Page 77 - C:\Users\Evans Moetji\Desktop\DPSA Guideline Digital Flipbook\
P. 77

ANNEXURE 1:  QUALITY CONTROL METHODS



                       C      LOGICAL AND PHYSICAL CONSISTENCY






                       Logical  and physical consistency looks at  whether spatial features are consistent  when
                       it  comes  to  topology, their  extent, spatial  relationship with  other features,  and geographic
                       location.  You need to check the location of spatial features and their related attributes to see
                       whether they consistently and logically make sense.  This can be done by:
                        Producing thematic maps using variables associated with spatial features, including the
                          categorical and numeric fields.

                        Randomly selecting between 20 and 30 points that can be identified in the study area to
                          check consistencies in the target population, government service point and road datasets.

                        Queries of the spatial and attribute databases can also be done to see if the variable values
                          associated with spatial features are correct.



                       D      REFERENTIAL INTEGRITY






                       You need to check how well attribute tables are able to link to spatial features and other tables.  By
                       using a relational join (joining two tables together by matching values in fields that are common
                       to both tables) or geo-coding functions in GIS, you can easily determine the extent that attribute
                       tables are able to link to spatial features.

                       The spatial and attribute tables must have unique identifiers or primary and foreign keys to enable
                       them  to link effectively.    It is important  to have a unique identifier  to distinguish spatial
                       features from one another that may have the same geographic coordinates and to enable the
                       set of attributes associated with spatial features to be linked to them.

                       Consider the following:

                       ❏  The target population dataset should have the unique enumeration area number if you are
                          using census enumeration area data.

                       ❏  Service points should have a unique number that allows information from various divisions
                          in the department (e.g.  finances, human  resources,  operations, beneficiaries,  results) to
                          be linked to the  spatial features.    A good example of this is the Education Management
                          Information System (EMIS) number of the Department of Education.

















                                                                                                                    72
   72   73   74   75   76   77   78   79   80   81   82