Activity Space Analysis
- Introduced by
- Reginald Golledge & Robert Stimson
- Year
- 1997
- Type
- Measure of the spatial extent of an individual's routine activities
- Family
- Behavioural and movement geography
What does it do?
Activity space analysis measures the geographic area within which an individual moves and carries out their routine daily activities — home, work, shopping, leisure — and the travel that links them.
By delineating this lived spatial footprint from observed visit locations, it reveals how far and in what directions people actually range, and what environments they are exposed to in the course of ordinary life. It bridges the behavioural geography of Golledge and Stimson with modern mobility and health research that links where people go to the contexts they encounter.
The intuition behind the method
Pin every place you visited over a week onto a map — home, the office, the gym, the children's school, the supermarket — and you get a scatter of dots that is usually not a tidy circle but an elongated smear stretched along the routes you travel most. The cloud is denser where you spend more time and thinner at the edges you visit rarely. Drawing a shape that captures that cloud — an ellipse tilted along your main axis of movement, or a smooth density surface — summarizes the slice of the world you actually inhabit, which is typically far smaller and more lopsided than the whole city you nominally live in.
History
The concept of an activity space grew out of mid-20th-century behavioural geography, where the action space and awareness space described the parts of the environment an individual knows and uses; Reginald Golledge and Robert Stimson synthesized this tradition in their 1997 text Spatial Behavior: A Geographic Perspective. Early measures relied on travel diaries and the standard deviational ellipse, long used to summarize directional point patterns. The spread of GPS and smartphone tracking in the 2000s made fine-grained activity points abundant, and kernel-density and network-based delineations became standard. Recent health and segregation research has used activity spaces to define mobility-based, realized exposures, reviving the behavioural-geography insight that residence alone poorly describes the contexts people inhabit.
When to use it
Use activity space analysis when the question concerns where individuals actually go and what they are exposed to, rather than only where they live — for studies of mobility, segregation experienced through daily movement, access to amenities, or environmental and health exposures. It suits data made of geolocated visits per person, from travel diaries or GPS, and is the right tool when residential-neighbourhood measures are too coarse because people spend much of their day elsewhere. It is less appropriate when only aggregate or place-based patterns matter, when activity locations are unobserved, or when the analytical interest is in feasibility and timing, where the space-time prism of time geography is more exact.
Limitations and considerations
- The standard deviational ellipse imposes a smooth, symmetric shape that misrepresents multi-centred or linear activity patterns.
- Results depend on choices of kernel bandwidth, ellipse standard-deviation multiplier, and density threshold.
- Quality hinges on how completely and accurately the underlying activity points are observed.
- Aggregating points over a period blurs the timing of activities, losing the temporal sequencing that time geography retains.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Captures the real, mobility-based extent of daily life rather than assuming people stay near home.
- Offers a graded toolkit — ellipses for a quick summary, kernel density and network buffers for detail.
- Enables realized-exposure measures that improve on static residential-neighbourhood definitions.
- Summarizes complex movement into interpretable shape, size, and orientation metrics for comparison.
Applications
- Estimating individuals' realized exposure to greenspace, air pollution, or unhealthy food environments.
- Measuring experienced segregation through the diversity of places people actually visit.
- Comparing the size and orientation of activity spaces across age, gender, or income groups.
- Assessing how mobility limitations shrink the daily reach of elderly or disabled individuals.
- Defining mobility-based service catchments for transport, retail, and public-health planning.
How it works
How it works
The analysis begins by assembling, for each individual, the set of geolocated points where they conduct activities or that they visit — collected from travel diaries, surveys, or GPS traces. Each point may carry attributes such as activity type, duration, and time of day. This point set is the raw evidence of where the person's life actually unfolds and is the input to every subsequent delineation.
The simplest summary is the standard deviational ellipse, centred on the mean centre of the activity points with semi-axes scaled from the standard deviations of the x and y coordinates and rotated by an angle θ that aligns the long axis with the dominant direction of spread. The ellipse captures the size, elongation, and orientation of the activity space in one interpretable shape, showing both how widely and along what axis a person ranges.
Formula
\sigma_x = \sqrt{\tfrac{1}{n}\sum_i (x_i - \bar{x})^2},\quad \sigma_y = \sqrt{\tfrac{1}{n}\sum_i (y_i - \bar{y})^2},\quad \theta = \text{orientation}A richer delineation builds a kernel density estimate over the activity points, placing a smooth kernel of bandwidth h on each location and summing them into a continuous intensity surface that is high where activity concentrates and low at the margins. Thresholding this surface — for example at a chosen density contour — outlines the activity space as one or several patches, capturing irregular, multi-centred shapes that a single ellipse cannot.
Formula
\hat{f}(s) = \frac{1}{n h^2}\sum_{i=1}^{n} K\!\left(\frac{s - s_i}{h}\right)Once the activity space is delineated, it is intersected with environmental or contextual layers — greenspace, air pollution, food outlets, deprivation — to estimate realized exposure: the contextual conditions an individual actually encounters, weighted by where within the space they spend time. This moves beyond residential-neighbourhood measures, which assume exposure occurs only at home, to capture the full mobility-based context of daily life.
Formula
E = \frac{\int_{AS} w(s)\,c(s)\,ds}{\int_{AS} w(s)\,ds}Common pitfalls
- Reporting a one-standard-deviation ellipse as if it captured the full activity space, when it covers only the central spread.
- Choosing a kernel bandwidth arbitrarily, which can either over-smooth distinct activity nodes or fragment them.
- Using Euclidean ellipses and buffers where travel follows a road network, overstating reach across barriers.
- Inferring exposure from the activity space without weighting by time spent, treating a fleeting pass-through like a daily destination.
Frequently asked questions
What is the difference between an activity space and a space-time prism?
An activity space is observed: it describes where a person actually went, delineated from their visited locations. A space-time prism is potential: it describes everywhere a person could have reached given fixed anchor points and a time budget. Activity space answers 'where do they go?', while the time-geographic prism answers 'where could they go?', so the activity space typically lies inside the prism that bounds it.
Why use a standard deviational ellipse to summarize an activity space?
The ellipse condenses a scatter of activity points into three interpretable numbers: a centre, a size from the coordinate standard deviations, and an orientation angle showing the dominant axis of movement. It quickly reveals whether someone ranges widely or narrowly and in what direction, and it is easy to compare across people. Its weakness is that it assumes a single, smooth, symmetric cluster, so kernel density is preferred for irregular, multi-centred patterns.
How does activity space improve exposure measurement in health research?
Traditional measures assume people are exposed only to the conditions of their residential neighbourhood, ignoring that much of the day is spent elsewhere — at work, school, or while travelling. By delineating the full activity space and intersecting it with environmental layers, researchers estimate realized exposure across all the places a person frequents, weighted by time spent, yielding a more accurate account of the contexts that influence health behaviours and outcomes.
Related methods
Sources
Golledge, R. G., & Stimson, R. J. (1997). Spatial Behavior: A Geographic Perspective. Guilford Press, New York.
ISBN: 9781572300507Copied.