Module 1.3: Data Quality - Assessment

    For this week’s lab assignment, I assessed the completeness of two road datasets, TIGER Roads and Street Centerlines. In the analysis, I determined completeness by which source featured a greater length of road. The analysis loosely follows the steps taken by Haklay 2010. Below are a detailed account of the steps that I took to analyze the two road datasets. 

    I started by clipping the road shapefile to the extent of the grid layer which removed road segments of the road that would not be relevant to the analysis. Then, I used the intersect tool to determine the points where the roads were crossing each of the grids. The output of the intersect tool was then included as an input in the split the line tool which segmented the roads at the intersection point. I set the search radius to 5 feet to account for any error with the intersect point. This ensured that roads that stretched over multiple grids were segmented at the point where they intersected a grid boundary. This step was crucial because otherwise roads that span multiple grids may not be properly sorted in future steps. 

    After segmenting the roads, I spatially joined the segmented roads layer and the grid shapefile using the largest overlap match option. Next, I dissolved the roads by the GridCode field so that all of the roads within a grid would be lumped together as a single attribute. Once dissolved, I recalculated the length of road within each grid. Last, I completed the same steps with the other roads layer and compared the data by joining the tables and calculating the percent difference.

Note: there is likely a more simplistic way to complete this analysis but these were the geoprocessing steps that I took to complete the analysis.

    It is important to note that more data is not always better. For instance, I did not assess the data for duplicated attributes which could influence the results of this analysis. Completeness assessments and other data quality assessments should be conducted in conjugation with other data quality assessments to ensure the utmost quality of the data. Without further assessment of the data, this analysis will not provide a complete picture of the data’s overall quality.