Data quality components

Road accident data quality can be improved by considering the components that constitute quality data as relatively mentioned in section 0 and subsection 2.5.3 respectively. These components are used as directives towards a better way of attaining procedural steps for quality data assessment. They enhance the possibility of identifying the productive way of acquiring quality data in the process of assembling data, whereby paving the way for the possibility of identifying the particular areas that are problematic towards the acquisition of data.

According to statement made by James O’ Day (1993) in a science magazine, “from a quality point of view, accuracy and completeness are probably more important than timeliness, but excessive processing time should be avoided” (O’ Day 1993). On the contrary, from my personal viewpoint, accuracy and completeness as quality characteristics, should be built on timeliness. Simply because the expected outcome from the implementation of these two quality characteristics along the evaluation protocols depends on the time taken to process the data. This illustration could be used as a yardstick for determining the competence of the reporting officers. This could improve the rate at which the completed ARFs get to the DCO for further evaluations. Although, the productivity rate of an information system can be measured by the degree of quality of data produced and the satisfactory comments of the data users (Wang &

Strong 1996). In addition, an illustrative flowchart showing the relational connectivity between the quality components is shown in Figure 75_.

 Completeness of reportage –this is the primary phase, wherein the evaluation process is ensued to ascertain the relevance of the data collected regarding all specified cases through the data collection method established (O’ Day 1993; Vogel & Bester 2004; Sinclair 2011). Considering this aspect of quality component, the data supervisory officers, in charge of inspecting the completed ARF queries the experience of reporting officer regarding the application of the ARF, provided that errors are discovered.

 Consistency of reportage –this phase complements the evaluation performs in the first phase. The action performs here is basically to identify the factors that affect the degree of reportage (O’ Day 1993; Sinclair 2011), which could be varied by jurisdiction, time, personal characteristics, weather, and other unspecified factors (O’ Day 1993), to ascertain the flexibility and reliability of the database (Vogel & Bester 2004).

 Missing data –the activities perform in this phase strengthen the assurance of knowing the actuality of the problems or factors frustrating the consistency of accident reporting system, by identifying the probable causes of missing data attributing to the negligence of the officers executing the assessment of completeness of the reportage (O’ Day 1993; Vogel & Bester 2004). Missing data can be described as unavailable data for analysis, which is classified important to the real-world scenarios (Batini et al. 2009). According to article published by Paolo Atzeni & Valeria De Antonellis in (1993), as cited by Batini et al (2009), “a data or value can be missing either because it exists, but it is not known, or because it does not exist, or because it is not known whether it exists” (Batini et al. 2009).

 Consistency of interpretation –this is principally one of the essential parts as regards application of the quality components. This certify the correlation of accident reportage across the authorised locals and provinces (Vogel & Bester 2004; O’ Day 1993), to ascertain whether the accident coverage is reported in a similar way, across the localities or regions by different reporting officers regarding the basics information required as instructed in the ARF (O’ Day 1993).

 Right data –this actually paves the way for better understanding of the analysis of the data, if only the right data are provided for actual purpose used for (Vogel & Bester

5 _{The contents in the diagram illustrated in the Figure 7 are adapted from James O’ Day (1993), Liesel Vogel & Christo Johannes}

2004). An illustration is provided by James O’ Day (1993) to support the understanding of this particular data quality component, which states as thus “for the road accident data to be most useful for Safety Authorities, vehicle characteristics is important; for the Highway Authorities, roadway details are important; for the Psychologists, personal details of the driver are important” (O’ Day 1993).

 Appropriate degree of detail –this is useful in the area of broad analysis, wherein some variables or data elements need to be more detailed, so as to achieve an improved compilation and dissemination of information (Vogel & Bester 2004). Actually, this is dependent on the variables and the questions inquired. For instance; for Weather

Conditions and Visibility, variables such as Clear, Overcast, Rain, Mist/Fog, Hail/Snow,

Dust and others like Fire/Smoke, Severe Wind, and Unknown are relevant (O’ Day 1993). Completeness of reportage Missing data Consistency of reportage Consistency of interpretation

Right data Appropriate degree of detail

Correct entry procedures Freedom from response error

Figure 7: The relational connection between the quality components

 Correct entry procedures –this aspect scrutinises the procedure followed in the preceding phases. Here, further thorough assessment on the information provided in the ARF is executed, in order to ascertain the degree of quality of the information [data] inputting into the data set (O’ Day 1993; Vogel & Bester 2004; Sinclair 2011). The information [data] received here is treated with utmost care to evaluate the tasks performed by the preceding phases. However, after evaluation process has been completed, then the officers in charge will embark on assigning appropriate codes to each relevant field where necessary.

 Freedom from response error –this aspect is basically to ascertain the degree of accuracy and completeness in reporting accidents, wherein RTI is observed to be incorrectly represented due to human errors which involved the physical measures of such accident scene, description of the location, including the entering process of the

information into the road accident database (O’ Day 1993; Vogel & Bester 2004). In addition, James O’ Day (1993) offered a simple explanation of what ‘Freedom from

response error’ means, by stating that “when something was measured, was it

measured correctly?” (O’ Day 1993). An illustration that elucidate the inaccuracy level of the officers [human] reporting the incident.

However, it is important to control the quality of data by manual and automatic edit checks at the Correct Entry Procedures phase (O’ Day 1993). Ultimately, analyses that identify shortcomings in the data gathered should be reported back to the data collectors, and the data entry officers so that the system will improve (O’ Day 1993). The aforementioned quality components for data processing determine how accurate any data gathered can be, if processed in conformity to the practicality of the quality components. In addition, James O’ Day (1993) suggested that the application of these quality components should be carried out manually to ascertain easy unearthing of shortcomings, and applied the corrections accordingly through automatic means if necessary.