number of administrations of each dose in each of the three conditions
(automaintenance, rate-increasing and rate-decreasing schedules)...........50
7 Mean grain-access time in seconds for each trial and ratio of head-in-hopper
time to mean grain-access time as a function of dose of cocaine during acute dosing under automaintenance and behavior-correlated contingencies (direct
and inverse relations between response rate and reinforcer magnitude).......60
8 R2 values, slopes, and standard error of estimate for data shown in Figure 14. 70
VI


LIST OF FIGURES
Figure page
1 Frequency distributions of number of responses per 8-s key illumination
during the initial automaintenance condition (baseline)..................15
2 Responses per minute across the first 40 successive sessions during the
initial automaintenance condition (baseline) for all pigeons...............22
3 Number of trials in which 2-, 4-, and 8-s access to grain was earned as
a function of successive sessions during the rate-increasing schedule........27
4 Number of trials in which 2-, 4-, and 8-s hopper durations were earned
across successive sessions during a return to a rate-increasing schedule for Pigeons 11, 922, 198, and 446...................................... 30
5 Number of trials in which 2-, 4-, and 8-s durations were earned during a
rate-increasing schedule for Pigeons 922 and 446....................... 34
6 Number of trials in which 2-, 4-, and 8-s hopper durations were earned
during a rate-increasing schedule for Pigeon 198.......................38
7 Number of trials in which 2-, 4-, or 8-s access to grain was earned as
a function of successive sessions for Pigeons 4986, 11, 922, 446, and 198... .41
8 Number of trials in which 2-, 4-, or 8-s access to grain was earned as a
function of successive sessions for Pigeons 11, 922, and 4986.............49
9 Responses per minute as a function of dose of cocaine (0.1 to 10.0 mg/kg)
for all pigeons during the initial automaintenance condition (baseline).......55
10 Normalized dose-response functions (0.1 to 10.0 mg/kg) during
automaintenance (baseline) and return to automaintenance................57
11 Mean head-in-hopper time (seconds) as a function of dose of cocaine
(0.1 to 10.0 mg/kg) during automaintenance (baseline) and a return to automaintenance for Pigeons 11, 446, 922, 198, 4986, and 829 ............59
vn


12 Percent of saline response rates across a range of doses of cocaine
(0.1 to 10.0 mg/kg) during behavior-correlated schedule with
direct relations (i.e., rate-increasing schedule) for Pigeons 11, 446, 922,
and 198........................................................65
13 Percent of saline response rates as a function of dose of cocaine
(0.1 to 10.0 mg/kg) during the initial automaintenance condition
and rate-decreasing schedule for Pigeons 11, 446, 922, 198, and 4986.......67
14 Rate-dependency plots display the logarithm of drug response rate as a
function of control response rate for Pigeons 446, 922, 198, 11, and 4986. .. .69
*

Vlll


Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
EFFECTS OF BEHAVIOR-CORRELATED REINFORCEMENT MAGNITUDE
AND COCAINE USING AN AUTOMAINTENANCE PROCEDURE
By
Michelle L. Miller December 2004
Chair: Marc N. Branch
Major Department: Psychology
Effects of psychomotor stimulants on operant behavior often depend on baseline rates of responding, a ubiquitous phenomenon known as rate dependency. Key pecking of pigeons often has been analyzed in rate-dependency studies. Autoshaping, a Pavlovian method, can induce key pecking. In the current experiment, six White Carneau pigeons were exposed to an automaintenance procedure in which 4-s access to grain was delivered independent of pecks on a disk that was illuminated for 8 s before each food presentation. Key pecking was established for all pigeons. A range of doses of cocaine (0.1 to 10.0 mg/kg) was then administered. Rates of responding under small doses were similar to control rates, whereas responding was diminished at large doses for all pigeons. Operant responding was established when behavior-correlated schedules were added to the automaintenance procedure. The amount of food delivered at the end of each 8-s key illumination depended on the number of responses that occurred during the interval. In one correlated schedule, 2-, 4-, or 8-s access to grain was available depending on whether
ix


the number of responses per trial was less than 34%, between 34 and 65%, or greater than 65% of baseline values, respectively. Response rates increased relative to those generated under automaintenance for 5 of the 6 pigeons. In the other correlated schedule, 8-, 4-, or 2-s access to grain was available if the number of responses per trial was less than 34%o, between 34 and 65%), or greater than 65% of baseline values, respectively. Response-rate decrements were observed for all pigeons. Dose-response functions for cocaine were determined under both correlated schedules. The main findings of the study were that a) operant contingencies produced changes in response rates but not in latencies to respond; b) rates of responding and reinforcement magnitude decreased systematically as dose of cocaine increased during automaintenance and behavior-correlated schedules; and, c) effects of cocaine were correlated with baseline rates of responding (i.e., "rate-dependent" effects were observed). These results are consistent with the view that operant contingencies must be in effect for rate dependency to be observed.
x


CHAPTER 1 INTRODUCTION
Drug effects on behavior are determined, to some extent, by schedules of reinforcement (cf. Dews, 1955a,b, 1958, 1964). In response-dependent schedules, reinforcer delivery is contingent upon a response. In other words, a response is necessary for delivery of a reinforcer. Some types of response-dependent schedules include fixed ratio (FR), variable ratio (VR), fixed interval (FI), and variable interval (VI). In these schedules, a certain number of responses are required (ratio schedules) or at least one response is required after a specified amount of time (interval schedules) in order for a reinforcer to be delivered (cf. Ferster & Skinner, 1957). In response-independent schedules, a reinforcer is delivered at the end of a specified time interval, regardless of whether a response occurs. Thus, a response is not necessary for reinforcer delivery. Examples include autoshaping, fixed-time (FT), and variable-time (VT) schedules.
An autoshaping procedure is a response-independent schedule of reinforcement
because a known reinforcer (e.g., grain) is delivered periodically, regardless of whether or not a response occurs. Despite this lack of dependency between responding and reinforcer delivery, organisms will approach and make contact with a stimulus that immediately precedes the delivery of a reinforcer. This phenomenon is called autoshaping and was first shown in pigeons by Brown and Jenkins (1968). In their study, an 8-s key light illumination was followed by 4-s access to grain, and the intertrial interval (ITI) was on average 45 s. All pigeons pecked at the illuminated key light, even though grain deliveries were response independent. Autoshaping has been demonstrated
1


2
across several species, including fish (Squier, 1969), human children (Zeiler, 1972), dogs (Smith & Smith, 1971), chicks (Wasserman, 1973; Zolman, Chadler & Black, 1972), quail (Gardner, 1969), rats (Leslie, Boakes, Linaza, & Ridgers, 1979; Peterson, Ackil, Frommer, & Hearst, 1972), and monkeys (Sidman & Fletcher, 1968).
In drug studies, schedules of reinforcement have often been used to establish baseline responding and drug effects have been measured as deviations from steady-state responding (cf. Sidman, 1960). That is, the schedule of reinforcement is in effect until the measured behavior shows stability. Effects of drugs are then tested using the steady-state responding as a baseline. Acute dosing occurs when a drug is administered intermittently before sessions and behavior is allowed to return to baseline levels before the next administration. This type of regimen allows for investigation of how responding under drug conditions differs from responding during baseline (i.e., non-drug conditions).
Dews (1955a) was among the first to show schedules of reinforcement as important determinants of drug effects. In Dews' study, pigeons were trained to respond on a multiple FI FR schedule of reinforcement. A multiple schedule of reinforcement is one in which two or more schedules are presented sequentially, and each is signaled by a different stimulus (cf. Ferster & Skinner, 1957). In Dews' study, pigeons were required to peck once after 15 min during the FI component to earn reinforcement, and to peck 50 times during the FR component to earn reinforcement. Once responding stabilized under this schedule, the effects of a range of doses of pentobarbital were investigated. Different effects were engendered by the same doses of pentobarbital depending on whether responding was maintained by the interval or the ratio component of the multiple FI FR schedule. At a moderate dose of pentobarbital, rates of responding under the ratio schedule were increased relative to baseline rates, whereas the rates of responding under


3
the interval schedule were reduced relative to baseline. Thus, whether the drug served to stimulate or depress behavior depended on the schedule of reinforcement.
To investigate further the relationship between schedules of reinforcement and drug effects, Dews (1958, 1964) conducted a series of experiments. In one of these experiments (Dews, 1958), pigeons' key pecking was maintained on either a) an VI 1 min; b) an FR 50; c) an FI 15 min; or, d) an FR 900. Rates of responding were relatively high during the VI 1-min and FR-50 schedules, and were relatively low during the FI 15-min and FR-900 schedules. When a range of doses of methamphetamine was administered, response rates were decreased under the FI 1-min and FR-50 schedules, and were increased under the FI 15-min and FR-900 schedules. That is, whether the same dose of methamphetamine increased or decreased rates of responding depended on the baseline rate of responding. These results led Dews to conclude that an effect of a drug on responding may depend on the baseline rate. This was the original formulation of the rate-dependency hypothesis.
Rate dependency is a robust effect that has been observed under a variety of schedules (Clark & Steele, 1966; Dews, 1958; Heffner, Drawbaugh, & Zigmond, 1974; Kelleher & Morse, 1968; Leander & McMillan, 1974; McKearney, 1970; McMillan, 1968a,b, 1969; Smith, 1964; Wuttke, 1970;) and across classes of drugs (Dews 1955b; McKearney, 1970; McMillan, 1968a,b; Smith, 1964), including amphetamine (Clark & Steele, 1966; Kelleher & Morse, 1968; Leander & McMillan, 1974; Smith, 1964; Wuttke, 1970) and cocaine (Gonzalez & Goldberg; 1977; Smith, 1964; Spealman, Goldberg, Kelleher, Goldberg, & Charlton, 1977; see review by Sanger & Blackman, 1976). In these studies, low rates of responding were increased and high rates of responding were


4
decreased after drug administration. Thus, the effect of the drug depended on the baseline rate of responding maintained by a particular schedule of reinforcement.
Rate dependency describes a relationship between the control (non-drug) and drug rates of responding, which can be quantified by a linear log-log relationship. The logarithm of the drug effect is a linear function of the logarithm of the baseline rate. The slope of this linear function approaches zero as the dose of the drug increases. A slope of less than 1.0 indicates that low rates increased more or decreased less than higher rates. A slope of 1.0 indicates equal rates of responding both under drug and no drug (Dews & Wenger, 1977; Kelleher & Morse, 1968).
One way to test for rate-dependent effects of a drug is to use a schedule of reinforcement to produce different rates of responding. Response-dependent schedules have been used to investigate drug effects, and reinforcer magnitude has been shown to interact with drug effects under such schedules (Balster & Schuster, 1973; Spealman & Kelleher, 1979; Lemaire & Meisch, 1985; Meisch, 2000; Meisch & Stewart, 1995; Stewart, Wang, Bass, & Meisch, 2002; Woolverton & English, 1997). One type of response-dependent schedule that allows for investigation of relations between reinforcer magnitude and drug effects is a behavior-correlated schedule (Miller, Brodkorb, & Branch, 2001).
In behavior-correlated schedules, the amount of each reinforcer is contingent upon some aspect of the subject's performance (Dodd, 1980; Gentry & Eskew, 1984; Hendry, 1962; Hendry & Van-Toller, 1964). For example, Gentry and Eskew (1984) showed that response rates were sensitive to increased reinforcer magnitude when a direct relation was established between number of responses and duration of access to grain. In their study, pigeons were initially trained to key peck using an automaintenance


5
procedure (cf. Brown & Jenkins, 1968) in which key light illuminations were followed by access to grain. Once key pecking was reliably established, a behavior-correlated schedule was implemented in which each response emitted during the 8-s trial generated 0.25-s access to grain delivered at the end of the trial. The number of responses per trial increased relative to control (automaintenance) conditions wherein access to grain was held constant. Other studies provided support that responding is sensitive to correlated reinforcement magnitude using this type of schedule (Balsam, Brownstein, & Shull, 1978; Dodd, 1980). Balsam, Brownstein, and Shull (1978) investigated the effects of different feeder durations on automaintained responses. They found no effect of different durations as long as the same duration was used throughout the session. When different magnitudes were available in the same session, higher rates of responding were observed in the presence of the stimulus cue that preceded the longer feeder duration.
The goal of the current study was to assess the effects of cocaine on behavior under automaintenance and behavior-correlated schedules arranged to be identical to automaintenance except that food amounts at the end of the interval depended on number of responses in each interval. Although few studies have investigated the effects of amphetamine on automaintained responding (Poling & Appel, 1979; Poling & Thompson, 1977), no studies have investigated the effects of cocaine on automaintained responding. By using behavior-correlated reinforcement magnitude, rates of responding were manipulated to determine whether the effects of cocaine were rate dependent. All previous research examining rate-dependent effects of drugs have used schedules under which reinforcement depended on and was necessarily temporally contiguous with the measured responses. The current research, therefore, examined the generality of rate-dependent effects of cocaine either in circumstances where reinforcement was


6
independent of responding (i.e., under an automaintenance procedure) or under which temporal contiguity between response and reinforcer was not guaranteed.
Experiment 1 investigated the effects of behavior-correlated contingencies on automaintained responding. Experiment 2 investigated effects of cocaine on responding under each of the schedules generated during Experiment 1.


CHAPTER 2 GENERAL EXPERIMENTAL METHOD
Subjects
Six experimentally naive adult White Carneau pigeons (obtained from Double "T" Farm, Glenwood, Iowa, USA) served as subjects. Pigeons were housed individually in a temperature-controlled colony room under a 16:8 hour light/dark cycle. All pigeons had continuous access to vitamin-enriched water and health grit in their home cages, and were maintained at 80% of their ad libitum weights. These weights ranged from 383 to 487 grams.
Apparatus
A custom-built operant-conditioning chamber for pigeons was used for all sessions. Interior chamber dimensions were 30 cm x 31 cm x 34 cm, within which 3 walls and the ceiling were painted flat black, and the remaining wall was a brushed-aluminum work panel. Situated horizontally on the work panel were three 2-cm diameter response keys, located 22 cm above the wire grid floor. General illumination was provided by two 1.1-W, 28-Vdc lamps (houselights) mounted above small aluminum reflectors that deflected light toward the ceiling. These lamps were situated in the upper corners of the work panel. The center key was transilluminated by a 1.1-W, 28-Vdc lamp, whereas the side keys were dark and inoperative during all conditions. Pecks with a force of 0.1 N or greater to the center key resulted in a 30-ms operation of a relay behind the front panel and were recorded as key pecks. An aperture (through which
mixed grain could be obtained) measured 6 cm x 5 cm and was located 9.5 cm below the
7


8
center key. Grain was made available by a solenoid-operated feeder. During the grain deliveries, the aperture was illuminated by a 1.1-W, 28-Vdc bulb and all other lights (i.e., key light and houselights) were extinguished. A fan mounted on the ceiling provided continuous ventilation inside the chamber, and white noise (95 dB) was present in the experimental room. A MED Associates Single I/R Control generated an infrared beam across the opening of the aperture. This allowed measurement of the total time that the pigeon's head was in the hopper. Pigeons could be observed through a peep-hole (2 cm diameter) located in the door of the operant chamber.
Experimental contingencies and data collection were executed under the ECBasic control system (Palya & Walter, 1993) interfaced with an IBM-compatible computer located in an adjacent room. Continuous recordings of responses were obtained using a GerbrandsR Model C-3 cumulative-response recorder.
General Procedures
Sessions were conducted 7 days per week at approximately the same time every day. Each session was preceded by a 5-min blackout period in which all lights in the chamber were off and no programmed contingencies were in effect. After that, the houselights were illuminated to begin the session. Criteria for changing conditions were based on a lack of trends in session-to-session performance.
Training
Magazine training was completed within two sessions for each pigeon, at which time eating from the raised hopper occurred reliably. These sessions provided 60 4-s opportunities to eat grain from the hopper. Timing of the 4-s grain delivery began when the pigeon's head entered the hopper.


9
Drug-Administration Procedures
Cocaine hydrochloride (provided by the National Institute on Drug Abuse) was dissolved in 0.9% saline solution and injected intramuscularly in a volume of 1.0 mL/kg. A range of doses of cocaine (0.1 to 10.0 mg/kg) and saline was administered once per week before the experimental session for all pigeons. Dosages were administered in the following descending series: saline, 10.0 mg/kg, 5.6 mg/kg, 3.0 mg/kg, 1.0 mg/kg, 0.3 mg/kg, and 0.1 mg/kg, and then the series was repeated. A fixed order was used to facilitate detection of systematic differences across repeated determinations of effects of each dose (cf. Sidman, 1960). At least two determinations of each dose were assessed, except when Pigeon 922 received only one injection of 0.3 mg/kg during the rate-decreasing schedule due to experimenter error. After two iterations of the descending series, the effects of doses that revealed sufficient variability were reassessed until the effects at those doses were deemed to be accurately determined. A control session was defined as one that preceded a session during which cocaine was administered.
The details of the drug-administration procedure are described here to clarify the chronology of events in Experiments 1 and 2. Once responding was stabilized in each condition (automaintenance and behavior-correlated schedules), the effects of a range of cocaine doses were examined. Thus, conditions of Experiments 1 and 2 were interspersed, although they are described throughout the manuscript as separate experiments.


CHAPTER 3 EXPERIMENT 1
Automaintenance (Baseline)
During automaintenance, grain was presented response-independently for 4 s after an 8-s keylight illumination (cf. Brown & Jenkins, 1968). Sessions were 30 min in duration and consisted of 36 trials separated by intertrial intervals (ITIs) of 16, 32, 48, or 64 seconds. During the ITI, the key light was darkened and the houselights remained lit. The ITI values were randomly selected in blocks of 4, yielding an average duration of 40 s per block (i.e., across 4 trials). Key pecking was established within 4 sessions for each pigeon. Numbers of sessions conducted under this condition, and all conditions, are shown in Table 1. For the first 20 sessions, the session began with key light and houselight illumination, that is, commenced with a trial. Subsequent sessions began with only houselight illumination (i.e., an ITI).
Behavior-Correlated Schedule: Direct Relation between
Responding and Grain Availability
A direct relation was programmed between responding and grain availability such
that the greater the number of pecks during the 8-s key illumination compared to
baseline, the greater the amount of grain presented at the end of the trial. Hereafter, this
schedule is called the rate-increasing schedule. Response requirements were determined
by the baseline number of responses in each trial. Figure 1 shows frequency distributions
for the number of pecks that occurred during each 8-s key illumination from control
sessions during the dosing regimen (Experiment 2) during the automaintenance condition
10


11
for all pigeons except Pigeon 829. Control sessions were defined as those that occurred the day before sessions that included drug or vehicle injections. Frequency distribution
data are not shown for Pigeon 829 because out of 576 trials (16 sessions), it did not peck on 529 trials, pecked once on 45 trials, and pecked twice on 2 trials (it never pecked more than twice on a trial during this condition). The number of responses emitted under conditions of automaintenance varied widely across subjects, with number of responses per trial ranging from 0 to nearly 30. Each pigeon's distribution was divided into thirds, and the following response requirements were implemented: if the number of responses in a trial was less than the 34 percentile, between the 34 and 65 percentile, or greater than the 65 percentile of baseline rates, then the duration of grain access on that trial was 2, 4, or 8 s, respectively. All other procedural details were identical to those described for the automaintenance condition.
One goal was to produce, at least initially, a relatively equal number of 2-, 4-, and 8-s grain-access times during the behavior-correlated schedule. This was accomplished for all pigeons except for Pigeon 446, which earned 8-s grain deliveries on every trial. Therefore, this pigeon was again exposed to an automaintenance condition so that its response requirements could be re-determined. Table 1 shows the number of sessions in the second automaintenance condition. The frequency distribution shown in Figure 1 for Pigeon 446 is from its second exposure to automaintenance. These values served as the new response requirements for Pigeon 446 during a subsequent behavior-correlated schedule.


Table 1. Response requirements, grain-access times, and number of sessions in each condition for all pigeons during automaintenance and
behavior-correlated schedules. Values in the first column indicate reinforcement magnitude with corresponding response requirements shown in the subsequent columns. Automaintenance conditions are shown in columns labeled by "auto." Conditions in which behavior-correlated schedules were in effect are designated by columns labeled either "inc" for those in which a rate-increasing schedule was programmed or "dec" when a rate-decreasing schedule was in effect. Number of sessions in each condition is listed in the row labeled "sess" and number of sessions conducted under acute dosing are listed under "acute", and these totals are exclusive of drug sessions. The experimental conditions are shown in chronological order, left to right. Conditions in which ITI values and/or grain-access times were modified are designated by an asterisks (see text for details). Grain-access times are listed as "2s", "4s", or "8s" in the left column for each pigeon. When grain-access time was manipulated, the value is listed in parentheses next to the response requirement.
Conditions
Pigeon 1 2 3 4 5 6 7 8 9 10 11 12
11 auto inc dec dec dec* dec* dec auto inc ---
2s 0-4 >9 >2 >2 2(3s) >2 0-4 ---
4s >0 5-8 5-8 1 1 l(4s) 1 >o 5-7 --- -
8s >9 0-4 0 0 0(5s) 0 >8 --- ---
sess 98 74 52 23 8 17 15 51 37 --- ---
acute 181 0 0 0 0 0 61 151 107 ---
446 auto inc auto dec auto inc inc inc --- ---
2s 0 >9 0-1 0-18 0 --- --- ---
4s >o 1 >o 2-8 >0 2-3 19-22 1-19 --- --- ---
8s >2 0-1 >4 >23 >20 --- ---
sess 97 28 53 31 56 16 25 62 --- ---
acute 227 0 82 82 158 0 68 145 --- -
922 auto inc dec dec* dec* dec auto inc inc inc inc ---
2s 0 >2 >2 2(3s) >2 0 0-8 0-20 0-3
4s >o 1 1 1 l(4s) 1 >0 1 9-14 21-23 4-5 ---
8s >2 0 0 0(5s) 0 >2 >15 >24 >6 ---
sess 98 97 66 6 10 16 50 40 19 28 56 ---
acute 175 0 0 0 0 51 109 0 0 54 78


Table 1 continued
Conditions
Pigeon 1 2 3 4 5 6 7 8 9 10 11 12
198 auto inc dec auto inc inc* inc* inc* inc* inc* inc* inc
2s 0-1 >4 0 0 0(ls) 0(.5s) 0(1 s) 0(2s) 0(ls) 0
4s >o 2-3 2-3 >o 1 1
8s >4 0-1 >2 >1 l(12s) l(12s) 1 (12s) l(8s) l(8s) 2
sess 98 57 66 53 29 34 49 82 29 69 40 14
acute 191 0 140 154 0 0 0 0 0 0 0 74
4986 auto inc dec dec auto --- --- --- mmm ---
2s 0-5 >12 >8 --- --- --- ---
4s >o 6-11 6-11 3-7 >o --- --- --- --- --- --- ---
8s >12 0-5 0-2 --- --- --- --- ---
sess 97 61 44 25 103 --- --- --- ---
acute 224 0 64 83 169 --- --- ---
829 auto inc inc inc* auto* --- --- mm ---
2s 0 0 0 --- --- --- --- ---
4s >o 1 --- --- --- --- -
8s >2 >1 >1 >o --- --- --- --- --- ---
sess 98 32 20 13 11 --- --- --- ---
acute 189 0 0 0 0 --- ---


14
Figure 1. Frequency distributions of number of responses per 8-s key illumination during
the initial automaintenance condition (baseline). These data are from control sessions during acute cocaine administration. Each graph shows data for one subject. Note that the axes differ for each subject. Data for Pigeon 829 are not shown (see text for details).


15
50 i 40 30 ] 20 -10 -0
4986
140 -, 120 -100 -
80
60
40 -j
20 -0
40
o 30 H
Q) 20 -
10 -
0
JZL
n
n
l~l n n
n
11
922
nnn
n n n
80 60 -I 40 20 -I
0
an
446
171 r-i l~l r-i n ?_nnn
400
300
200 -\ 100 -
0
198
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Number of Pecks


16
Additional Manipulations
Additional manipulations were conducted for 2 pigeons to determine if response rates could be increased by manipulation of response requirements, reinforcement magnitude, or both. The manipulations conducted for Pigeons 829 and 198 occurred when response rates were low despite a direct relation between response requirements and grain availability.
For Pigeon 829, response requirements were such that 0 or 1 (or more) responses in the trial would produce either 2- or 8-s access to grain, respectively, at the end of that trial for 20 sessions (Table 1, column 3). In the next condition, these response requirements were held constant and the ITIs were changed from 16, 32, 48, and 64 s to 32, 64, 96, and 128 s. To hold constant overall session duration, the number of trials was reduced from 36 to 16; this condition lasted for 13 sessions (Table 1, column 4). Then, a condition was implemented in which no pecks were required to earn 8-s access to grain. The ITI values and number of trials per session was the same as in the previous condition, and this lasted for 11 sessions (Table 1, column 5). At this point, experimentation was terminated for Pigeon 829 because responding was not maintained under any of these conditions.
For Pigeon 198, grain-access times in the correlated schedule were systematically varied such that 0 or 1 (or more) responses produced, respectively a) 2- or 8-s access for 34 sessions; b) 1- or 12-s access for 49 sessions; c) 0.5- or 12-s access for 82 sessions; d) 1- or 12-s access for 29 sessions; e) 2- or 8-s access for 23 sessions; and, f) 1- or 8-s access for 40 sessions (Table 1, columns 5-11). These manipulations will be referred to as a 'two-valued procedure' hereafter. Then, response requirements and grain-access times were modified so that 0, 1, or 2 (or greater) responses would produce 2-, 4-, or 8-s


17
access to grain, respectively (Table 1, column 12). After 14 sessions of this last procedure, acute dosing was implemented and will be discussed in Experiment 2.
Behavior-Correlated Schedule: Inverse Relation between
Responding and Grain Availability
For five pigeons (Pigeons 11, 446, 922, 198, and 4986), a behavior-correlated schedule was implemented with an inverse relation between the number of responses that occurred on each trial and grain availability (hereafter called a 'rate-decreasing schedule'). In this condition, the greater the number of pecks during the 8-s key illumination compared to baseline, the less grain that was delivered at the end of that trial. Response requirements were the inverse of what they were during the behavior-correlated schedule with a direct relation for Pigeons 11, 922, 198, and 4986. Response requirements for Pigeon 446 were determined from its second exposure to the automaintenance condition. For all pigeons, if the number of responses was less than the 34th percentile, between 34 and 65th percentile, or greater than 65th percentile of automaintenance (baseline) rates, then grain delivery was 8, 4, or 2 seconds, respectively. All other procedural details were the same as previously described for automaintenance condition.
Additional Manipulations
Pigeons 11 and 922 earned almost all 8-s grain-access times during their initial exposure to a rate-decreasing schedule. Therefore, additional manipulations were conducted for these pigeons to determine if a relatively equal number of grain-access times could be generated by modification of response requirements, grain availability, and/or ITI values.


18
During the first exposure to a rate-decreasing schedule for Pigeon 11, the response requirements were such that 9 (or more), 5-8, and 0-4 responses produced 2-, 4-, and 8-s access to grain, respectively, for 52 sessions (Table 1, column 3). A frequency distribution was determined from the last 10 sessions of this condition. Response requirements were changed such that 2 (or more), 1, and 0 responses would produce 2-, 4-, and 8-s access to grain, respectively, for 23 sessions (Table 1, column 4). These response requirements were also in effect for Pigeon 922 during the rate-decreasing schedule, which lasted for 66 sessions for this pigeon (Table 1, column 3). Then, ITI values were increased from 16, 32, 48, and 64 s to 24, 48, 72, and 96 s for 8 and 6 sessions for Pigeons 11 and 922, respectively. Response requirements remained the same (Table 1, column 5 for Pigeon 11, column 4 for Pigeon 922). Then, grain-access time was changed from 2, 4, and 8 seconds to 3, 4, and 5 seconds for 17 sessions for Pigeon 11 and for 10 sessions for Pigeon 922 (Table 1, column 6 for Pigeon 11, column 5 for Pigeon 922). Response requirements for Pigeons 11 and 922 were then changed such that 2 (or more), 1 or 0 responses produced 2-, 4-, and 8-s access to grain, respectively. The ITI values were also changed from 24, 48, 72, and 96 s back to the original ITI values of 16, 32, 48, and 64 s. This condition was implemented for 17 sessions for Pigeon 11 and 16 sessions for Pigeon 922 (Table 1, column 7 for Pigeon 11, column 6 for Pigeon 922). Acute dosing was then implemented using these response requirements, and will be discussed in more detail in Experiment 2.
The response requirements for Pigeon 4986 during its first exposure to a rate-decreasing schedule were such that 12 (or more), 6-11, and 0-5 pecks produced 2-, 4-, and 8-s access to grain, respectively (Table 1, column 3). After 44 sessions, an acute-dosing regimen was implemented. During this acute-dosing regimen, however, there was


19
a shift in the number of 8-s grain deliveries earned in a session. Therefore, response requirements were re-determined from control sessions during this acute-dosing regimen. Response requirements were such that 8 (or more), 3-7, and 0-2 pecks produced 2-, 4-, and 8-s access to grain, respectively (Table 1, column 4).


CHAPTER 4 RESULTS: EXPERIMENT 1
Autoshaping (Acquisition)
Figure 2 shows responses per minute during periods when the key was illuminated during each of the first 40 sessions during autoshaping (acquisition). For 5 of 6 pigeons (Pigeons 4986, 11, 922, 198, and 829), response rate was zero in the first session and response rate for Pigeon 446 was 0.21. Overall mean rates of responding during the first 40 sessions ranged from 3.28 (Pigeon 829) to 66.99 (Pigeon 198) with a median of 27.01 across all pigeons.
Table 2 shows that the mean number of responses in a trial did not vary systematically as a function of duration of the immediately preceding ITI during automaintenance (baseline) for individual pigeons.
Table 2. Mean number of responses per trial as a function of preceding intertrial interval
(seconds) during automaintenance. Data are from the last 10 sessions of automaintenance before the initial acute-dosing regimen.
Intertrial Interval (ITI)
Pigeon 16s 32s 48s 64s
4986 9.94 10.87 10.10 9.91
11 6.17 5.09 4.50 4.75
829 0.00 0.02 0.09 0.03
922 1.60 1.61 1.47 0.98
446 1.48 1.01 1.38 1.16
198 3.93 3.09 2.24 2.33
means: 3.85 3.62 3.30 3.19
20


21
Figure 2. Responses per minute across the first 40 successive sessions during the initial
automaintenance condition (baseline) for all pigeons. Note that the y-axis varies for each subject.


22
CD
30 25 20 \ 15 10
5
0
120 100
80 \
60
40
20
0
4986
11
CD
to
CD
O
CL
CD
CD
O
80 60 40
20 1
0 20
15 10 5
0
140 120 100 80 60 40 20 0
12 10
446
8
6 4 2 | 0
0
829
5
10
15
20
25
30
35
40
Sessions during Autoshaping (Baseline)


23
Behavior-Correlated Schedules
Table 3 shows mean overall responses per minute across every exposure to the automaintenance, rate-increasing and rate-decreasing schedules. Data are from the last 10 sessions of each condition. Pigeons 11, 922, 198, and 4986 were exposed to automaintenance twice, Pigeon 446 was exposed to automaintenance three times, and Pigeon 829 was exposed to automaintenance once. Values in Table 3 are the means from every exposure to the automaintenance, rate-increasing and rate-decreasing schedules, respectively. However, data were not included when 'additional manipulations' were conducted during any schedule, such as when ITI values were changed or a two-valued procedure was used. These 'additional manipulations' are designated by asterisks next to the condition label in Table 1.
These results showed that a) automaintenance generated low-to-moderate responses per minute across all pigeons (range: 0.27 to 59.73); b) rates of responding were increased relative to automaintenance for all pigeons (except Pigeon 829) when a rate-increasing schedule was implemented (range: 0.05 to 110.01); and, c) rates of responding decreased relative to baseline for all 5 pigeons exposed to a rate-decreasing schedule (range: 1.29 to 28.83).
Table 3. Mean overall responses per minute across conditions of automaintenance, rate-increasing, and rate-decreasing schedules. Data are from the last 10 sessions of each condition, and were averaged across all exposures to each condition, respectively.
Pigeon Auto Increasing Decreasing
11 38.12 59.04 3.46
446 59.73 110.01 12.94
922 5.48 103.09 1.29
198 15.29 30.98 14.93
4986 54.27 81.91 28.83
829 0.27 0.05 mm MM


24
Behavior-Correlated Schedule: Direct Relation Between
Responding and Grain Availability
Figure 3 shows the number of trials (out of 36) in which 2-, 4-, and 8-s grain deliveries were earned across successive sessions during the rate-increasing schedule. This provides only an indirect behavioral measure because the y-axis shows the number of trials in which a particular reinforcer magnitude was earned rather than a direct measure of responding. Describing these data in this manner, however, is more straightforward than labeling the y-axis as the number of small, medium, and large response bins that occurred during a session, where response bins would be defined by the response requirements that were determined by the frequency distributions.
One would predict a relatively equal number of 2-, 4-, and 8-s grain deliveries
initially because response requirements were based on a frequency distribution that was divided into thirds. This prediction was true for Pigeons 4986, 11, and 198. It was not the case for Pigeons 922 and 446. For Pigeon 922, the frequency distribution was composed of unequal response values during automaintenance. The number of responses per trial for this pigeon ranged from 0 to 2. Mostly, this pigeon did not peck or if it did, it pecked only once per trial during automaintenance. By the end of this condition, Pigeon 922 was pecking at least twice per trial and earned 8-s grain deliveries on almost every trial in a session. Pigeon 446 earned almost all 8-s grain deliveries and few, if any, 2-and 4-s grain deliveries on each trial during its initial exposure to a rate-increasing schedule. Additional manipulations were then conducted for this pigeon.
When response-dependent contingencies were superimposed onto an automaintenance schedule such that more responding produced more grain (direct relation), the number of 8-s grain deliveries was increased relative to the number of 2-


25
and 4-s grain deliveries for Pigeons 922 and 198. The number of 8-s grain deliveries increased relative to 2-s grain deliveries for Pigeons 4986 and 11. For these two pigeons, the number of 4- and 8-s grain deliveries was about equal by the end of the condition. These findings show that responding was sensitive to increased reinforcer magnitude for 4 of 5 pigeons.


26
Figure 3. Number of trials in which 2-, 4-, and 8-s access to grain was earned as a
function of successive sessions during the rate-increasing schedule. Filled and open circles represent 2- and 4-s access; squares represent 8-s access to grain. Response requirements are listed for each pigeon.


27
40 35 30 25 20 15 10
5
0
35
30 j
25 -
20
15
10 5 \ 0
4986
0 2s
1 4s >2 8s
0-5 2s
6-11 >12
4s 8s
40 35 30 25 20 15 10 5 0
0-4 5-8 >9
2s 4s 8s
40 35 30 25 20 15 10 5 0
. 922
c 40 35 30 25 20 15 10 5 0
2 s 4 s 8s
0-1 2-3 >4
2s 4s 8s
0
20
40
60
80
Successive Sessions


28
Replication of Behavior-Correlated Schedule with a Direct Relation
Figure 4 shows the results of a second exposure to a rate-increasing schedule for Pigeons 11, 922, 198, and 446. Pigeon 829 was removed from the experiment by the time this condition was implemented.
By the end of this condition, the number of 8-s grain deliveries increased from an average of 5 to 33 per session for Pigeons 446 and 922, and increased from 5 to 17 for Pigeon 11. An unequal distribution of 2-, 4-, and 8-s hopper durations was observed for Pigeon 922 when the schedule was initially in effect. This may be attributed to the frequency distribution that was composed of unequal response values of 0, 1, and 2 (or more). That is, many trials with no responses occurred during the initial exposure to automaintenance, whereas 1 or 2 responses were emitted on a few trials during this condition.
Pigeon 198 earned 2-s grain deliveries on almost every trial across the entire condition; 4-s grain deliveries were earned approximately 4 times per session, and 8-s grain deliveries were earned on only two occasions during this condition. This implies responding was insensitive to increased reinforcer magnitudes for this pigeon. Additional manipulations were conducted for this pigeon, and the results are discussed in the next section.


29
Figure 4. Number of trials in which 2-, 4-, and 8-s hopper durations were earned across
successive sessions during a return to a rate-increasing schedule for Pigeons 11, 922, 198, and 446. Response requirements are listed for each pigeon.


30
40 35 30 j 25 20 15 10 \
5
0
11
0-4 2s 5-7 4s >8 8s
40 35 30 25 20 15 10 5 0
0 10 20 30 40
Successive Sessions


31
Additional Manipulations
Figure 5 shows the number of 2-, 4-, and 8-s grain deliveries across successive sessions during additional manipulations for Pigeons 922 and 446 during the rate-increasing correlated schedule. These additional manipulations were conducted for several reasons. Pigeon 922 earned 8-s access to grain on almost every trial during the rate-increasing schedule (left panel for Pigeon 922). The rate-increasing schedule generated a high rate of responding for this pigeon (157.37 responses per minute) with at least 15 pecks on almost every trial. Because we wanted to generate a relatively equal number of 2-, 4-, and 8-s grain-access times before assessing the acute effects of cocaine, we reassessed the frequency distribution based on the performance of Pigeon 922 during this rate-increasing schedule. From this frequency distribution, response requirements were re-determined and the number of 2-, 4-, and 8-s grain durations earned during this condition are shown in the middle panel for Pigeon 922. These results show that the number of 2-, 4-, and 8-s grain durations were approximately equal. At this point, an acute-dosing regimen was implemented (details will be discussed in Experiment 2). Rate of responding for Pigeon 922 at the beginning of the cocaine administration was 160.73 responses per minute. However, after exposure to the first cycle of doses, rate of responding decreased to 73.19 responses per minute during non-drug (control) sessions. Because of this baseline shift, response requirements were again re-determined for Pigeon 922. The frequency distribution from which these values were obtained was determined by rates of responding during the last 5 control days. The number of 2-, 4-, and 8-s grain deliveries earned with these re-determined response requirements are shown in the right panel of Figure 5 for Pigeon 922.


32
The data in the left panel for Pigeon 446 show that the number of 2-, 4-, and 8-s grain deliveries was approximately equal. An acute-dosing regimen was then implemented. After one cycle of doses, however, a baseline shift occurred. That is, rate of responding for Pigeon 446 before cocaine administration was 132.43 responses per minute and decreased to 78.35 responses per minute during non-drug (control) sessions during cocaine administration. The response requirements were re-determined for Pigeon 446 in a manner previously described for Pigeon 922. The data in the right panel for Pigeon 446 show the results of these response requirements on the number of 2-, 4-, and 8-s grain deliveries. For this pigeon, the number of 2-s grain deliveries was highest and 8-s grain deliveries were lowest during the first 23 sessions of this condition. Gradually, the number of 2-s durations decreased, 8-s durations remained the same, and 4-s durations increased bv the end of this condition.


33
Figure 5. Number of trials in which 2-, 4-, and 8-s durations were earned during a rate-increasing schedule for Pigeons 922 and 446.


34


35
Additional manipulations also were conducted for Pigeon 198 during a behavior-correlated schedule with direct relations in an attempt to bolster response rates. Because response rates were low for Pigeon 198 (less than 1 response per minute), a two-value procedure was implemented (Table 1, columns 6-11). That is, two different amounts of grain were available upon completion of the interval, as opposed to the three-value procedure (e.g., 2-, 4-, and 8-s grain access times) used previously.
Figure 6 shows that the amount of grain earned for Pigeon 198 increased systematically as the ratio of small-to-large grain amount decreased. In the first manipulation for this pigeon, response requirements were such that no responses produced 2-s access to grain and at least 1 response produced 8-s access to grain. The data in the first panel of Figure 6 show the numbers of 2- and 8-s grain deliveries earned under these response requirements; Pigeon 198 earned almost all 2-s and few 8-s grain deliveries across 34 sessions. Next, at least 1 response during the trial produced 12-s access to grain and no responses produced 1-s access. The data in the second panel show that the number of 12-s grain deliveries was less than the number of 1-s grain deliveries for 49 sessions. When grain amount was changed in a subsequent condition such that no responding would produce 0.5-s access to grain and at least 1 response would produce 12-s access to grain, the number of 0.5-s durations decreased and 12-s durations increased across 82 sessions (shown in the third panel from the right). Data in the fourth panel show that when grain amount was changed such that no responses again produced 1-s access to grain, and at least 1 response produced 12-s access to grain, the number of 12-s durations earned was greater than the number of 1-s durations across the entire condition (29 sessions). This condition was identical to the one that produced the data in the second panel but the effect was not the same. Next, no responses produced 2-s access


36
to grain and at least 1 response produced 8-s access to grain. The number of 8-s durations were initially higher than the 2-s durations, and then decreased relative to the 2-s durations across 69 sessions (shown in the fifth panel). These response requirements were the same as those that produced the data in the first panel. The data in the last panel show the effects of a grain-amount manipulation when no responses produced 1 -s access and at least 1 response produced 8-s access to grain. The number of 8-s durations was initially less than the number of 1 -s durations, and increased relative to the number of 1 -s durations across this condition, which lasted for 40 sessions.


37
Figure 6. Number of trials in which 2-, 4-, and 8-s hopper durations were earned during a
rate-increasing schedule for Pigeon 198. Response requirements for each manipulation are shown above each panel of data.


38
0 0.5s 0 1s 0 2s 0 1s
>1 12s >1 12s >1 8s >1 8s
100 150 200 250
Successive Sessions
short grain-access times (sec) long grain-access times (sec)


39
Behavior-Correlated Schedule: Inverse Relation between
Responding and Grain Availability
Figure 7 shows that the number of 8-s grain deliveries increased across successive
sessions during the first exposure to a rate-decreasing schedule for Pigeons 4986, 11, 922, 446, and 198. Pigeon 829 had been removed from the experiment by the time this condition was implemented and, therefore, was not exposed to a rate-decreasing schedule. For Pigeon 446, the number of 8-s grain deliveries remained stable across successive sessions, and was somewhat higher than the number of 4-s grain deliveries. For all 5 pigeons, the number of 2-s grain deliveries was reduced relative to the number of 2-s grain deliveries earned at the beginning of the condition. These results show that responding was sensitive to reinforcement magnitude when a contingency was established between a low response requirements and high reinforcement magnitude.


40
Figure 7. Number of trials in which 2-, 4-, or 8-s access to grain was earned as a function
of successive sessions for Pigeons 4986, 11, 922, 446, and 198. Filled and open circles represent 2- and 4-s access; squares represent 8-s access to grain. Data are from sessions in which a behavior-correlated schedule (i.e., rate-decreasing schedule) was in effect.


41
>9
2s
5-8 4s 0-4 8s
40 ,922
40 35 30 25 20 15 10 5 0
0
20
40
60
80
Successive Sessions


42
Table 4 shows the mean overall mean latency to first key peck during automaintenance (baseline), reversal to automaintenance, rate-increasing and rate-decreasing correlated schedules. Data from the initial exposure to automaintenance are shown in the first column under 'Auto 1' for all pigeons. Data from the second exposure (third exposure for Pigeon 446) to automaintenance are shown in the second column labeled 'Auto 2'. The automaintenance data (Table 6, columns 1 and 2) were averaged and are shown in the'third column labeled 'Auto 1/2'. The criteria for data selection were stability in overall response rates and number of 2-, 4-, and 8-s grain-access times, and only data from three-valued procedures were used. Data from all conditions in which additional manipulations were conducted were excluded from this analysis, such as when ITI values were changed, reinforcer magnitude was changed, or a two-valued procedure was used. Therefore, data from the rate-increasing schedules were from the following exposures: third exposure for Pigeon 198 (Table 1, column 12), first exposure for Pigeons 4986 (Table 1, column 2) and 829 (Table 1, column 2), second exposure for Pigeon 11 (Table 1, column 9), fourth exposure for Pigeon 446 (Table 1, column 8), and fifth exposure for Pigeon 922 (Table 1, column 11). Data from the rate-decreasing schedules were from the following exposures: first exposure for Pigeons 446 (Table 1, column 4) and 198 (Table 1, column 3), second exposure for Pigeons 4986 (Table 1, column 4) and 922 (Table 1, column 6), and third exposure for Pigeon 11 (Table 1, column 7). Pigeon 829 was not exposed to a second automaintenance condition or a rate-decreasing schedule.
Mean overall latency to the first key peck showed no systematic changes from the initial automaintenance condition to a return to automaintenance across pigeons. For all 5 pigeons that were exposed to the rate-decreasing schedule the overall mean latency to


43
first peck decreased relative to baseline during the rate-decreasing schedule. For 4 of the
5 pigeons exposed to the rate-increasing schedule, the overall mean latency to first peck
increased relative to baseline during the rate-increasing schedule. Further, the finding
that the latency to first peck was increased while rates of responding increased during the
rate-increasing schedule (Table 3) suggests that the run rate was increased relative to
those during the initial automaintenance condition for all pigeons. Run rate is defined as
the number of responses during the 8-s key light illumination divided by the duration of
the response bout (i.e., latency of the last response minus the latency of the first
response). The opposite was true for performance under the rate-decreasing schedule:
latency to first key peck was decreased while rates of responding were reduced relative to
automaintenance (baseline), which suggests that run rates decreased. These data
emphasize that changes occurred in the temporal organization of behavior.
Table 4. Mean overall latency in seconds to first key peck on each trial during the
initial automaintenance condition (Auto 1), return to automaintenance (Auto 2), combined automaintenance data (Auto 1/2), rate-increasing (Inc) and rate-decreasing (Dec) correlated contingencies. Data are from the last 5 control days (sessions preceding cocaine administration) in which a response occurred. For Pigeons 4986 and 829, cocaine was not administered during the rate-increasing schedule so the data shown are those from the last 5 sessions in which a response occurred during the rate-increasing schedule.
Pigeon Auto 1 Auto 2 Auto 112 Inc Dec
4986 1.46 1.21 1.34 1.06 1.14
11 1.41 1.72 1.57 1.65 0.93
446 0.89 0.77 0.83 1.21 0.74
922 1.65 1.92 1.79 2.26 1.45
198 1.93 1.49 1.71 2.11 1.65
829 0.35 --- 0.35 0.24 ---
Table 5 shows the mean overall latency in seconds from the last key peck during the 8-s key illumination to grain delivery for automaintenance (Autol), return to automaintenance (Auto2), combined automaintenance data (Auto 1/2), rate-increasing


44
(Inc) and rate-decreasing (Dec) schedules for Pigeons 4986, 11, 446, 922, and 198. Data were from the same conditions described for Table 5. Table 6 shows that the mean peck-to-grain latency was a) similar during all automaintenance conditions, with a group mean of 4.43 seconds; b) was lowest during the rate-increasing schedule (group mean 3.19 seconds); and c) was greatest during the rate-decreasing schedule (group mean of 6.14 seconds).
Table 5. Mean overall time in seconds from last key peck during 8-s key illumination
to grain delivery for automaintenance (Autol), return to automaintenance (Auto2), combined automaintenance data (Auto 1/2), rate-increasing (Inc) and rate-decreasing (Dec) schedules for Pigeons 4986, 11, 446, 922, and 198. Data are from the last 5 control days (sessions preceding cocaine administration) in which a response occurred. For Pigeons 4986 and 829, cocaine was not administered during the rate-increasing schedule so the data shown are those from the last 5 sessions in which a response occurred during the rate-increasing schedule. Pigeon 829 was not exposed to a second automaintenance condition or a rate-decreasing schedule.
Pigeon Autol Auto2 Auto 112 Inc Dec
4986 3.26 3.32 3.29 2.64 5.12
11 0.92 1.20 1.06 0.60 7.14
446 3.92 5.06 4.49 0.42 6.90
922 5.90 6.06 5.98 1.76 5.84
198 4.74 6.58 5.66 5.86 5.70
829 7.81 7.81 7.88 ----
means: 4.43 4.44 4.43 3.19 6.14
Additional Manipulations
Additional manipulations were conducted for Pigeons 11, 922, and 4986 during their initial exposure to a rate-decreasing schedule because they initially earned 8-s grain deliveries on most trials, and our goal was to produce an approximately equal number of grain-access times. Figure 8 shows the number of trials in which 2-, 4-, or 8-s access to grain was earned as a function of successive sessions for these pigeons.
During each modification of the rate-decreasing schedule, Pigeon 11 earned mostly 8-s grain deliveries and approximately equal number of 2- and 4-s grain


45
deliveries. Data in the first panel of Figure 8 are from the condition in which 2 (or more), 1, or 0 pecks produced 2-, 4-, or 8-s access to grain, respectively, for 23 sessions (Table 1, column 4). Pigeon 11 did not peck on most trials and, therefore, earned mostly 8-s access to grain and relatively few 2- and 4-s grain deliveries during each session of this condition. Data in the second panel show the results of when ITI values were changed
for 8 sessions from 16, 32, 48, and 64 s to 24, 48, 72, and 96 s, and response requirements were the same as in previous condition (Table 1, column 5). Rate of responding was not influenced by the change in ITI values for Pigeon 11, as evidenced by approximately the same number of 2-, 4-, and 8-s earned grained deliveries as in the first panel. Data in the third panel show the results of changing response requirements such that 2 (or more), 1, or 0 responses produced access to grain for 3, 4, and 5 s, respectively, for 17 sessions (Table 1, column 6). The number of 8-s grain deliveries was slightly lower and the number of 4-s grain deliveries was slightly higher than in the previous panel; the number of 2-s grain deliveries was the same. These data suggest that there was a marginal increase in response rate for Pigeon 11 relative to the previous manipulations (i.e., panels 1 and 2). Data in the fourth panel show the results of changing response requirements such that 2 (or more), 1, and 0 pecks produced 2-, 4-, or 8-s access to grain, and the original ITI values of 13, 32, 48, and 64 s were reinstated for 15 sessions (Table 1, column 7). The number of 8-s grain deliveries was slightly higher, and number of 2- and 4-s grain deliveries was slightly lower than those shown in the third panel. The data in panel four are more similar to those shown in panels one and two, which suggests that rates of responding for Pigeon 11 under these response requirements were approximately the same.


46
For Pigeon 922, the number of earned grain deliveries was highest for the 8-s durations, and about the same for 4- and 2-s grain deliveries across all modifications during the rate-decreasing schedule. Data in the first panel show the effect of changing the ITI values from 16, 32, 48, and 64 s to 24, 48, 72, and 96 s, while the response requirements were such that 2 (or more), 1, or 0 responses produced 2-, 4-, and 8-s access to grain, respectively, for 6 sessions (Table 1, column 4). Modification of the ITI values did not influence the number of 2-, 4- and 8-s grain deliveries earned by Pigeon 922 (see Figure 7). Data in the second panel for Pigeon 922 show the results of changing grain amounts so that 2 (or more), 1, or 0 responses produced 3-, 4-, or 5-s access to grain. ITI values remained at 24, 48, 72, and 96 s (Table 1, column 5). There was a decrease in the number of 8-s grain deliveries and an increase in the number of 2- and 4-s grain deliveries earned by the end of this condition relative to those in the beginning of this condition. This suggests that response rate was somewhat higher at the end of the condition than it was at the beginning for Pigeon 922. Data in the third panel show the effects of modifying grain amount such that 2 (or more), 1, or 0 responses produced 2-, 4-and 8-s grain deliveries, and ITI values were changed back to 16, 32, 48, and 64 s for 16 sessions. The number of 8-s grain deliveries increased and 2- and 4-s grain deliveries decreased across sessions in this condition relative to those at the start of the condition. These data suggest that rates of responding were lower at the end of the condition than they were at the start of the condition. Acute dosing was then implemented, and the details will be discussed in Experiment 2.
The first panel of Figure 8 shows that Pigeon 4986 earned more 8-s grain deliveries than 2- and 4-s grain deliveries when its response requirements were such that 12 (or more), 6-11, and 0-5 responses produced 2-, 4-, and 8-s access to grain,


47
respectively, for 44 sessions (Table 1, column 3). Despite the unequal number of 2-, 4-, and 8-s grain deliveries earned in a session, acute dosing was then implemented. However, after one cycle of acute doses, the number of 8-s grain deliveries increased from approximately 20 prior to acute dosing to approximately 30 per session during non-drug sessions during acute dosing (second panel). The number of 4-s durations decreased from approximately 14 per session to approximately 5 per session during acute dosing. The number of 2-s grain deliveries remained about the same. Response requirements were therefore re-determined, and these requirements were such that 8 (or more), 3-7, and 0-2 pecks would produce 2-, 4-, and 8-s access to grain, respectively (Table 1, column 4). Acute dosing began after 25 sessions with these requirements, and the details will be discussed in Experiment 2.


48
Figure 8. Number of trials in which 2-, 4-, or 8-s access to grain was earned as a function
of successive sessions for Pigeons 11, 922, and 4986. Filled and open circles represent 2- and 4-s access; squares represent 8-s access to grain. Data are from sessions in which a behavior-correlated schedule with inverse relations (i.e., rate-decreasing contingency) was in effect.


49
40 i 11
35
30
25 1 J
20 Bl
15
10
5
0
2 3s
1 4s
0 5s
s}r-
a] T*^ [a

40 35 30 25 20 15 10 5 0
1 4s 0 8s
2 3s 1 4s 0 5s
2 2s 1 4s 0 8s
40 35 30 25 20 15 10 5 0
4986
12up 2s 6-11 4s
12up 2s 6-11 4s 0-5 8s
Successive Sessions


CHAPTER 5 EXPERIMENT 2
Acute effects of cocaine were examined under each of the 3 types of schedules generated in Experiment 1 (automaintenance, rate-increasing and rate-decreasing schedules). Table 6 shows the number of administrations of each dose in each of the three conditions (automaintenance, rate-increasing and rate-decreasing schedules), and
Table 1 shows under which conditions drug tests were conducted and the number of sessions held during the period of drug testing.
Table 6. Number of administrations of each dose in each of the three conditions
(automaintenance, rate-increasing and rate-decreasing schedules). Pigeon and dose of cocaine (mg/kg) are shown in the first two columns. The first exposure to automaintenance is shown by 4 Auto' and the second exposure shown by cAuto2'; data for the behavior-correlated schedules are shown by 'Inc' for rate-increasing schedule and 'Dec' for rate decreasing schedule. For Pigeon 446, the data from its third exposure to automaintenance are shown in parenthesis under the column labeled 4Auto2.' For Pigeon 922, data from its first exposure to a rate-increasing schedule were omitted and the data shown in the column labeled 'Inc' are from its second exposure to a rate-increasing schedule. See text for details.
Dose Auto Auto2 Inc Dec
Pigeon (mg/kg)
11 c 15 17 12 12
s 2 2 2 2
0.3 2 2 2 2
1.0 2 2 2 2
3.0 2 3 2 2
5.6 2 2 2 2
7.4 2 2 0 0
10.0 3 3 2 2
50


51
Table 6 continued.
Dose Auto Auto2 Inc Dec
Pigeon (mg/kg)
446 c 15 14(14) 15 15
2(2) 2 2
2(2) 2 2
2(2) 2 2
2(2) 2 2
0(0) 0 0
2(2) 3 2
2(2) 2 3
2(2) 2 2
c 15
s 2
0.1 0
0.3 2
1.0 2
4.2 3
3.0 2
5.6 2
10.0 2
c 14
s 2
0.3 2
1.0 3
3.0 3
5.6 2
10.0 2
922 c 14 12 12 12
2 2 2
2 2 1
2 2 2
2 2 2
2 2 3
2 2 2
198 c 18 17 15 18
s 2 3 2 3
0.3 4 4 2 2
1.0 2 2 2 2
3.0 3 2 2 3
4.2 0 0 0 2
5.6 2 2 2 4
7.4 2 2 J) 0
10.0 3 2 2 2
4986 c 14 12 14
s 2 2 2
0.3 2 2 2
1.0 2 2 2
1.7 2 2 2
3.0 2 2 3
5.6 2 2 3
10.0 2 0 0
829 c 16 -
s 4 -
0.3 4 -
1.0 2 -
3.0 2 -
5.6 2 -
10.0 2


CHAPTER 6 RESULTS: EXPERIMENT 2
Automaintenance (Baseline)
Figure 9 shows responses per minute as a function of dose of cocaine (range 0.1 to 10.0 mg/kg) for all 6 pigeons during the initial automaintenance condition. The number of administrations of each dose and saline are listed in Table 6. There was no overall consistent pattern in drug effects on this measure across pigeons, but large doses decreased response rates for all subjects. Dose-response curves were inverted U-shaped functions for Pigeons 11, 922, and 198. Generally, rates of responding under small doses were similar to control rates.
Figure 10 shows normalized dose-response curves (percent of saline response rates as a function of dose of cocaine) obtained during the first and second exposures to automaintenance. Data from all three exposures to automaintenance are shown for Pigeon 446. Response rates were normalized by dividing the mean rate of responding at each dose (and saline) by the mean rate of responding during saline administration for each pigeon. For Pigeons 922 and 198, dose-response functions were shifted to the right with respect to the initial assessment of doses. For Pigeons 11, 446, and 4986, dose-response functions were shifted to the left with respect to the initial assessment of doses.
Figure 11 shows mean head-in-hopper times during automaintenance (baseline)
and a return to automaintenance (third exposure to automaintenance for Pigeon 446)
across a range of doses of cocaine (0.1 to 10.0 mg/kg) and saline for all pigeons. Mean
head-in-hopper time decreased as dose of cocaine increased. Low rates of responding
52


53
were correlated with low head-in-hopper times during both automaintenance conditions for Pigeons 446, 922, 198, and 829, and during the initial automaintenance condition for Pigeons 11 and 4986 (Figure 9). During a return to automaintenance for Pigeons 11 and 4986, head-in-hopper functions were elevated compared to dose-response functions (Figure 10) at all doses for Pigeon 11 and at all doses except 5.6 mg/kg for Pigeon 4986. This means that head-in-hopper times were insensitive to response-rate decreases produced by a range of doses of cocaine, at least for Pigeons 11 and 4986.


54
Figure 9. Responses per minute as a function of dose of cocaine (0.1 to 10.0 mg/kg) for
all pigeons during the initial automaintenance condition (baseline). Open circles show the values for each administration of cocaine or saline, and filled circles show the mean of those determinations. Points above "C" are from the session before cocaine or saline administration; points above "S" are from sessions in which saline (vehicle) was administered. Note that the y-axis differs for Pigeons 446 and 829.


55
120 i 11
100
80
60 9
40
20 o
0 -VA
CS 0.3
20
446
15
10
5
0
1.0
3.0 5.6 10.0
CS
0.3
1.0
O
3.0 5.6
Q.
(D
O
(/) CD
DC
0) 120 1922
100 80 60 40 20 0
C S
0.3
120 100 80 60 40 20
0
1.0
3.0 5.6 10.0
198
o
I
T 1
c s
0.3
1.0
3.0 5.6 10.0
120 14986
100
80 60 40
20 O
0
O
8 8
e0
c s
0.3
5 i829
4
3
2
1
0
1.0 1.7 3.0 5.6 10.0
C S
o
0.3
1.0
O Probes Means
3.0 5.6 10.0
Cocaine (mg/kg)


56
Figure 10. Normalized dose-response functions (0.1 to 10.0 mg/kg) during
automaintenance (baseline) and return to automaintenance. Baseline determinations are shown by filled circles and return to automaintenance is shown by open triangles. The data for Pigeon 446 are from its first, second, and third exposures to automaintenance.


57
400
300
200
100
11
4
o
CO
CD
C S 0.3
CD
a:
CO
400
o
Q.
CD
EC
CD
c
CO
300
200
100
922
o
CD O
0
A
1-r
C S 0.3
- 200
150
100
4986
A
50
0
C S 0.3
1
1
200
150
100
50
0
446
A
3 5.6 10
C S 0.1
0.3
1
400
300
200
100
0
198
A
A
T-17
3 5.6 10
C S 0.3
1
3 5.6 10
3 5.6 10
Cocaine (mg/kg)
Automaintenance Automaintenance 2 Automaintenance 3
1 1.7 3 5.6 10
Cocaine (mg/kg)


58
Figure 11. Mean head-in-hopper time (seconds) as a function of dose of cocaine (0.1 to
10.0 mg/kg) during automaintenance (baseline) and a return to automaintenance for Pigeons 11, 446, 922, 198, and 4986. Filled circles are from the initial automaintenance condition, and open triangles are from a return to automaintenance. Pigeon 829 was exposed to only the initial automaintenance condition (baseline). Data shown by triangles for Pigeon 446 are from the third exposure to automaintenance.


59
8
11
8
446
6
4
2
0
tr
c s
0.3
1.0
6
4
2
0
3.0 5.6 10.0
0
ft
CS 0.1
0.3
1.0
3.0 5.610.0
8
922
6
4
2
0
60
/Z.
7 x
c s
0.3
1.0
8
6
4
2
198
o
3.0 5.6 10.0
Oft
c s
0.3
1.0
3.0 5.6 10.0
8
4986
6
4
2
0
c s
0.3
8
6
4
2
0
829
1.0 1.7 3.0 5.6 10.0
T-r
c s
Automaintenance 1 A Automaintenance 2
0.3
1.0
3.0 5.6 10.0
Cocaine (mg/kg)


60
Table 7 shows mean grain-access time in seconds per trial and ratio of head-in-hopper time to grain-access time as a function of dose of cocaine during automaintenance and behavior-correlated schedules. These data provide an indicator of consumption, which is described by the ratio of head-in-hopper time to grain-access time. Generally, as dose of cocaine increased consumption decreased under conditions of automaintenance and both correlated schedules. Access to grain and consumption was greater for all pigeons when no cocaine was administered than when cocaine was administered, except for Pigeon 922 at 0.3 mg/kg and Pigeon 198 at 3.0 mg/kg during the rate-decreasing schedule and Pigeon 198 at 0.3 and 3.0 mg/kg during the rate-increasing schedule.
Table 7. Mean grain-access time in seconds for each trial and ratio of head-in-hopper
time to mean grain-access time (in parenthesis) as a function of dose of cocaine during acute dosing under automaintenance and behavior-correlated schedules.
Dose Auto 112 Decreasing Increasing
Pigeon (mg/kg)
ii Con 4.00(.69) 7.42(.72) 6.72(.87)
Sal 4.00(.71) 7.34(.80) 6.92(.89)
0.3 4.00(.73) 7.67(.74) 7.03(.84)
1.0 4.00(.63) 7.61(.74) 5.28(.81)
3.0 4.00(.68) 6.00(.69) 2.39(.78)
5.6 4.00(.68) 7.89(37) 2.34(.37)
7.4 4.00(.40) -
10.0 4.00(.27) 7.95(.32) 2.00(.00)
446 Con 4.00(.67) 6.15(.48) 5.26(.48)
Sal 4.00(.58) 5.89(.61) 4.78(.63)
0.1 4.00(.66) 6.22(.53) 5.27(.60)
0.3 4.00(.63) 7.17(.54) 3.30(.54)
1.0 4.00(.60) 7.62(.47) 2.95(.43)
4.2 4.00(.51) -
3.0 4.00(.41) 6.95(.47) 3.37(.57)
5.6 4.00(.31) 7.15(.18) 3.25(37)
10.0 4.00(.01) 8.00(.00) 2.10(35)


61
Table 7 continued.
Dose Auto 112 Decreasing Increasing
Pigeon (mg/kg)
922 Con 4.00(.86) 7.74(.77) 4.55(.80)
Sal 4.00(.87) 7.70(.76) 4.25(.86)
0.3 4.00(.86) 7.89(.93) 4.23(.86)
1.0 4.00(.87) 7.95(.82) 3.70(.80)
3.0 4.00(.68) 7.53(.46) 2.31(.68)
5.6 4.00(.14) 7.35(.46) 2.00(.24)
10.0 4.00(.01) 7.28(.15) 2.00(.00)
198 Con 4.00(.86) 6.64(.92) 2.80(.79)
Sal 4.00(.85) 6.80(.91) 2.89(.81)
0.3 4.00(.79) 7.30(.91) 3.11(.86)
1.0 4.00(.84) 6.42(.87) 2.37(.81)
3.0 4.00(.79) 7.06(.94) 2.89(.84)
4.2 5.64(.60) -
5.6 4.00(.44) 7.70(.00) 2.61(.09)
7.4 4.00(.00) 2.30(.42)
10.0 4.00(.02) 8.00(.00) 2.03(.00)
4986 Con 4.00(.86) 6.34(.83) -
Sal 4.00(.83) 5.60(.94) -
0.3 4.00(.82) 6.06(.83) -
1.0 4.00(.80) 7.08(.78) -
1.7 4.00(.40) 7.39(.71) -
3.0 4.00(.33) 8.00(.25) -
5.6 4.00(.02) 8.00(.00) -
10.0 4.00(.00) -
829 Con 4.00(.88) -
Sal 4.00(.88)
0.3 4.00(.88) -
1.0 4.00(.79)
3.0 4.00(.55) -
5.6 4.00(.20)
10.0 4.00(.05) -
Behavior-Correlated Schedule with a Direct Relation
Figure 12 shows normalized dose-response curves during automaintenance (baseline) and a rate-increasing schedule for the 4 pigeons in which drug effects were examined when a rate-increasing procedure was arranged. The number of administrations of cocaine and saline are shown in the third column of Table 6. Dose-


62
response curves were generally shifted to the left and/or downward with respect to baseline functions for Pigeons 11, 446, 922, and 198.
Behavior-Correlated Schedule with an Inverse Relation Figure 13 shows normalized dose-response curves during a rate-decreasing schedule with inverse relation for the 5 pigeons tested. Table 6 shows the number of administrations of cocaine and saline for the rate-decreasing schedule (fourth column). Dose-response curves obtained during automaintenance are shown for comparison. Curves for 4 of the pigeons were to the right of the functions obtained under simple automaintenance.
Overall, dose-response curves revealed no completely consistent effect of cocaine across pigeons under conditions of automaintenance and behavior-correlated schedules (i.e., rate-increasing and rate-decreasing schedules), but under the rate-increasing schedule dose-response curves tended to be to the left of the functions obtained during automaintenance. Under a rate-decreasing schedule, they tended to be to the right. That is, the higher rates engendered by the rate-increasing schedule tended to be decreased at smaller doses than the lower rates engendered by the rate-decreasing schedule, with effects being somewhat intermediate when automaintenance was in effect. This is a pattern of effects consistent with rate dependency so a more detailed analysis was conducted.
A quantitative rate-dependency analysis revealed that effects of cocaine were related systematically to baseline rates of responding across all conditions for Pigeons 446, 922, 198, 11, and 4986. Specifically, when drug-influenced response rates were plotted as a function of baseline (non-drug) response rates on logarithmic scaling, the distribution of points often fell along a straight line (cf. Gonzalez & Byrd, 1977a,b), as


63
shown in Figure 14. In these graphs, a straight line through the origin with a slope of 1.0 shows equal rates both under drug and no drug; that is, no effect. A slope less than 1.0 shows either that lower rates were increased more or decreased less than higher rates. That is, slopes less than 1.0 reflect a differential effect of drug correlated with whether baseline rates of responding were relatively high or low. Figure 14 shows that as dose of cocaine increased, the slopes of the functions decreased, a pattern of effects often seen with rate-dependent effects of drugs. These results are amplified in Table 8, which shows the slopes of the functions displayed in Figure 14. Also shown in Table 8 is information regarding the goodness of fits of regression lines. Given the small number of points in each graph, the fits are reasonably good.


64
Figure 12. Percent of saline response rates across a range of doses of cocaine (0.1 to 10.0
mg/kg) during behavior-correlated schedule with direct relations (i.e., rate-increasing schedule) for Pigeons 11, 446, 922, and 198. Dose-response functions obtained during automaintenance (baseline) are shown by filled circles; those from the rate-increasing schedule are shown by open diamonds.


65
C S 0.3 1 3 5.6 10 C S 0.1 0.3 1 3 5.6 10
C S 0.3 1 3 5.6 10 C S 0.3 1 3 5.6 10
Cocaine (mg/kg)


66
Figure 13. Percent of saline response rates as a function of dose of cocaine (0.1 to 10.0
mg/kg) during the initial automaintenance condition (shown by filled circles) and rate-decreasing schedule (shown by squares with crosses) for Pigeons 11, 446, 922, 198, and 4986.


67
0
CO 0
c
o
0
DC
0
03 0)
O
0 O
0
Q_
700 600 -I 500 400 300 200 100
0
300 i
250
200
150
100
50
0
C S 0.3
200
150
100
50 -
0
4986
a
C S 0.3
1
446
a
C S 0.1
0.3
1
198
T-r
3 5.6 10
C S 0.3
1
3 5.6 10
3 5.6 10
Automaintenance Rate Decreasing
Cocaine (mg/kg)
1 1.7 3 5.6 10
Cocaine (mg/kg)


68
Figure 14. Rate-dependency plots display the logarithm of drug response rate as a
function of control response rate for Pigeons 446, 922, 198, 11, and 4986. Plots from automaintenance conditions are shown by filled circles (baseline) and open triangles (reversal); plots from the behavior-correlated schedules are shown by filled diamonds (rate-increasing schedule) and open squares with a cross (rate-decreasing schedule). Cocaine doses range from 0.1 to 10.0 mg/kg; see text for criterion for dose selection.


69
446 922 198 11 4986
-1 01 23 -1 01 23 -1 01 23
Log Control Response Rate


70
Table 8. R2 values, slopes, and standard error of estimate for data shown in Figure 14
Dose r slope std err of Pigeon estimate
446 0.1 .82 0.97 .26
0.3 .77 0.64 .19
1.0 .92 0.12 .02
3.0 .11 0.06 .09
5.6 .00 0.09 .28
922 0.3 .86 1.09 .44
1.0 .62 0.28 .90
3.0 .43 0.50 .58
5.6 .00 -0.04 .53
10.0 .11 -0.09 .26
198 0.3 .72 1.34 .41
1.0 .94 1.78 .22
3.0 .89 1.15 .20
5.6 .48 0.94 .49
7.4 .60 0.35 .33
11 0.3 .87 1.26 .42
1.0 .49 1.03 .90
3.0 .15 0.23 .46
5.6 .80 1.25 .54
4986 0.3 .95 0.66 .06
1.0 .51 0.63 .26
1.7 .30 -0.90 .56


CHAPTER 7 DISCUSSION
Experiment 1
The results from Experiment 1 showed that by arranging a contingency between number of key pecks per trial and amount of grain delivered at the end of that trial altered behavior. Autoshaping has been considered a Pavlovian conditioning procedure because grain is contingent on and contiguous with the offset of key illumination; responses have no programmed consequences. From this perspective, the key light becomes a conditional stimulus (CS) because it has been repeatedly paired with grain delivery, an unconditional stimulus (US). Key pecks to the illuminated key are conditional responses (CR) and resemble pecks towards the grain, or unconditional responses (UR). This characterization has been supported by studies that showed the form of the automaintained response (CR) is similar to the UR (cf. Jenkins & Moore, 1973). Further support was provided by studies that showed responding of pigeons persisted even when responding resulted in the omission of the delivery of grain at the end of the signal (cf. Williams & Williams, 1969).
Although in its design, the autoshaping procedure is Pavlovian, elements
important for instrumental conditioning may be inherent in the procedure. During
automaintenance, it is probable that a response will be followed relatively shortly by
grain delivery. This is important because it establishes temporal contiguity between these
two events, a response and delivery of grain. Thus, there may be some instrumental
conditioning at work during the ostensibly Pavlovian arrangement.
71


72
The results from this experiment show both respondent and operant effects. The design of the automaintenance procedure is respondent but it allows for operant contingencies to be established by contiguity between the last peck and the delivery of grain. When a contingency was systematically programmed between responding and reinforcer magnitude, responding varied in accordance with reinforcer magnitude. Other researchers have also shown that automaintained key pecks are sensitive to variations in reinforcer magnitude (Dodd, 1980; Gentry & Eskew, 1984; Hendry, 1962).
In the present experiment, there was some evidence of Pavlovian conditioning when the behavior-correlated contingencies were in effect. Overall mean latency to respond across pigeons during key-light illumination was shorter during the rate-decreasing schedule than during automaintenance for all 5 of the pigeons tested, and longer during the rate-increasing schedule compared to automaintenance for 4 of the 6 pigeons tested (Table 4). These differences, however, were modest. That latencies did not differ much with respect to schedule provides evidence that latency to first key peck may have been under respondent control.
Rates of responding increased or decreased when behavior-correlated contingencies were in effect (Table 3), and this provides evidence for operant control over responding. Under both of the behavior-correlated schedules, responding adjusted so that the maximum reinforcement often was produced (Figures 3 and 7). As would be expected from these results, the time from the last response in a trial and the delivery of grain at the end of a trial varied with respect to the schedule of reinforcement that was in effect (Table 5). The overall time since the last key peck to grain delivery across all pigeons (i.e., group means) was 4.43 s during automaintenance (initial and second exposures), 3.19 s during the rate-increasing schedule, and 6.14 s during the rate-


73
decreasing schedule. A classical conditioning account of the results for the rate-increasing schedule could be interpreted in the following manner: responding (CR) increased systematically because the amount of grain delivered (US) increased on average from 4 seconds to 4.67 seconds (the mean of 2, 4, and 8 seconds). The CR varies systematically as a function of increasing amounts of US (Pavlov, 1927). But this reasoning is violated by the findings in the rate-decreasing schedule wherein low rates of responding produced greater amounts of grain and high rates of responding produced smaller amounts of grain. To be in accordance with a classical conditioning paradigm, the results for the rate-decreasing schedule should have been such that high rates of responding were maintained even though it decreased grain amount. This was not the case in the current experiment. Rates of responding varied systematically as a function of the contingencies between reinforcer magnitude and responding, and these results are consistent with previous findings (Balsam, Brownstein, & Shull, 1978; Dodd, 1980; Gentry & Eskew, 1984; Hendry, 1962; Hendry & Van-Toller, 1964).
A study by Williams and Williams (1969) provided support for the notion that autoshaped responding is Pavlovian. In their study, pigeons were exposed to the following arrangement: if no key peck occurred during a 6-s key illumination, with averaged ITIs of 30 s, then 4-s access to grain was available upon termination of the illuminated key light (called an omission procedure). Despite this negative contingency
between key pecks and grain delivery, all pigeons in their experiment pecked at the illuminated key light. The degree to which responding was maintained under this arrangement varied across pigeons.
Although the present experiment was not designed to investigate systematically the effects of an omission procedure on automaintained responding, some components of


74
the behavior-correlated schedule may be similar to an omission procedure. For example, in the rate-decreasing schedule, high rates of responding produced less grain than it had
previously. That is, there was a relative reduction in reinforcer amount rather than a total elimination, as in the case of an omission procedure. In the current study, there was a reduction in rates of responding for all pigeons when the rate-decreasing contingency was implemented but responding was completely suppressed for only 1 of the 6 pigeons (Pigeon 198). This was similar to the findings reported by Williams and Williams (1969), yet a conclusion of Pavlovian (to the exclusion of operant) control in the current study is unwarranted. Although response rates were decreased, that responding was sensitive to alterations in the behavior-correlated contingencies and could be either increased or decreased based on the reinforcer magnitude, leads to the conclusion that responding was under operant control.
The data from the current experiment support a functional approach to an operant-respondent distinction. That is, the automaintenance procedure could be considered an FT 8-sec schedule (with ITIs that averaged 40 s) that contained elements of respondent conditioning. It was made to be functionally operant because once contingencies were placed on the number of responses that occurred per trial, response rates varied systematically (Table 3). The first instance in which an automaintained response came under operant control (during the automaintenance procedure) is open to speculation because it was not systematically investigated in this experiment. This is because when a presumably elicited key peck was followed shortly with grain presentation, which established temporal contiguity between a response and an event known to be an effective reinforcer for key pecks, a distinction between respondent and operant conditioning was obscured. It is assumed that the first response during autoshaping (i.e.,


75
acquisition of responding) was a result of respondent processes because a contingency between key pecks and food delivery was not explicitly programmed, nor had any accidental contingencies been experienced. That is, food delivery was unconditional upon key pecks up to that point in time. All subsequent pecks, however, might have been influenced by the potential operant effects of the first peck (and its successors) having occurred with some degree of temporal contiguity to food (Schwartz & Gamzu, 1977; although see Baum, 1973). However, once a behavior-correlated schedule was superimposed on the automaintenance schedule and responding varied systematically as a function of the response requirements, then operant effects were clearly evident. The findings from this experiment extend previous findings that showed operant contingencies alter behavior that originally was established using a respondent conditioning technique (Dodd, 1980; Gentry & Eskew, 1984; Hendry, 1962).
In the current experiment, the number of responses per trial did not vary systematically as a function of preceding ITI during the automaintenance procedure (Table 2). The mean number of responses across pigeons ranged from 3.19 following a 64-s ITI to 3.85 following a 16-s ITI. This differs from previous studies that showed the speed with which autoshaped responding was acquired varied as a function of the ratio of ITI to trial duration (cf, Gibbon, Baldock, Locurto, Gold, & Terrace, 1977; Terrace, Gibbon, Farrell, & Baldock, 1975). However, in the studies by Gibbon et al. (1977) and Terrace et al. (1975), the effect of ITI on responding was studied across conditions whereas in the current study it was examined within session.
Specifically, trial duration and ITIs were manipulated systematically in a study by Gibbon et al. (1977) to examine acquisition and maintenance of autoshaped key pecks by pigeons. In Experiment 1 of their study, the number of trials to acquisition and rates of


76
responding (once key pecking was established) were examined across 25 groups of pigeons. Trial durations, or key light illuminations, were fixed (range 1 to 64 s) and ITIs varied (range: 6 to 768 s) across groups. The ratios of ITI to trial duration ranged from 2:1 to 96:1). When the ratio was high, autoshaped key pecks were acquired more rapidly than when the ratio was low. Once key pecking was established, rates of responding varied inversely with absolute trial duration. Additionally, there was no clear effect of ITI on the overall level of responding, which is consistent with the findings from the current study. In the study by Gibbon et al. (1977), the relative (rather than absolute) durations of ITI and trial duration controlled the speed with which autoshaped responding was acquired. Once autoshaped responding emerged, there was a shift toward more control over sustained responding by trial duration. Latency to the first response was also insensitive to ITI values and ratio of ITI to trial duration, and showed a modest relationship to trial durations.
In Experiment 2 of their study, the ratio of ITI to trial duration was fixed at 3, 6, 12, and 24 across 4 groups of pigeons. Intertrial intervals were 48, 96, 192, and 384 s and trial durations were 8, 12, 16, and 32 s. Under this arrangement, they found that the number of trials to acquisition increased as the ratio of ITI to trial duration increased. They also found that the speed of acquisition of an autoshaped response did not depend on whether the ITI was fixed or variable. Similar results have been reported by other researchers (Terrace, Gibbon, Farrell, & Baldock, 1975; Perkins, Beavers, Hancock, Hemmendinger, Hemmendinger, & Ricci, 1975).
The correlated-reinforcement contingency influenced the number of trials in which differential grain deliveries were presented across successive sessions for 4 of the 6 subjects (Pigeons 4986, 11, 922, and 198; Figures 3 and 7). These results were


77
consistent with those of other studies of correlated-reinforcement magnitude under automaintenance conditions in that the number of responses per trial increased relative to baseline rates when a relation was established between the number of responses and differential grain deliveries (cf. Dodd, 1980; Gentry & Eskew, 1984). The results obtained using a rate-decreasing schedule extend previous findings that showed responding was sensitive correlated-reinforcement contingencies. That is, in the current study, the number of responses per trial decreased relative to baseline rates when a contingency was established between a low number of responses during the 8-s key illumination and a large amount of grain delivered upon termination of the key illumination. This extends previous findings because it demonstrated that responding could be increased or decreased when differential grain deliveries were presented. For example, Dodd (1980) used a behavior-correlated schedule with 8-s key illuminations and found that the median number of responses increased or decreased when reinforcer rate was contingent upon high or low rates, respectively.
Sensitivity to differential grain amounts was not evident (at least initially) for Pigeon 198 during the rate-increasing schedule (Figure 4). In fact, responding was not sensitive to differential grain amounts until a two-valued procedure with extreme grain durations (e.g., 0.5 vs. 12 s) was in effect (Table 1, columns 6-11; Figure 6). When grain amount was varied such that two, rather than three, different magnitudes were available responding showed sensitivity to differential grain amounts. The amount of grain earned during these manipulations increased as a function of the ratio of small-to-large grain durations (Figure 6). That is, the greater the difference in the two grain amounts delivered in a trial, the more times the larger amount of grain was earned. For example, when grain amount was either 0.5 or 12 s, the number of earned 12-s grain deliveries gradually


78
increased while the number of those with 0.5-s durations decreased across 82 sessions (Figure 6, third panel). When the difference in access times was small, such as when grain-access time was either 2 or 8 s, then the number of earned 2-s durations was higher than the number of earned 8-s durations (Figure 6, panels 1 and 5).
Response-independent schedules may generate responding not explicitly programmed that may interfere with the operant under investigation. In an experiment by Skinner (1948), grain was delivered on a time-based schedule to pigeons irrespective of their behavior. The pigeons developed stereotyped and idiosyncratic behavior patterns, which Skinner called "superstitious" responding. When behavior is followed by a reinforcing stimulus, such as food, the likelihood of that behavior increases (Skinner, 1953). Even though a contingency was not explicitly programmed between a particular pattern of responses and the delivery of grain, responses that occurred prior to the delivery of grain were strengthened when grain was delivered periodically (Morse & Skinner, 1957; although see Staddon & Simmelhag, 1971; Timberlake & Lucas, 1985). In the current study, there was some anecdotal evidence of "superstitious" responding for two pigeons whose rate of responding was diminished and did not recover after exposure to the rate-decreasing correlated contingencies (Pigeons 829 and 198). When the key light was illuminated, Pigeon 829 would "air peck" around the illuminated key in a clockwise pattern, and only occasionally would its beak make contact with the key.
Pigeon 198 displayed a peck-and-check pattern during the key illumination; that is, it would peck the key and then move its head to the hopper before pecking again. This evidence is anecdotal but suggests that adventitiously reinforced behavior may have played a role in its response-rate reductions.


79
Movements like those exhibited by Pigeons 829 and 198 could be quantified by increasing the precision of measurement (e.g., via a tracking system) in the operant chamber. By casual observation, these patterns of responding appeared orderly because they occurred only in the presence of the illuminated key light, and they were highly repetitive and stereotypical. Perhaps improved measurement of this kind of behavior would allow for analysis of these responses as a function of reinforcer magnitude using an automaintenance procedure. Data of this sort would also provide evidence of whether such stereotyped patterns of responding are respondent or operant. For example, based on observed differences in head movements of pigeons and distance of the head from the key during schedules of reinforcement and autoshaping, Pear and Eldridge (1984) suggested that the operant-respondent distinction can be elucidated by increasing the precision of the measurement system. They suggest the framework should be modified to account for disparate data using topographical and functional approaches (see also Eldridge & Pear, 1987).
Experiment 2
The results from Experiment 2 showed that following cocaine administration a) rates of responding decreased in a dose-dependent manner under automaintenance and behavior-correlated schedules; and, b) rate-dependent effects were observed.
Dose-related effects on response rates were observed across pigeons under
automaintenance and behavior-correlated schedules (rate-increasing and rate-decreasing schedules; Figures 9, 10, 12, and 13). In general, rates of responding under small doses were similar to control rates, whereas responding was diminished at large doses. There were several instances during the automaintenance and behavior-correlated schedules, however, in which rates of responding were increased relative to control rates. During


80
the initial automaintenance condition, response-rate increases were observed for Pigeon 11 at 1.0 mg/kg (and marginally at 3.0 mg/kg), Pigeon 446 at 4.2 mg/kg, and Pigeon 922 at 3.0 mg/kg. During the return to automaintenance, such increases were observed for Pigeon 922 at 1.0 mg/kg, 3.0 mg/kg, and 5.6 mg/kg, and for Pigeon 198 at 3.0 mg/kg and 5.6 mg/kg. During the rate-increasing schedule, rates of responding were increased relative to control rates for Pigeon 198 at 0.3 mg/kg. During the rate-decreasing schedule, this effect was observed for Pigeon 11 at 3.0 mg/kg, for Pigeon 922 at 5.6 mg/kg and 10.0 mg/kg, and for Pigeon 198 at 5.6 mg/kg. Although dose-dependent decreases in rate of responding were generally observed across all 3 conditions for most pigeons, there were several instances of dose-dependent increases in responding (relative to control rates) across most doses of cocaine (range: 0.3 to 10.0 mg/kg) during the automaintenance and behavior-correlated schedules.
Dose-dependent decreases in responding have been previously reported in studies using amphetamine and response-independent schedules (Poling & Appel, 1978, 1979; Poling & Thompson, 1977; Thompson & Corr, 1974). In a study by Poling and Thompson (1977), an automaintenance procedure was implemented, and then the relationship between key light illumination and the delivery of grain was systematically manipulated. The probability of grain delivery after the 8-s key illumination was 100, 50, 25, 10 and 2.5 % across conditions. Effects of d-amphetamine (0.5, 1.0, and 2.0 mg/kg) were examined under each condition. The mean number of responses per key illumination decreased as a function of increasing dose of d-amphetamine across all 5 conditions. Thompson and Corr (1974) also found dose-dependent decreases in responding during a VI component of a multiple VI VT schedule. Rates of responding decreased as dose of d-amphetamine increased under the VI component; rates of


81
responding in the VT component were low (less than 1 response per minute) during control and saline sessions, and d-amphetamine produced no measurable effects. The results from the current study also showed that cocaine administration under a response-independent schedule can produce dose-dependent decreases in responding.
In the current experiment, a head-in-hopper measure (i.e., amount of time that a pigeon's head was in the grain aperture) was included to validate grain consumption and, thus, reinforcer magnitude. The ratio of head-in-hopper time to mean grain-access time is an indicator of consumption because it's likely that the pigeon is eating grain when its head is in the aperture and the grain is available. The head-in-hopper times and consumption ratios decreased as dose of cocaine increased across automaintenance and behavior-correlated schedules (Table 7). Generally, low rates of responding were correlated with low head-in-hopper times during acute dosing during the initial and second exposures to automaintenance (Figure 11). The exceptions were for Pigeons 11 and 4986 during the second exposure to automaintenance. That is, head-in-hopper functions were elevated compared to dose-response functions at all doses for Pigeons 11 and 4986 (except at 5.6 mg/kg for Pigeon 4986; Figures 10 and 11). For these two pigeons, head-in-hopper times were insensitive to response-rate decreases across a range of doses of cocaine.
Analysis of the effects of cocaine across the automaintenance and behavior-correlated schedules revealed rate-dependent effects. The rate-dependency hypothesis is that effects of a drug on behavior are related to the rate of responding under control conditions (cf. Dews, 1958). That is, a systematic relationship exists between control rate and effect of a drug, and is often characterized by a linear log-log relationship. Many other researchers have provided support for the notion of rate dependency (Kelleher &


82
Morse, 1968; Leander & McMillan, 1974; McMillan, 1968a,b, 1969; Smith, 1964; Clark & Steele, 1966; Wuttke, 1970; McKearney, 1970). Rate dependency has been observed under a variety of schedules and across classes of drugs, (McKearney, 1970; McMillan, 1968a,b; Smith, 1964), including amphetamine (Clark & Steele, 1966; Kelleher & Morse, 1968a; Leander & McMillan, 1974; Smith, 1964; Wuttke, 1970) and cocaine (Gonzalez & Goldberg; 1977; Smith, 1964; Spealman, Goldberg, Kelleher, Goldberg, & Charlton, 1977).
Despite these robust findings of rate dependency across response-dependent schedules, there have been few studies demonstrating this effect using behavior-correlated schedules, or schedules of response-independent stimulus presentations. One such study was conducted by Miller et al. (2001). In that study, rate dependency was observed when cocaine was administered using a behavior-correlated FI 8-min schedule. Responding in the first 4 minutes of the interval determined the amount of grain to be delivered at the end of the 8-min interval. Behavior-correlated contingencies produced no increase in the amount of grain delivered at the end of the FI 8-min schedule, that is the correlated-schedule contingency exerted no effect on behavior. When cocaine was administered acutely, an increase grain amount was observed at low to moderate doses for all 4 pigeons in that study, a result of increased responding early in the FI. Additionally, response patterning was disrupted during acute cocaine administration. Response rates often decreased in the last half of the interval (i.e., the last 4 minutes) after cocaine deliveries. As such, these within-interval effects showed rate dependency.
In the current study, drug administration may be considered a response-independent manipulation because the pigeons received cocaine injections regardless of their schedule performance. Thus, perhaps a crucial component of rate-dependency is the


83
lack of contingency between a stimulus (i.e., drug) and a response. Future studies could investigate whether, and to what extent, rate dependency occurs using a self-administration protocol (see Griffiths, Brady, & Bradford, 1979; Johanson, 1978 for self-administration techniques). In this manner, drug delivery would be response dependent. The importance of the role of a contingency between a response and drug administration could be systematically investigated by manipulating schedule parameters and dose of a drug.
One possible implication of these data, when contrasted with those of Thompson and Corr (1974), is that the baseline rate of responding is one of the most crucial features for determination of a drug's effect in experiments that used a single schedule, or single operant. This conclusion is related to Skinner's (1938, 1953) view that rate of responding is a fundamental measure of behavior, and also suggests that mechanisms underlying rate differences are "where the action is."
Rate-dependent effects of drugs are usually limited to operant behavior. Such effects have not been observed with respondent behavior, such as automaintained responding (Thompson & Corr, 1974). In the present experiment, rate dependency was observed when automaintenance data were analyzed in conjunction with behavior-correlated schedule data. That is, operant contingencies were necessary for observation of rate dependency.


LIST OF REFERENCES
Balsam, P.D., Brownstein, A.J., & Shull, R.L. (1978). Effect of varying the duration of
grain presentation on automaintenance. Journal of the Experimental Analysis of Behavior, 29, 27-36.
Balster, R.L. & Schuster, CR. (1973). Fixed-interval schedule of cocaine
reinforcement: effect of dose and infusion duration. Journal of the Experimental Analysis of Behavior, 20,419-129.
Barrett, J.E. (1976). Effects of alcohol, chlordiazepoxide, cocaine and pentobarbital on
responding maintained under fixed-interval schedules of food or shock
presentation. Journal of Pharmacology and Experimental Therapeutics, 196, 605-615.
Baum, W.M. (1973). The correlation-based law of effect. Journal of the Experimental
Analysis of Behavior, 20, 137-153.
Brown, P.L. & Jenkins, H.M. (1968). Auto-shaping of the pigeon's key-peck. Journal of
the Experimental Analysis of Behavior, 11, 1-8.
Clark, F.C., & Steele, B.J. (1966). Effects of d-amphetamine on performance under a
multiple schedule of reinforcement in the rat. Psychopharmacologia, 9, 157-169.
Dews, P.B. (1955a). Studies on behavior I. Differential sensitivity to pentobarbital of
pecking performance in pigeons depending on the schedule of reward. Journal of Pharmacology and Experimental Therapeutics, 113, 393-401.
Dews, P.B. (1955b). Studies on behavior II. The effects of pentobarbital,
methamphetamine and scopolamine on performances in pigeons involving discriminations. Journal of Pharmacology and Experimental Therapeutics, 115, 380-389.
Dews, P.B. (1958). Studies on behavior: IV. Stimulant actions of methamphetamine.
Journal of Pharmacology and Experimental Therapeutics, 122, 137-147.
Dews, P.B. (1964). A behavioral effect of amobarbital. Naunyn-Schmiedebergs Archiv
fur Experimentellen Pathologie und Pharmakologie, 248, 296-307.
84


85
Dews, P.B. & Wenger, G.R. (1977). Rate-dependency of the behavioral effects of
amphetamine. In T. Thompson & P.B. Dews (Eds.), Advances in behavioral pharmacology (Vol. 1, pp. 167-227). New York: Academic Press.
Dodd, P. (1980). Discrimination and differentiation of response number in stimulus
directed pecking of pigeons. Journal of the Experimental Analysis of Behavior, 33, 253-264.
Eldridge, G.D. & Pear, J.J. (1987). Topographical variations in behavior during
autoshaping, automaintenance, and omission training. Journal of the Experimental Analysis of Behavior, 47, 319-333.
Ferster, C.B., & Skinner, B.F. (1957). Schedules of reinforcement. New York: Appleton
Century Crofts.
Gardner, W.M. (1969). Auto-shaping in bob white quail. Journal of the Experimental
Analysis of Behavior, 12, 279-281.
Gentry, G.D., & Eskew, Jr., R.T. (1984). Graded differential reinforcement: Response-dependent reinforcer amount. Journal of the Experimental Analysis of Behavior, 41, 27-34.
Gibbon, J., Baldock, M.D., Locurto, C, Gold, L., & Terrace, H.S. (1977). Trial and
intertrial durations in autoshaping. Journal of Experimental Psychology: Animal Behavior Processes, 3, 264-284.
Gonzalez, F.A., & Byrd, L.D. (1977a). Mathematics underlying the rate-dependency
hypothesis. Science, 195, 546.
Gonzalez, F.A., & Byrd, L.D. (1977b). Rate-dependency hypothesis. Science, 198,
1182.
Gonzalez, F.A., & Goldberg, S.R. (1977). Effects of cocaine and d-amphetamine on
behavior maintained under various schedules of food presentation in squirrel monkeys. Journal of Pharmacology and Experimental Therapeutics, 201, 33.
Griffiths, R.R., Brady, J.V., and Bradford, L.D. (1979). Predicting the abuse liability
with animal self-administration procedures: Psychomotor stimulants and hallucinogens. In T. Thompson & P.B. Dews (Eds.), Advances in Behavioral Pharmacology (vol 2, pp. 163-208). New York: Academic Press.
Heffner, T.G., Drawbaugh, R.B., and Zigmond, M.J. (1974). Amphetamine and operant
behavior in rats: relationship between drug effect and control rate. Journal of Comparative and Physiological Psychology, 86, 1031-1043.
Hendry, D.P. (1962). The effect of correlated amount of reward on performance on a
fixed-interval schedule of reinforcement. Journal of Comparative and Physiological Psychology, 55, 387-391.


86
Hendry, D.P., & Van-Toller, C. (1964). Performance on a fixed-ratio schedule with
correlated amount of reward. Journal of the Experimental Analysis of Behavior, L 207-209.
Jenkins, H.M., & Moore, B.R. (1973). The form of the auto-shaped response with food
or water reinforcers. Journal of the Experimental Analysis of Behavior, 20, 163-181.
Johanson, C.E. (1978). Drugs as reinforcers. In D.E. Blackman & D.J. Sanger (Eds.),
Contemporary research in behavioral pharmacology (pp. 325-390). New York: Plenium Press.
Kelleher, R.T., & Morse, W.H. (1968). Determinants of the specificity of the behavioral
effects of drugs. Ergebnisse der Physiologie Biologischen Chemie und Experimentallen Pharmakologie, 60, 1-56.
Leander, J.D., & McMillan, D.E. (1974). Rate-dependent effects of drugs. I.
Comparisons of d-amphetamine, pentobarbital and chlorpromazine on multiple and mixed schedules. Journal of Pharmacology and Experimental Therapeutics, 188, 726-739.
Lemaire, G.A. & Meisch, R.A. (1985). Oral drug self-administration in rhesus
monkeys: interactions between drug amount and fixed-ratio size. Journal of the Experimental Analysis of Behavior, 44, 377-389.
Leslie, J.L., Boakes, B.A., Linaza, J., & Ridgers, A. (1979). Autoshaping using visual
stimuli in the rat. The Psychological Record, 29, 523-546.
McKearney, J.W. (1970). Rate-dependent effects of drugs: modification by
discriminative stimuli of the effects of amobarbital on schedule-controlled behavior. Journal of the Experimental Analysis of Behavior, 14, 167-175.
McMillan, D.E. (1968a). The effects of sympathomimetic amines on schedule-controlled
behavior in the pigeon. Journal of Pharmacology and Experimental Therapeutics, 160,315-325.
McMillan, D.E. (1968b). Some interactions between sympathomimetic amines and
amine-depleting agents on the schedule-controlled behavior of the pigeon and
squirrel monkey. Journal of Pharmacology and Experimental Therapeutics, 163, 172-187.
McMillan, D.E. (1969). Effects of d-amphetamine on performance under several
parameters of multiple fixed-ratio, fixed-interval schedules. Journal of Pharmacology and Experimental Therapeutics, 167, 26-33.
Meisch, R.A. (2000). Relative persistence of behavior: a fundamental measure of
relative reinforcing effects. Experimental and Clinical Psychopharmacology, 8, 333-349.


87
Meisch, R.A. & Stewart, R.B. (1995). Relative reinforcing effects of different doses of
orally delivered cocaine. Drug and Alcohol Dependence, 37, 141-147.
Miller, M.L., Brodkorb, G.W., & Branch, M.N. (2001). Tolerance to the effects of
cocaine on performance under behavior-correlated reinforcement magnitude. Journal of the Experimental Analysis of Behavior, 76, 217-234.
Morse, W.H. & Skinner, B.F. (1957). A second type of superstition in the pigeon.
American Journal of Psychology, 70, 308-311.
Palya, W.L. & Walter, D.E. (1993). A powerful, inexpensive controller or IBM PC
interface and experimental control language. Behavior Research Methods, Instruments & Computers, 25(2), 127-136.
Pavlov, LP. (1927). Conditioned reflexes: an investigation of the physiological activity
of the cerebral cortex. London and New York: Oxford University Press.
Pear, J.J., & Eldridge, G.D. (1984). The operant-respondent distinction: future
directions. Journal of the Experimental Analysis of Behavior, 42, 453-467.
Perkins, C.C., Jr., Beavers, W.O., Hancock, R.A., Jr., Hemmendinger, P.C.,
Hemmendinger, D., & Ricci, J. A. (1975). Some variables affecting rate of key pecking during response-independent procedures (autoshaping). Journal of the Experimental Analysis of Behavior, 24, 59-72.
Peterson, G.B., Ackil, J., Frommer, G.P., & Hearst, E. (1972). Conditioned approach
and contact behavior towards signals for food or brain-stimulation reinforcement. Science, 177, 1009-1011.
Poling, A. & Appel, J.B. (1978). Drug effects under automaintenance and negative
automaintenance procedures. Pharmacology, Biochemistry & Behavior, 9, 315-318.
Poling, A., & Appel, J.B. (1979). Drug effects on the performance of pigeons under a
negative automaintenance schedule. Psychopharmacology, 60, 207-210.
Poling, A. & Thompson, T. (1977). The effects of d-amphetamine on the
automaintained key pecking of pigeons. Psychopharmacology, 51, 285-288.
Sanger, D.J., & Blackman, D.E. (1976). Rate-dependent effects of drugs: Review of the
literature. Pharmacology, Biochemistry and Behavior, 4, 73-83.
i
Schwartz, B, & Gamzu, E. (1977). Pavlovian control of operant behavior. In Honig,
W.K., & Staddon, J.E.R. (Eds.), Handbook of Operant Behavior. (Pp. 53-97). New Jersey: Prentice-Hall, Inc. 201-216.


88
Sidman, M. (1960). Tactics of Scientific Research: Evaluating Experimental Data in
Psychology. Boston: Authors Cooperative, Inc.
Sidman, M., & Fletcher, F.G. (1968). A demonstration of auto-shaping with monkeys.
Journal of the Experimental Analysis of Behavior, 11, 307-309.
Skinner, B.F. (1938). Behavior of organisms: An experimental analysis. Acton: Copely.
Skinner, B.F. (1948). "Superstition" in the pigeon. Journal of Experimental
Psychology, 70,308-311.
Skinner, B.F. (1953). Science and human behavior. New York: Macmillan.
Smith, C.B. (1964). Effects of d-amphetamine upon operant behavior of pigeons:
enhancement by reserpine, Journal of Pharmacology and Experimental Therapeutics, 146, 167-174.
Smith, S.G., & Smith, W.M. (1971). A demonstration of auto-shaping with dogs.
Psychological Records, 21, 377-379.
Spealman, R.D., Goldberg, R.T., Kelleher, D.M., Goldberg, D.M., & Charlton, J.P.
(1977). Some effects of cocaine and two cocaine analogs on schedule-controlled behavior of squirrel monkeys. Journal of Pharmacology and Experimental Therapeutics, 202, 500-509.
Spealman, R.D. & Kelleher, R.T. (1979). Behavioral effects of self-administered
cocaine: responding maintained alternatively by cocaine and electric shock in squirrel monkeys. Journal of Pharmacology and Experimental Therapeutics, 210, 206-214.
Squier, L.H. (1969). Auto-shaping key responses in fish. Psychonomic Science, 17,
177-178.
Staddon, J.E.R., & Simmelhag, V.L. (1971). The "superstitious" experiment: a
reexamination of its implications for the principles of adaptive behavior. Psychological Review, 78, 3-43.
Stewart, R.B., Wang, N., Bass, A.A., & Meisch, R.A. (2002). Relative reinforcing
effects of different oral ethanol doses in rhesus monkeys. Journal of the Experimental Analysis of Behavior, 77, 49-64.
Terrace, H.S., Gibbon, J., Farrell, L., & Baldock, M.D. (1975). Temporal factors
influencing the acquisition and maintenance of an autoshaped keypeck. Animal Learning and Behavior, 3, 53-62.
Thompson, D.M., & Corr, P.B. (1974). Behavioral parameters of drug action: Signaled
and response-independent reinforcement. Journal of the Experimental Analysis of Behavior, 21, 151-158.


89
Timberlake, W., & Lucas, G.A. (1985). The basis of superstitious behavior: chance
contingency, stimulus substitution, or appetitive behavior? Journal of the Experimental Analysis of Behavior, 44, 279-299.
Wasserman, E.A. (1973). Pavlovian conditioning with heat reinforcement produces
stimulus-directed pecking in chicks. Science, 181, 875-877.
Williams, D.R., & Williams, H. (1969). Auto-maintenance in the pigeon: Sustained
pecking despite contingent non-reinforcement. Journal of the Experimental Analysis of Behavior, 12, 511-520.
Woolverton, W.L. & English, J. A. (1997). Further analysis of choice between cocaine
and food using the unit price model of behavioral economics. Drug and Alcohol Dependence, 49, 71-78.
Wuttke, W. (1970). The effects of d-amphetamine on schedule-controlled water licking
in the squirrel monkey. Psychopharmacologia, 17, 70-82.
Zeiler, M.D. (1972). Superstitious behavior in children: An experimental analysis.
Advances in Child Development and Behavior, 7, 1-29.
Zolman, J.F., Chandler, S.D., & Black, D. (1972). Visual discrimination learning of the
young chick: Key-pecking conditioning with heat-light reinforcement. Developmental Psychobiology, 5, 181-187.


BIOGRAPHICAL SKETCH
Michelle Miller's academic career began at the University of California, San Diego. There she worked under the supervision of Dr. Edmund Fantino on a project titled, "Negative instances in a delayed matching-to-sample test to assess the base-rate error in pigeons." After earning a B.A. in Psychology in 1995, Michelle worked in a residential treatment facility for children with behavior disorders, wherein pharmacological agents were used as a means by which to control behavior. As such, she became interested in the effects of drugs on behavior and sought to investigate some of these mechanisms in a laboratory.
At the University of Florida, Michelle conducted several behavioral pharmacology experiments with Dr. Marc Branch beginning in 1998. These experiments involved discovering behavioral mechanisms that contribute to the development of tolerance to cocaine. One of these studies served as the basis for an M.S. in psychology, which she earned in 2001. In 2002, Michelle received a National Research Service Award from the National Institute of Health and Drug Abuse. In 2004 she earned a Ph.D. in psychology from the University of Florida. She has published articles in the following journals:
Journal of the Experimental Analysis of Behavior
Psychopharmacology
Pharmacology, Biochemistry and Behavior
90