Spatial Thinking in Planning Practice: An Introduction to GIS

40 CHAPTER 6: TO STANDARDIZE OR NOT TO STANDARDIZE? "e idea of data quality and standards are important especially in the urban planning !eld because we make decisions based on data collected from di#erent institutions. If there is not a common standard to follow, it will be frustrating to work if these data came with di#erent quality. Besides, using low quality data might lead public o&cials and researchers to wrong conclusions, a#ecting the decision-making process. "is chapter by Yeung ex- plains the importance of data quality and data standards, and their inter-relationships. "ere are ways to quan- titatively assess the positional and attribute accuracy of geo-spatial data. If you are interested, you can explore census TIGER !le to see how the1990 !les di#er from thee 2000 dataset. "e chapter starts discussing concepts of geospatial data quality such as accuracy (degree to which data agree with the description of the real world that they represent); precision (how exactly are measured and stored); error (a measure of the deviation between the measured value and the true value), and uncertainty (lack of con!dence in the use of the data due to incomplete knowledge of the data). All these concepts are related to the description and evaluation of data quality. Yeung also discusses the sources and types of errors in geospatial data (inherent and operational errors), which are almost impossible to avoid. In sum, Yeung describes seven dimensions of geospatial data quality: (i) lineage (document the sources from which data is derived), (ii) positional accuracy (it is de!ned as the closeness of values in the database to the true positions of the real world), (iii) attribute accuracy (closeness of descriptive data to the assumed real world values), (iv) logical consistency (describes the !delity of the relationship between real world and encoded data), (v) completeness (refers to whether the data exhausts the universe of all possible items), (vi) temporal accuracy (refers to the representation of time in geospatial data), and (vii) semantic accuracy (measures how correctly spa- tial objects are labeled in the data set). "e positional and attribute accuracy are the most relevant. "e presence of errors is a norm rather than an exception. "us spatial errors need to be managed to reduce uncertainty. "ere are three perspectives to e#ect the management of spatial data errors: (i) data production (control de data quality during the data acquisition), (ii) data use (related to errors when data is used), and (iii) communication between data producer and data user (evaluating the quality of the data so that users are aware of the level of uncertainty). To make sure the quality assurance and quality control of geospatial data is the expected the process of data col- lection needs to be monitor because is the greatest source of errors in digital geospatial data. During the process of geographic analysis there might be an accumulation of the e#ects of errors. "is is known as the error prop- agation. Managing errors requires a pragmatic approach through, for example, sensitivity analysis, which is a modeling technique to assess the subjectivity and variability in the parameters of spatial problem-solving model. "e purpose of the sensitivity analysis is to test the model for output over a range of legitimate uncertainties. An- other relevant aspect is the reporting data quality: information need to be e#ectively communicated in the form of ‘accuracy indices’ and maps to all potential users. Geospatial data standards can provide a yardstick -- created by consensus by a recognized organization -- against which quality can be evaluated, through the provision of rules for common and repeated use. Yueng o#ers four categories of standards: (i) application standards, (ii) data standards –the most important-, (iii) technology standards, and (iv) professional practice standards. In general geospatial data standards provide the means of communication between suppliers and users. "ese are made up of one or more of these four components: (a) standard data products, (b) data transfer standards, (c) data quality standards, and (d) metadata standards. "e development and acceptance of data standards was crucial not only for allowing sharing data but most important, it helped to develop ‘open GIS’. Read the book chapter Albert Yeung. 2007. “Geospatial Data Quality and Standards” in Albert Yeung 2007 Concepts and Techniques in Geographic Information Systems. pp 108-42.

RkJQdWJsaXNoZXIy NTc4NTAz