Visit the National Academies Press online and register for

Educating the Student Body
data. The instruments avoid the subjective weaknesses of self-reported 
data but have their own problems. First, they must be properly calibrated 
to assess the existence and force of any movement. Moreover, accel-
erometers are reasonably good at detecting and quantifying common 
ambulatory movements (e.g., walking, running) but are not as good 
for stair climbing; cycling; water activities; or activities primarily of the 
upper body, such as sweeping or raking. In addition, the instruments 
cannot determine exactly what activity is being done, where, or why. The 
cut-points used to place movements in the common intensity categories 
(sedentary, light, moderate, vigorous) also may over- or underestimate 
the volume of activity depending on the predominant physical activity 
behaviors of the subjects under investigation (Matthews et al., 2008; 
Troiano et al., 2008), and they cannot assess relative intensity. Still, 
instruments measuring physical activity continue to improve and are 
becoming less costly, and they are useful across the full age spectrum. 
They appear to be the best method for collecting information about 
activities in the low and middle ranges of rate of energy expenditure and 
are expected to become even better with more research and experience.