Institutional repository statistics: reliable, consistent approaches for Canada

Introduction: “Usage metrics are an effective way for libraries to demonstrate the value of their institutional repositories, however, existing tools are not always reliable and can either undercount or overcount file downloads. As well, although statistics can sometimes be accessed through the various repository interfaces, without an agreed standard it is impossible to reliably assess and compare usage data across different IRs in any meaningful way.”[1] The Task Group for Standards for IR Usage Data has undertaken an information-gathering exercise to better understand both the existing practices of Canadian repositories, as well as the emerging tools and processes available for repositories to track and monitor usage more effectively. This exercise directly links to the broader goals of the Open Repositories Working Group, which are to “strengthen and add value to the network of Canadian open access repositories by collaborating more closely and adopting a broader range of services.”[2] Our recommended course of action is for all Canadian IRs to collectively adopt OpenAIREStatistics. This path aligns with the following recommendations which our group also advances: Recommendations: We suggest the following Mandatory (M) and Optional (O) recommendations: R1(M):All Canadian IRs should adopt the COUNTER Code of Practice. R2(M): All Canadian IRs should select a service that allows for interoperability with other web services via a fully open, or accessible, permissions-based API. R3(M): All Canadian IRs should usea statistics service that practices transparent communication and maintains a governance strategy. In addition, we strongly urge for the future that Canadian IRs consider the following advice. R4(O): Make further investments into understanding and utilizing the common log format (CLF). R5(O): Conduct research into the privacy implications of collecting use statistics via third party services with commercial interests and consider available alternatives. R6(O): Practice a healthy skepticism towards tools and solutions that promise “increased” usage statistics, and instead advocate for responsible collection assessment based on multiple aspects of use.

Institutional repository statistics: reliable, consistent approaches for Canada

Introduction: “Usage metrics are an effective way for libraries to demonstrate the value of their institutional repositories, however, existing tools are not always reliable and can either undercount or overcount file downloads. As well, although statistics can sometimes be accessed through the various repository interfaces, without an agreed standard it is impossible to reliably assess and compare usage data across different IRs in any meaningful way.”[1] The Task Group for Standards for IR Usage Data has undertaken an information-gathering exercise to better understand both the existing practices of Canadian repositories, as well as the emerging tools and processes available for repositories to track and monitor usage more effectively. This exercise directly links to the broader goals of the Open Repositories Working Group, which are to “strengthen and add value to the network of Canadian open access repositories by collaborating more closely and adopting a broader range of services.”[2] Our recommended course of action is for all Canadian IRs to collectively adopt OpenAIREStatistics. This path aligns with the following recommendations which our group also advances: Recommendations: We suggest the following Mandatory (M) and Optional (O) recommendations: R1(M):All Canadian IRs should adopt the COUNTER Code of Practice. R2(M): All Canadian IRs should select a service that allows for interoperability with other web services via a fully open, or accessible, permissions-based API. R3(M): All Canadian IRs should usea statistics service that practices transparent communication and maintains a governance strategy. In addition, we strongly urge for the future that Canadian IRs consider the following advice. R4(O): Make further investments into understanding and utilizing the common log format (CLF). R5(O): Conduct research into the privacy implications of collecting use statistics via third party services with commercial interests and consider available alternatives. R6(O): Practice a healthy skepticism towards tools and solutions that promise “increased” usage statistics, and instead advocate for responsible collection assessment based on multiple aspects of use.

Efficient Software Quality Assurance Approaches Oriented to UML Models in Real Life

UML is accepted as the standard notation for object-oriented (OO) development. UML models have a widespread use in today’s software practices. Any initiative to improve software quality assurance (SQA) should assume that a high percentage of deliverables to be controlled are currently based on the use of UML notation. However, real life projects are strongly influenced by the need of reaching tangible productivity and efficiency goals. SQA techniques should be customized to meet the balance between quality and budget supported by risk analysis. In this chapter, different strategies and techniques devised to follow the above philosophy of efficiency are presented, especially centred in automatic testing generation from specifications. Our proposal is mainly based on a recommended course of action as well as on integrated tool support for Eclipse environments.

Recommendations Implicit in Policy Defaults

Should people be considered organ donors after their death unless they request not to be, or should they not be considered donors unless they request to be? Because people tend to stay with the default in a variety of domains, policymakers' choice of default has large and often important effects. In the United States, where the organ-donation policy default is “not a donor,” about 5,000 people die every year because there are too few donors. Four experiments examined two domains—being an organ donor and saving for retirement—where default effects occur and have important implications. The results indicate that default effects occur in part because policymakers' attitudes can be revealed through their choice of default, and people perceive the default as indicating the recommended course of action. Policymakers need to be aware of the implicit messages conveyed by their choice of default.