Answer:
With the computerization of the state bureaucracy’s allowing for efficient data storage and
easy access to information, the interest in developing statistics based on administrative
records has reached new heights. Already, registry data is being used extensively for both
research purposes (Bakker, 2012), and in the production of official statistics. For example,
according to a newly completed governmental investigation, approximately 95% of the
Swedish official statistics originate from administrative processes (SOU 2012:83). In this
context, developing methods to understand, evaluate and improve the quality of this type of
statistics is essential.
Unfortunately, research investigating the characteristics and quality of registry data has not
been at par with the rapid increase in the amount of accessible data. What is probably the first
comprehensive theoretical book dedicated to registry-based statistics was published as late as
2007 (Wallgren & Wallgren, 2007). A few years later, when Statistica Neerlandica presented
their special edition on statistics based on administrative data, they stated in their introductory
note that “Theory of registry based data is scarcely out of the egg” (Bakker and Daas, 2012).
Furthermore, surprisingly little attention has been given to issues related to the quality of the
raw data, and the potential impact that this has on the final statistics. Instead, most of the
research so far has focused on other aspects, such as methods to integrate data from different
records (Zhang, 2012), the validity of administrative variables (Bakker, 2012) or differences
in estimates from surveys and administrative records (Groen, 2012). Exploring these topics
helps, but is not sufficient to get a good understanding of the quality issues related to the raw
data.
A significant part of the work in the field has been carried out by individual public authorities.
A few examples in Sweden include an evaluation of the coding system used by the
Educational Registry (SCB, 2006), evaluation of the coding of professions for the Swedish
occupational register (SCB, 2007) and evaluation of the quality in the Swedish mortality
registry (The National Board of Health and Welfare, 2010). Nevertheless, such studies are
often written in other languages than English, and their results are advertised mainly within
the institution that conducted them. This makes their findings inaccessible to a broader public,
and hinders faster developments in this field.
The Swedish National Council for Crime Prevention (Brå) has commenced a series of quality
studies to gain insights into issues associated to the raw data on which the official criminal
statistics are based. In this paper, the results from the first two studies are presented. By
analyzing the impact of the coding errors on the official criminal statistics, we reveal the
importance of investigating and quantifying the extent of such errors when working with
registry data. We conclude by formulating a few recommendations that are likely to be
applicable in any institution producing or using statistics based on registry data.
police protection from thief please make me brainslit please
Copyright © 2024 EHUB.TIPS team's - All rights reserved.
Answers & Comments
Verified answer
Answer:
With the computerization of the state bureaucracy’s allowing for efficient data storage and
easy access to information, the interest in developing statistics based on administrative
records has reached new heights. Already, registry data is being used extensively for both
research purposes (Bakker, 2012), and in the production of official statistics. For example,
according to a newly completed governmental investigation, approximately 95% of the
Swedish official statistics originate from administrative processes (SOU 2012:83). In this
context, developing methods to understand, evaluate and improve the quality of this type of
statistics is essential.
Unfortunately, research investigating the characteristics and quality of registry data has not
been at par with the rapid increase in the amount of accessible data. What is probably the first
comprehensive theoretical book dedicated to registry-based statistics was published as late as
2007 (Wallgren & Wallgren, 2007). A few years later, when Statistica Neerlandica presented
their special edition on statistics based on administrative data, they stated in their introductory
note that “Theory of registry based data is scarcely out of the egg” (Bakker and Daas, 2012).
Furthermore, surprisingly little attention has been given to issues related to the quality of the
raw data, and the potential impact that this has on the final statistics. Instead, most of the
research so far has focused on other aspects, such as methods to integrate data from different
records (Zhang, 2012), the validity of administrative variables (Bakker, 2012) or differences
in estimates from surveys and administrative records (Groen, 2012). Exploring these topics
helps, but is not sufficient to get a good understanding of the quality issues related to the raw
data.
A significant part of the work in the field has been carried out by individual public authorities.
A few examples in Sweden include an evaluation of the coding system used by the
Educational Registry (SCB, 2006), evaluation of the coding of professions for the Swedish
occupational register (SCB, 2007) and evaluation of the quality in the Swedish mortality
registry (The National Board of Health and Welfare, 2010). Nevertheless, such studies are
often written in other languages than English, and their results are advertised mainly within
the institution that conducted them. This makes their findings inaccessible to a broader public,
and hinders faster developments in this field.
The Swedish National Council for Crime Prevention (Brå) has commenced a series of quality
studies to gain insights into issues associated to the raw data on which the official criminal
statistics are based. In this paper, the results from the first two studies are presented. By
analyzing the impact of the coding errors on the official criminal statistics, we reveal the
importance of investigating and quantifying the extent of such errors when working with
registry data. We conclude by formulating a few recommendations that are likely to be
applicable in any institution producing or using statistics based on registry data.
Answer:
police protection from thief please make me brainslit please