Objective: Funnel plots (plots of effect estimates against sample size) may be useful to detect bias in meta-analyses that were later contradicted by large trials. We examined whether a simple test of asymmetry of funnel plots predicts discordance of results when meta-analyses are compared to large trials, and we assessed the prevalence of bias in published meta-analyses. Design: Medline search to identify pairs consisting of a meta-analysis and a single large trial (concordance of results was assumed if effects were in the same direction and the meta-analytic estimate was within 30% of the trial); analysis of funnel plots from 37 meta-analyses identified from a hand search of four leading general medicine journals 1993-6 and 38 meta-analyses from the second 1996 issue of the Cochrane Database of Systematic Reviews . Main outcome measure: Degree of funnel plot asymmetry as measured by the intercept from regression of standard normal deviates against precision. Results: In the eight pairs of meta-analysis and large trial that were identified (five from cardiovascular medicine, one from diabetic medicine, one from geriatric medicine, one from perinatal medicine) there were four concordant and four discordant pairs. In all cases discordance was due to meta-analyses showing larger effects. Funnel plot asymmetry was present in three out of four discordant pairs but in none of concordant pairs. In 14 (38%) journal meta-analyses and 5 (13%) Cochrane reviews, funnel plot asymmetry indicated that there was bias. Conclusions: A simple analysis of funnel plots provides a useful test for the likely presence of bias in meta-analyses, but as the capacity to detect bias will be limited when meta-analyses are based on a limited number of small trials the results from such analyses should be treated with considerable caution. Key messages Systematic reviews of randomised trials are the best strategy for appraising evidence; however, the findings of some meta-analyses were later contradicted by large trials Funnel plots, plots of the trials9 effect estimates against sample size, are skewed and asymmetrical in the presence of publication bias and other biases Funnel plot asymmetry, measured by regression analysis, predicts discordance of results when meta-analyses are compared with single large trials Funnel plot asymmetry was found in 38% of meta-analyses published in leading general medicine journals and in 13% of reviews from the Cochrane Database of Systematic Reviews Critical examination of systematic reviews for publication and related biases should be considered a routine procedure

/pdf/bias-in-meta-analysis-detected-by-a-simple-graphical-test-5erh428s5c.pdf

Bias in meta-analysis detected by a simple, graphical test

物件導向軟體之架構(Object-Oriented Software Construction)探討

From the Publisher:
Written by the inventors of the technology, The Java(tm) Language Specification, Second Edition is the definitive technical reference for the Java(tm) programming language If you want to know the precise meaning of the language's constructs, this is the source for you 
The book provides complete, accurate, and detailed coverage of the syntax and semantics of the Java programming language It describes all aspects of the language, including the semantics of all types, statements, and expressions, as well as threads and binary compatibility

/pdf/the-java-language-specification-1s92cud0yi.pdf

The Java Language Specification

We have found many programming problems for which neither procedural nor object-oriented programming techniques are sufficient to clearly capture some of the important design decisions the program must implement. This forces the implementation of those design decisions to be scattered throughout the code, resulting in “tangled” code that is excessively difficult to develop and maintain. We present an analysis of why certain design decisions have been so difficult to clearly capture in actual code. We call the properties these decisions address aspects, and show that the reason they have been hard to capture is that they cross-cut the system's basic functionality. We present the basis for a new programming technique, called aspect-oriented programming, that makes it possible to clearly express programs involving such aspects, including appropriate isolation, composition and reuse of the aspect code. The discussion is rooted in systems we have built using aspect-oriented programming.

/pdf/aspect-oriented-programming-15jlquln69.pdf

Aspect-Oriented Programming

Aspect-oriented programming (AOP) is a technique for improving separation of concerns in software design and implementation. AOP works by providing explicit mechanisms for capturing the structure of crosscutting concerns. This tutorial shows how to use AOP to implement crosscutting conerns in a concise modular way. It works with AspectJ, a seamless aspect-oriented extension to the Java(tm) programming language, and with AspectC, an aspect-oriented extension to C in the style of AspectJ. It also includes a description of their underlying model, in terms of which a wide range of AOP languages can be understood.

Aspect-oriented programming

Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Publications Dept, ACM Inc., fax +1 (212) 869-0481, or permissions@acm.org.

Efficient method dispatch in PCL

Parnas' seminal work [2] on separation of concerns in design has led to diverse innovations in programming language design, to support modularity. However, there has been a growing sentiment in many quarters that there are some concerns that stubbornly resist tidy confinement, when using established modularization mechanisms in pro- gramming languages. A diverse set of new approaches have emerged in response: aspects [1], monads [5], mixin lay- ers [3], and multi-dimensional separation of concerns [4]. These approaches arose more or less independently of each other, and have (to varying degrees) developed technical maturity, real-world credibility and strong user bases. We are also now beginning to see strong scholarly comparisions of the intellectual foundations and practical utility of these different aproaches. This panel aims to support this trend. In this panel, we bring together leading experts (Profs. Batory, Kiczales, and Launchbury, and Dr. Tarr) in these different areas. Each represents a particular perspective on how to evolve and adapt the old idea of modularization to deal with new challenges such as security, fault-tolerance, distribution, and auditing. In addition, we also have two pioneering researchers (Profs. Balzer and Parnas) to pro- vide us with a historical perspective on the evolution (sic) of program modularization and evolution techniques. Position statements of some of the panelists follow, pre- sented in alphabetical order of their names:

Modularity in the new millenium: a panel summary

Advice weaving can be efficiently supported with only lightweight enhancements to existing Virtual Machines. Performing weaving at the Java bytecode (JBC) level while preserving appropriate metadata enables the VM to understand the AspectJ-specific semantics of the code and optimize it. This allows the overhead of advice weaving and performing non-local advice dispatch optimization to occur prior to runtime. It also allows the VM to perform optimizations that are unavailable to a bytecode level weaver.An experimental implementation shows that this approach can take advantage of previously known macro optimizations of expensive constructs, including cflow, as well as micro optimizations including those based on improved type analysis unavailable to JBC-based advice dispatch. A thorough benchmark evaluation confirms that the use of this architecture does not result in runtime performance overhead and benefits from the implemented optimizations.

Lightweight virtual machine support for AspectJ

The primary implementations of AspectJ to date are based on a compile- or load-time weaving process that produces Java byte code. Although this implementation strategy has been crucial to the adoption of AspectJ, it faces inherent performance constraints that stem from a mismatch between Java byte code and AspectJ semantics. We discuss these mismatches and show their performance impact on advice dispatch, and we present a machine code model that can be targeted by virtual machine JIT compilers to alleviate this inefficiency. We also present an implementation based on the Jikes RVM which targets this machine code model. Performance evaluation with a set of micro benchmarks shows that our machine code model provides improved performance over translation of advice dispatch to Java byte code.

A machine code model for efficient advice dispatch

Operating system code is complex. But, while substantial complexity is inherent to this domain, other complexity is caused by modularity problems. The implementation of certain key system concerns seems to defy traditional modular boundaries and impede evolution. From OS/360 to Windows NT, systems suffer from unintentional interactions between modules [Lehman and Belady 1985] and require developers to be intimately familiar with implicit patterns of interaction between subsystems [Vogels 19991]. 
The thesis of this work is that aspect-oriented programming (AOP) can be used to improve evolution in operating system code by improving modularity. Specifically, AOP can alleviate modularity problems associated with concerns that are inherently crosscutting—no single modular decomposition can localize both the crosscutting concern and the concerns it crosscuts. Better modularization of crosscutting concerns requires that their implementation be localized, their interaction with the parts of the system they crosscut be explicit, and their internal structure be clear. By accomplishing these three things, AOP can provide better structural support for evolution. 
The first part of the thesis provides a case study comparing modularity involving three crosscutting concerns in the original versus aspect-oriented implementation within the FreeBSD operating system [Lehey 1999]. This comparison highlights specific improvements in modularity of both the crosscutting concerns and the concerns that are crosscut, or interacting concerns, in the AOP implementation. 
The second part of the thesis surveys evolutionary changes the three crosscutting concerns underwent between releases 2.2 (1997), 3.3 (1999) and 4.4 (2001) of FreeBSD, and identifies some of the specific impediments the original implementation poses with respect to evolution. A case study of the impact of these same evolutionary changes on the aspect-oriented implementation highlights improvements in locality of change afforded by the AOP implementation, and the ways in which improvements in modularity persist across the versions. 
The final part of the thesis presents inferences and generalizations based on the results of the case studies. We infer from the case studies that AOP can be used to improve evolvability of the three crosscutting concerns and their interacting concerns in three versions of FreeBSD, without harming non-interacting concerns. We then generalize these modularity benefits to more concerns, more systems, and more versions, and infer support for our main claim—that aspect-oriented programming can be used to improve evolvability of operating system code by providing better modularity of crosscutting concerns and their interacting concerns, without harming non-interacting concerns.

/pdf/improving-evolvability-of-operating-systems-with-aspectc-4l8r8rhvp9.pdf

Gregor Kiczales

Papers

Efficient method dispatch in PCL

Modularity in the new millenium: a panel summary

Lightweight virtual machine support for AspectJ

A machine code model for efficient advice dispatch

Improving evolvability of operating systems with aspectc