A detailed analysis of the effects of methamphetamine on schedule-controlled responding in pigeons

In behavioral pharmacology, overall response rate is the measure most often used to study the effects of drugs on an animal’s ongoing operant behavior. However, it has been argued that an overall measure of response rate may not provide the most valid description of behavior and, thus, a more detailed analysis might prove beneficial in classifying drug effects. The purpose of the present research was to (1) study effects of methamphetamine on schedulecontrolled responding, (2) determine if a log-survivor analysis recently used to study nose-poke responding in rats could be extended for use with key-pecking in pigeons and (3) examine microproperties of pigeon responding using an IRT analysis. Pigeons were trained to peck a lighted key under a multiple random-interval (RI) 1-min, RI 4-min schedule. Interresponse times (IRTs) were collected and then analyzed via log-survivor plots and an analysis described by Blough (1963) and used in greater detail by Palya (1992). Log-survivor plots show the proportion of IRTs greater than some time as a function of time in the session. The IRT analysis involved graphing each individual IRT as a function of its temporal position within the interval. Several doses of methamphetamine were administered (0.3 – 5.6 mg/kg) and the effect on both summary measures (overall response rates) as well as the more detailed analyses was observed. Log-survivor analyses did not produce the “broken-stick” responding seen in previous studies with rats and quantitative measures of bout-initiations and within-bout responses could not be obtained. The IRT plots showed clear bands of responding at 350 ms and 700 ms and responding in the initial (350 ms) band was affected more than responding during other bands or at other IRTs. These results strengthen the argument that not all responding (various IRTs) is uniformly affected by methamphetamine administration and suggest that a detailed analysis of responding might prove more useful than summary measures in characterizing drug effects on behavior