GISSA discusses spatial statistics and professional registration

March 16th, 2018, Published in Articles: PositionIT

The Geo-Information Society of South Africa’s (GISSA) Gauteng branch met on 23 February 2018 in Pretoria to continue the professional registration discussion and share GIS case studies and experiences.

The professional registration discussion followed a debate format, with the University of Pretoria director of the Centre for Geoinformation Science, Prof. Serena Coetzee, and City of Ekurhuleni’s director of Geo Informatics, Morena Letsosa, arguing in favour of registration, while Statistics South Africa’s Rob Anderson and Marc Leroy from the Gauteng Department of Rural Development and Land Reform argued against it. After hearing the main arguments, the floor was opened to input from members.

The debate follows the branch’s previous meeting in 2017, in which four of the five working groups tasked with outlining areas of expertise for job reservation questioned the need for professional registration.

Marc Leroy, Rob Anderson, Linda Tshabalala, Samy Katumba, Mimi Chauke, Dr. Arul Naidoo and GISSA Gauteng chairman Sam Osei.

Marc Leroy, Rob Anderson, Linda Tshabalala, Samy Katumba, Mimi Chauke, Dr. Arul Naidoo and GISSA Gauteng chairman Sam Osei.

In debate, Letsosa argued that there was a need for professional registration as the decisions taken by members in the profession could have legal implications. He also pointed out that Section 16 of the Geomatics Profession Act makes provision for professional registration that has been expanded to include other geospatial professionals in addition to land surveyors.

Anderson urged delegates to bear the purpose of professional registration in mind, namely to protect the general public. He also said that the skills outlined for job reservation in previous meetings (such as data management) were not unique to geospatial professions, and while Section 16 makes accommodation for professional registration, it does not make it compulsory.

Leroy further cautioned that job reservation could create silos and limit collaboration by excluding other professions (such as botanists and others) who use GIS as tools in their quest to better understand the world, and on whom GIS professionals often rely in conducting their work. Anderson added that the very accessibility that led to the widespread adoption of GIS could be threatened by professional registration. Furthermore, he said, professional registration is not necessarily a sign of competency, but a legal instrument to enforce responsibility.

Prof. Coetzee, who argued in favour of job reservation, questioned whether GIS professionals are the best drivers of professional registration, which she warned might be interpreted as a tool for exclusion. She used the example of how job reservation by town planners backfired when it was challenged by the Competition Commission of South Africa and deemed anti-competitive.

Audience members shared the same mixed sentiment, and the debate concluded that the current outline and basis for professional registration is problematic and needs refinement around the types of job and role for reservation.

The day also included GIS presentations and an update from the Committee for Spatial Information’s (CSI’s) work and progress. Two of the presentations illustrated the merging of statistics with geography.

Dr. Arulsivanathan Naidoo, a statistician at Statistics South Africa, showed how he used Esri’s Space Time Cube add-in for ArcGIS Pro to analyse schools’ pass rates for trends which could aid further investigations and interventions. The visualisation add-in tool brings a time component to spatial relationships, allowing for deeper insight to be derived and more detailed trend analyses of data.

Using the add-in for defined locations allowed data to be aggregated at school level. Dr. Naidoo analysed ten years’ pass rates data using Getis-Ord statistics, which looks at each feature within the context of neighbouring features, to identify hotspots – areas where features with either high or low values cluster spatially. He then used Mann-Kendall statistics for trend analyses. Together, these tools and methods allowed him to mine patterns for new and consecutive hotspots (failure rates for two consecutive years), intensified; persistent; diminishing; sporadic and oscillating hotspots.

Samy Katumba from the Gauteng City Region Observatory (GCRO) also demonstrated the value of spatial statistics in assessing the spatial extent and directional trends of inequality in Gauteng, based on the research unit’s Quality of Life Survey data. Using the Gauteng Multidimensional Poverty Index (GMPI) – which is based on Stats SA’s MPI – in a GIS model, Katumba identified both where the poor are located and the intensity of poverty. He found that poverty is often located on the periphery of cities. By comparing data for different years, the movement and directional changes in poverty can be determined. This is important in evaluating public policy aimed at elevating poverty and inequality, and also to identify areas for intervention.

The day concluded with a presentation by the Department of Rural Development and Land Reform’s (DRDLR) Mimi Chauke, who provided a progress update on the Committee for Spatial Information’s (CSI) work and spatial data infrastructure. She addressed two policies: the pricing of data, and base dataset custodianship. The SASDI compliance guidelines have been finalised and approved, and the first compliance audit for custodianship is being conducted on the DRDLR. Chauke also said that the EMC is functional (www.sasdi.net) and those interested can speak to Prof. Coetzee about advanced metadata training. The SDI Amendment Bill is also in the works and will be gazetted soon. The CSI has also published the national land classification standard.

A photo gallery from the event can be viewed here.

Send your comments to positionit@ee.co.za