Ozri 2013: Trends and Clusters

ozri-2013-blog-banner.jpg

In geography, the way we map our results (using colours, class breaks, symbology etc.) can change the message that our map communicates. Consumers also interpret maps subjectively; perceiving patterns in maps that may or may not exist. Inappropriate decisions can sometimes be made as a result. At Ozri 2013, Ebony and I presented several techniques to minimise subjectivity in the presentation of spatial information.

I presented on the use of descriptive spatial statistics in ArcGIS. Measuring properties of your data such as the centrality, concentration or orientation of a distribution filters out much of the complexity in interpreting the map, and makes it easier for you to compare populations or track change. For instance, a law enforcement authority may visualise the centre and extent of different graffiti tags to understand the potential home base and transport corridors used by offenders. A health department can track the spread of disease over time.

After describing a distribution, it is important to understand if perceived spatial patterns are statistically significant, or if they could have come about by chance. Ebony explained how a new tool in ArcGIS 10.2, Optimised Hot Spot Analysis, makes this easier. Hot Spot Analysis works be examining the value for each feature in the context of neighbouring feature values; effectively identifying where high values are surrounded by other high values (hot spots), and where low values are surrounded by other low values (cold spots). The analysis not only identifies the hot and cold spots, but also the probability that they could have arisen by chance, which minimises subjectivity in the interpretation of the map. For instance, health authorities may wish to objectively understand patterns in diabetes in order to direct funding to the areas of greatest need.

Peter - ObesityHotspotAnalysis

The ultimate goal of geographic analysis is actionable intelligence. Understanding the spatial processes that cause the patterns in our maps provides that intelligence. For instance, if a health authority could determine the influence that income, ethnicity, access to healthy food, concentration of fast food, access to parks and exercise all have on the incidence of diabetes – they could take appropriate remedial action. Ebony demonstrated how regression analysis in ArcGIS can be used to model spatial relationships, better understand the factors behind observed spatial patterns, and predict outcomes based on that understanding.

Peter C

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s