Physical activity measurement: Data collection methods ShareThis

< Go to guide main menu

Physical activity can be measured by self-report and/or objectively. In practice, both are likely to provide complementary value to large-scale studies. The common methods for data collections are outlined below.


Self-report physical activity questionnaires have been a staple of population-based research for many decades due to their practicality and low-cost [74]. A variety of self-report questionnaires exist with some focussing on assessing recent physical activity trends and others that attempt to classify lifetime activity levels [11]. The International Physical Activity Questionnaire (IPAQ) is the most widely used self-report method [75]. It contains 31 questions (9 in the short-form version) on how much time people spend sitting, in light activity (e.g. walking), moderate activity (e.g. leisure cycling), and vigorous activity (e.g. running). The reference period is either the past seven days, or in a typical week. Scores on the IPAQ can then be used to estimate total energy expenditure, via converted activities to metabolic equivalents (METs).

Self-report questionnaires have some advantages that extend beyond their practicability, such as their capacity to record the context and perceived intensity of physical activity [74]. For example, by using self-report questionnaires it is possible to categorise physical activity into different domains, which can be used to better understand and promote physical activity behaviour. There are also concerns relating to the reliability of self-report questionnaires; several studies have found large differences between self-reported and objectively measured physical activity [40, 76]. While such differences are commonly attributable to reporting biases in self-reported data, they may also be due to methodological differences between the two measures. Indeed, some self-reported measures correlate highly with directly observed measures [77].



Accelerometers are small electromechanical devices that allow for the objective measurement of physical activity [78]. Accelerometers are usually worn on the hip, wrist or chest and detect incidences of acceleration that are interpreted as bodily movements. Incidences of acceleration are recorded as ‘counts’. The number of counts that are recorded over a pre-specified time epoch, usually one minute, can be used to determine the intensity of activity. Accelerometers are validated with a moderate-to-strong correlation with direct measures of oxygen consumption, such as doubly labelled water and calorimetry [79]. Modern triaxial accelerometers, such as the ActiGraph GT3X+, can record activity across three directional planes and are strongly correlated with direct measures of oxygen consumption, such as gas analysis [80]. Additionally, GPS accelerometery tracking records movement and position in the world, called ‘inertia measurements’. Using these measurements, the tracker can calculate the position and moving pattern [81]. However, the use of these is limited due to governance and ethical reasons.

Accelerometers can be expensive for large-scale studies, but the cost per unit is falling, particularly with the rise in commercial grade accelerometers that also perform well against direct measures of oxygen consumption [82]. However, methods of processing and analysing accelerometer data are highly variable and the devices remain poor at estimating non-ambulatory activities, such as cycling or weight lifting [74].


Learn more about the individual studies covered by this guide and their measurement of physical activity:

Explore the measures by physical activity domain and their cross-study comparability:

Further information:

This page is part of the CLOSER resource: ‘Physical activity across age and study: a guide to data in six CLOSER studies’.