VOLUNTEERED GEOGRAPHIC INFORMATION: AN ASSESSMENT OF DATA QUALITY IN MALAWI

Abstract

Volunteered Geographic Information (VGI) is touted as the modern-day approach to spatial data acquisition. Around the globe, VGI is slowly being integrated with authoritative spatial data while completely replacing it in other cases. However, the issue of quality in VGI remains combative, mainly from a lack of specialised skills of its contributors. VGI has great potential in revolutionising spatial data sourcing, hence the need to establish adequate confidence in the status of VGI quality to boost its uptake, particularly in Malawi. This interpretive philosophical study that followed a mixed-methodology aimed at ascertaining the status of VGI quality in Malawi through analysis of VGI expert user experiences and systematic quality testing of VGI. Interviews, questionnaires, and online spatial data repositories as secondary data sources were used for data collection. Narrative analysis extracted themes from narratives of expert user experiences with VGI quality. Analytical Hierarchical Processing (AHP) was used to rank six key data quality dimensions in the order of priority to determine a quality dimension on which VGI quality was tested. As a result, the OpenStreetMap (OSM) Education Facilities dataset for Malawi was tested for horizontal accuracy against the Malawi Ministry of Education Primary Schools dataset. The test used the National Standard for Spatial Data Accuracy (NSSDA) methodical approach to estimate positional accuracy and the results were validated using Euclidean distance buffers as an application of Horizontal Positional Error (HPE). The study observed that VGI suffers from presumed notions of lack of quality primarily attributed to misinformation about VGI quality, however VGI used by the expert users in their practice surpassed their quality expectations. It was also discovered that VGI expert users prioritised accuracy among key data quality dimensions. From the quality test, the OSM Education Facilities dataset for Malawi was found to have a Circular Standard Error (CSE) of 19.20 metres and an NSSDA of 47.0150 metres for featured Primary schools. This study found VGI to be within tolerable levels of accuracy when compared against authoritative spatial data.