Age-Related Slowing in Movement Parameterization Studies: Not what you Might Think

Paul C. Amrhein*    University of New Mexico
* Correspondence should be sent to Paul C. Amrhein, Department of Psychology, University of New Mexico, Logan Hall, Terrace and Redondo Streets, NE, Albuquerque, NM 87131, U.S.A. email address: amrhein@hydra.unm.edu

INTRODUCTION

One of the staple, if not classic, methodologies used to study the effects of aging on human performance has been the reaction time (RT) task (see, e.g., Salthouse, 1985; Welford, 1959, 1977; Spirduso & MacRae, 1990). In particular, two reaction time tasks have been used extensively: Simple reaction time (SRT) and choice reaction time (CRT). Based, respectively, on Donders' (1869/1969) Type A and B tasks, they provide a means to separately assess age effects on sensory-motor and intervening cognitive processes (Dawson, 1988; Teichner & Krebs, 1974). As such, they provide a useful way to assess at a process level the pervasive response slowing seen in older persons (see Botwinick, 1984; Goggin & Stelmach, 1990; Welford, 1977).

In the SRT task, a pre-specified stimulus is presented and the subject responds with a pre-instructed response. (A variant of this task is where the stimulus is presented but subjects respond upon a latent “GO” signal; for example, see the delayed pronunciation task of Balota & Duchek, 1988.) By knowing the stimulus and the response to it, subjects are likely to prepare this response prior to actually receiving the stimulus (or “GO” signal). In a typical CRT task, subjects respond to one of a number of stimuli with a pre-instructed response unique to each potential stimulus. SRT and CRT tasks share perceptual and motor aspects in their task demands; that is, in both tasks (excluding the latent “GO” signal version), subjects must detect that a stimulus has been presented, and (including the latent “GO” signal version) the corresponding response must be prepared and executed. What distinguishes SRT and CRT tasks is the uncertainty concerning which stimulus is actually presented. Whereas there is no stimulus uncertainty for the SRT task, there is for the CRT task. As numerous studies have reported over the years, increases in this uncertainty yield corresponding increases in response latency across the adult lifespan (see, e.g., Kausler, 1991; Salthouse, 1985; Welford, 1959, 1977).

Many motor performance tasks are built upon SRT and CRT task methodologies. Indeed, SRT and CRT tasks typically require a manual (i.e., aimed movement) response. In the SRT task, response parameters (concerning which finger, hand, arm, foot or leg will be used) are prepared by the subject prior to target stimulus onset (see e.g., Amrhein, Stelmach, & Goggin, 1991). In the CRT task, by contrast, such preparation does not appear to occur (Amrhein et al., 1991; Klapp, Wyatt, & Lingo, 1974). Thus, SRT and CRT tasks actually represent two extremes on the scale of response preparation, and as such represent useful reference points when studying movement plan preparation, maintenance, restructuring and execution. While most SRT and CRT studies have assessed response initiation (reaction time, RT) and execution (movement time, MT) in aggregate (but still refer to the data as “RT” even though it might be better referred to as “Total Time”, TT), there have been some studies which have used RT/MT assessment.

Methodologically, what distinguishes the larger set of “RT” from the smaller set of “RT/MT” studies is that subjects in the former set simply press a target button upon stimulus response, often with little experimental control over the initial resting location of their responding body part, whereas in the latter set, upon stimulus presentation, subjects release a button (often called a “Home button” or “Home key”), and then move to press a target button. In the aging literature, these RT/MT studies include: Amrhein et al. (1991), Amrhein, Von Dras, and Anderson (1993), Clarkson (1978), Goggin, Stelmach, and Amrhein (1989), Larish and Stelmach (1982), Spirduso (1975), and Goggin (1988), and Amrhein (1988), and Garcia-Colera (1987), Singleton (1954), Szafran (1951), and Welford (1959, 1977). Generally, these studies have revealed slower RTs and MTs for older (e.g., age range 50-87 years) relative to younger (e.g., age range 18-31 years) individuals. However, many of these studies failed to separate the role of visual guidance from motor performance. That is, subjects in the other studies were allowed to use vision to guide their movement responses. This is not a trivial problem; there is a sizable literature which documents perceptual-motor interaction (e.g., see Rosenbaum, 1991; Szafran, 1951). Accordingly, I will constrain the scope of this chapter to cover only those studies where the role of visual perception is limited to stimulus processing. Moreover, each of the studies reviewed assessed healthy, community dwelling elderly (age range 63-80 years) and young (age range 18-31 years) individuals. Also, subjects in these studies received sufficiently numerous trials to allow an assumption that both subjects attained their respective asymptotic levels of practice on the various tasks (see, e.g., Spirduso & MacRae, 1990, concerning differential practice effects preceding asymptotic performance).

Two popular aimed movement tasks in the aging literature are the movement plan specification and restructuring tasks. Common to both tasks is the manipulation of movement parameters such as arm, direction, extent, velocity, force, etc. Such parameters take on values which are specific to a generalized motor program that defines a particular pattern of physical activity (Schmidt, 1988). Latency to initiate (RT) and execute (MT) a planned movement is assessed, as well as errors which may occur for movement initiation and execution. Overall, both tasks have exhibited age-related slowing like that seen for the SRT/CRT tasks; this is not surprising, because these movement tasks also manipulate stimulus uncertainty in a fashion similar to SRT and CRT tasks (Amrhein et al., 1991). However, the age-related slowing observed in these movement tasks has rarely received statistical analysis beyond experiment-specific determination of elderly/young latency ratios (see Goggin & Stelmach, 1990). As will be detailed later, regression analysis across-experiments may offer a more detailed assessment of the loci of this slowing, specifically concerning “sensory-motor” and intervening “computational” processes (Cerella, 1990). Details of each task are presented below.

Movement plan specification task

The movement plan specification task involves a precuing paradigm in which the subject is given partial or complete parameter information concerning the impending movement response (e.g., use of the left arm in a movement away from oneself). RTs indicating additional latency to respond to a stimulus, relative to that for a completely specified response, are related to the parameters left to be specified for that stimulus after preceding partial specification of the other parameters by the precue stimulus (e.g., left arm is specified but direction – away or toward oneself – is...