Winter, Y. (1998). Energetic cost of hovering flight in a nectar-feeding bat measured with fast-response respirometry.

Quantitative theories with free parameters often gain credence when they closely fit data. This is a mistake. A good fit reveals nothing about the flexibility of the theory (how much it cannot fit), the variability of the data (how firmly the data rule out what the theory cannot fit), or the likelihood of other outcomes (perhaps the theory could have fit any plausible result), and a reader needs all 3 pieces of information to decide how much the fit should increase belief in the theory. The use of good fits as evidence is not supported by philosophers of science nor by the history of psychology; there seem to be no examples of a theory supported mainly by good fits that has led to demonstrable progress. A better way to test a theory with free parameters is to determine how the theory constrains possible outcomes (i.e., what it predicts), assess how firmly actual outcomes agree with those constraints, and determine if plausible alternative outcomes would have been inconsistent with the theory, allowing for the variability of the data.Many quantitative psychological theories with free parameters are supported mainly or entirely by demonstrations that they can fit data-that the parameters can be adjusted so that the output of the theory resembles actual results. The similarity is often shown by a graph with two functions: one labeled observed (or data) and the other labeled predicted (or theory or simulated). That the theory fits data is supposed to show that the theory should be taken seriously-should be published, for example. This type of argument is common; judging from a search of Psychological Abstracts , the research literature probably contains thousands of examples. Early instances involved sensory processes (Hecht, 1931) and animal learning (Hull, 1943), but this reasoning is now used in many areas. Here are three examples.1. Cohen, Dunbar, and McClelland (1990) proposed a parallel distributed processing model to explain the Stroop effect and related data. The model was meant to embody a "continuous" view of automaticity, in contrast to an "all-or-none" view (Cohen et al.,

Three experiments were conducted to determine the effect of methamphetamine on the performance of rats in two timing tasks. When food sometimes followed the first response after T sec of a signal, the response rate increased to a peak near T sec and then declined. Methamphetamine decreased the time of the peak (Experiments 1 and 2). When one response (called a "short response") was reinforced following a short signal and a different response (a "long response") was reinforced following a long signal (where the short and long signals were 1 and 4, 2 and 8, or 4 and 16 sec), the probability of a long response increased as a function of signal duration. The point of indifference (50% long response) occurred near the geometric mean of the extreme durations, and methamphetamine decreased the point of indifference by about 10%. These results suggest that methamphetamine increases the speed of an internal clock used by rats in time discrimination tasks.

No abstract

Three experiments studied the time discrimination of rats. The purpose of the experiments was to determine some properties of the internal clock being used. Experiment 1 used a fixed-interval procedure, with 60-sec intervals. After the time discrimination was established, occasional trials contained a 15-sec break (lever withdrawn, noise on). Experiment 2 used a choice procedure, with signal durations from 4 to 22 sec. After the time discrimination was established, occasional signals contained a 2-sec or 4-sec break (light off, noise on). Experiment 3 used a fixed-interval procedure, with 30-sec intervals (signaled by light) and 60-sec intervals (signaled by sound). After the signal and time discriminations were established, occasional trials began with the 30-sec signal but later shifted to the 60-sec signal. Taken together, the results suggest that the rat's clock has many of the qualitative properties of a stopwatch: It can be stopped temporarily (Experiments 1 and 2) ; time before stopping can be added without loss to time after stopping (Experiment 2) ; it times signals from different modalities (Experiment 3); it times intervals of different lengths (Experiment 3) ; it times intervals of different lengths using the same rate (Experiment 3); and it times up (Experiment 3).With fixed-interval (FI) reinforcementwhere food is primed a fixed time after food is given-a rat pressing a lever responds more often as the time of food approaches (e.g., Meltzer & Brahlek, 1970;Sherman, 1959;Skinner, 1938). With the related choice procedure-where the correct choice depends on the duration of a signal-a rat's performance can be much better than chance (e.g., Church, Getty, & Lerner, 1976;Cowles & Finan, 1941;Heron, 1949). Both results show that rats can discriminate time, and thus suggest that rats have a "clock" (some-