Behavioral Momentum - Resistance To Change

Resistance To Change

According to behavioral momentum theory, there are two separable aspects of the discriminated operant that independently govern the rate at which a response occurs and the persistence of that response in the face of operant disruption such as punishment, extinction, differential reinforcement of alternative behaviors, etc. (see Nevin & Grace, 2000, for a review). The positive reinforcement contingency between the response and a reinforcing consequence governs response rates (i.e., a response–reinforcer relation) by shaping a particular special pattern of responding, according to the relative law of non-absolute effect (i.e., the matching law; Herrnstein, 1970). Conversely, the Pavlovian relation established between the context (as a discriminative stimulus) and the rate or magnitude (but not both) of reinforcement obtained in the context (i.e., a stimulus–reinforcer relation) governs the resistance of the behavior class to behavioral operations such as implementing an extinction contingency. Resistance to change is assessed as responding during conditions of environmental disruption (e.g., satiation, extinction) relative to stable, pre-disruption behavior response rates.

Resistance to disruption has been considered a superior measure to stable response rates as an expression of the underlying strength of a response (Nevin, 1974). This is because drastic differences in response rates can occur by reinforcement contingency shaping particular patterns of behavior (e.g., differential-reinforcement-of-high- or low-response-rate schedules) while reinforcement rates are equal. It is unclear, therefore, whether these differences in response rates necessarily indicate differences in the underlying strength of a response (see Morse, 1966, for a discussion).

According to behavioral momentum theory, the relation between response rates and resistance to change is analogous to the velocity and mass of a moving object, according to Newton's second law of motion (Nevin, Mandell & Atak, 1983). Newton's second law states that the change in velocity of a moving object when an outside force is applied is directly related to that force and inversely related to the object's mass. Similarly, behavioral momentum theory states that the change in response rates under conditions of disruption (Bx) relative to baseline response rates (Bo) are directly related to the force or magnitude of disruption (f) and inversely related to the rate of reinforcement in a stimulus context (r):

(1)

The free parameter b indicates the sensitivity of resistance to change to the rate of reinforcement in the stimulus context (i.e., the stimulus–reinforcer relation). Resistance to disruption typically is assessed when two distinctive discriminative stimulus contexts alternate and signal different schedules of reinforcement (i.e., a multiple schedule). Equation 1 can be rewritten to account for resistance to change across two stimulus contexts (Nevin, 1992; Nevin, Grace, & McLean, 2001) when a disrupter is uniformly applied across contexts (i.e., f₁ = f₂):

(2)

The subscripts indicate the different stimulus contexts. Thus, Equation 2 states that relative resistance to change is a power function of the relative rate of reinforcement across stimulus contexts, with the a parameter indicating sensitivity to relative reinforcement rate. Consistent with behavioral momentum theory, resistance to disruption often has been found to be greater in stimulus contexts presenting higher rates or magnitudes of reinforcement (see Nevin, 1992, for a review). Strong support for resistance to change being determined by stimulus–reinforcer relations and independent of response–reinforcer relations comes from studies that add response-independent (i.e., free) reinforcement to one stimulus context. For instance, Nevin, Tota, Torquato, and Shull (1990) had pigeons pecking lighted disks on separate variable-interval 60-s schedules of intermittent food reinforcement across two components of a multiple schedule. Additional free reinforcers were presented every 15 or 30 s on average when the disk was red, but not when the disk was green. Thus, the response–reinforcer relation was degraded when the disk was red because each reinforcer was not immediately preceded by a response. Consistent with the matching law, response rates were lower in the red context than in the green context. However, the stimulus–reinforcer relation was enhanced in the red context because the overall rate of food presentation was greater. Consistent with behavioral momentum theory, resistance to presession feeding (satiation) and discontinuing reinforcement in both contexts (extinction) was greater in the red context. Similar results have been found when reinforcers are added to a context by reinforcing an alternative response.

The findings of Nevin et al. (1990) have been extended across a number of procedures and species including goldfish (Igaki & Sakagami, 2004), rats (Harper, 1999a, 1999b; Shull, Gaynor & Grimes, 2001), pigeons (Podlesnik & Shahan, 2008), and humans (Ahearn, Clark, Gardenier, Chung & Dube, 2003; Cohen, 1996; Mace et al., 1990). The behavioral momentum framework also has been used to account for the partial-reinforcement extinction effect (Nevin & Grace, 1999), to assess the persistence of drug-maintained behavior (Jimenez-Gomez & Shahan, 2007; Shahan & Burke, 2004), to increase task compliance (e.g., Belfiore, Lee, Scheeler & Klein, 2002), and to understand the effects of social policies on global problems (Nevin, 2005).

Although behavioral momentum theory is a powerful framework for understanding how a context of reinforcement can affect the persistence of discriminated operant behavior, there are a number of findings that are inconsistent with the theory (see Nevin & Grace, 2000, and accompanying commentary). For instance, with equal reinforcement rates across stimulus contexts, resistance to change has been shown to be affected by manipulations to response–reinforcer relations, including schedules that produce different baseline response rates (e.g., Lattal, 1989; Nevin, Grace, Holland & McLean), delays to reinforcement (e.g., Bell, 1999; Grace, Schwendimann & Nevin, 1998; Podlesnik, Jimenez-Gomez, Ward & Shahan, 2006; Podlesnik & Shahan, 2008), and by providing brief stimuli that accompany reinforcement (Reed & Doughty, 2005). Also, it is unclear what factors affect relative resistance to change of responding maintained by conditioned reinforcement (Shahan & Podlesnik, 2005) or two concurrently available responses when different rates of reinforcement are arranged within the same context for those responses (e.g., Bell & Williams, 2002).