Table 1 Additional POSTDAT items and derived methods

Activities

Any activity space present

A binary variable indicating the presence of any activity space (Yes/No) (from a list of 12 different sports or active recreation spaces captured on PSDAT) was computed. Playground equipment was excluded.

Environmental quality

Number of trees present in park (Replaces POSDAT item: Estimate the approximate number of trees present 0, 1–50 or 50–100)

All tree canopies within the parks were manually digitised in ArcMap and the number of trees within each park calculated. For the purpose of this study, the number of trees were placed into two categories <25 trees or >25 trees.

Tree canopy area

All digitised tree canopies within each park were merged in ArcMap to determine a total tree canopy area for each park.

The proportion of the park area covered by tree canopy

The total area of tree canopy was calculated as a proportion of total park area (m²). Total tree canopy area÷park size.

Path shade

Original POSDAT categories (no pathshade, very poor, poor, medium, good, very good) were collapsed into two categories: High (medium, good and very good) and low (no shade, poor and very poor).

Amenities

Amenity count

The number of different amenities present (includes; barbeques, seating, picnic tables, toilets, public art, lighting) were summed to give a total count per park.

Safety

Park surrounded by minor roads only Replaces POSDAT item ‘Are all roads surrounding the POS minor roads or cul-de-sacs?’

Road types were objectively categorised (major or minor) using a classification of major and minor roads according to Western Australian Planning Commission policy and local municipality classification.

All roads were attributed as major or minor. All roads surrounding each park were identified Parks were coded as being surrounded by minor roads only (Yes/No).

Number of lots surrounding park

All residential lots (cadastre obtained from Landgate, dated 2006) surrounding the parks (i.e., within a 25 m buffer of the park perimeter) were identified and a count of lots surrounding each park computed. [36].

Number of lots orientation ratio

Each of the selected residential lots were inspected and classified according to the orientation of the dwelling on the lot to the park it surrounded; whether it was orientated towards (i.e. facing/fronting) or away from (i.e. backing onto) the park. [36]. The number of residential lots facing the park was divided by the total number of residential lots surrounding the park. Higher ratios (towards 1.0) indicated a higher proportion of the bordering houses being orientated towards the park [36].

Perimeter orientation ratio

The lot orientation ratio was further refined to determine the proportion of the park perimeter that was bordered by lots fronting the park. The length of the park perimeter surveilled by facing cadastre was identified. This was divided by the total perimeter of the park [36].

Environmental quality

Greenness

The presence of greenness in each park was calculated using the Extract Normalized Difference Vegetation Index (NDVI) tool and Landsat TM remote sensing imagery (summer 2006). NDVI provides an indication of the presence and condition of green vegetation with values ranging from −1 to 1. Values of −1 generally represent water, while values of zero correspond to bare surfaces such as rock, sand and roads. Higher values (0.2 to 0.4) represent grassland or bushland. Mean NDVI values for each park were determined.

Additional items

Park size (m²)

Using the calculate geometry function in ArcMap the area of each park was determined.

Proximity to beach

Distance calculated between the closest point on park perimeter and closest beach access points manually digitized at sites of beach entry. Distances were calculated using the road/pedestrian network.

No. of participants in 800 m buffer

Points were generated at 10 m intervals around each park and a service area generated extending 800 m from each from each point along the road/pedestrian network . The service area around each of the points was dissolved and the number of participant residential points in the service area was determined.

Giles-Corti adult attractiveness score

The attractiveness score previously used by Giles-Corti et al. (ref) was applied to each park.