Citing and Referencing Others’ Work

　　Notes: Citing References in Scientific

　　Research Papers

　　This is a collection of additional notes on the guide “Citing References in Scientific Research Papers “ by Timothy T. Allen, revised 2000, found at

　　http://tim.thorpeallen.net/Courses/Reference/Citations.html

　　关于“显示家庭作业”：本质上，您已经为要证明的事情选择了合适的背景……

　　a)如果是新方法，则该工作以前从未完成过，并且是新颖的。

　　b)您所依赖的理论工作已经建立(如果有的话)。 良好的引用可以节省大量的精力和精力，以证明以前已经过测试。

　　c)确保潜在的读者已经阅读或了解该领域的重要论文。 (例如，有关LASSO方法的任何论文都应引用Tibshirani最初介绍LASSO的论文)

　　关于“常识”：例如，您应该提到在哪里可以看到有关正态分布的事实，即使您认为“每个人”都知道这一点。

　　在“ Rodgers(1983)”示例中：(1983)表示这是被引用的特定手稿，而不只是某人的意见。

　　((名称，年份)格式表示此事实来自外部来源。)

　　统计和其他科学著作很少在引用中注明出版物的页码。 APA样式指南中不建议这样做。

　　在同一来源中引用两个或更多作者时：

　　对于三位或更多的作者，最好使用“ et al。”，而对于“ and others”则缩写为拉丁语。 请注意，单词“ al。”具有缩写词，但单词“ et”则没有。

　　关于寻求口头交流：这在统计中并不常见。

　　关于互联网资源：这不是可靠的引用互联网资源的来源，因为它是在2000年撰写的，当时还没有出现像Google Scholar和Wikipedia这样的当今许多高质量资源。

　　要查找旧页面，请考虑使用Archive.org/web上的Wayback Machine

　　查找页面，使其看起来尽可能接近访问日期。 您有责任在您引用的任何网页中包含访问日期。

　　On Formatting Reference Lists (This is mostly or entirely handled by

　　Bibtex now. See the chapter on Latex and bibtex.

　　Another useful resource to consider:

　　First, a look at this resource (which BibTex will do for you, mostly)

　　http://www.scientificstyleandformat.org/Tools/SSF-Citation-Quick-Guide.html

　　Example: Citations in a Service Research Paper

　　这篇论文是由统计学家合着的服务论文。 这样的服务工作对于统计学家来说是普遍且有利可图的。 粗斜体字的内容都不是原始论文的一部分，而是对如何以及为何引用某些内容的评论。 将此与方法论文的评论进行比较。

　　示例1 –服务论文：“南非青少年抑郁症的高发率：对艾滋病流行地区的年轻人进行艾滋病毒预防规划的意义” *(论文的好标题，海报的坏标题……不同的听众 ，相同的材料)

　　Introduction

　　Adolescents (ages 14-19 years) and young adulthood (20-24 years) are typically characterized by the initiation of sexual relationships, alcohol consumption and sometimes drug use. These behaviours occur to some degree almost ubiquitously among adolescents around the globe. In countries such as South Africa, the initiation of these behaviours during adolescence is complicated by the endemic presence of HIV. Adolescents and young adults (AYAs) are now considered the fastest growing population at risk for HIV transmission and 42% of all new HIV infections worldwide occur among youth aged 15-24 (UNAIDS, 2012a). In others words: UNAIDS is the group that considered young to be at risk of HIV, it’s their opinion. They were the ones that found the 42% figure. Also, the ‘year’ is 2012a because there is more than one work cited from UNAIDS in 2012 here.

　　Unlike other age groups that have seen declining transmission rates, AYAs have not benefitted to the same degree from advances in testing, treatment and prevention strategies and continue to be heavily impacted by HIV, highlighting this age group as a key demographic for targeted and informed HIV prevention programming (UNAIDS, 2012b). This sentence is long because we wanted to be clear that the whole sentence is attributed to UNAIDS. If it were two sentences, it would be best to have done the citation twice.

　　Example: Unlike other age groups… rates. (UNAIDS 2012b) Treatment and prevention… . (UNAIDS 2012b)

　　South Africa has the highest global HIV prevalence, affecting more than 6 million people

　　(UNAIDS, 2011). (Fact, attribution) The prevalence of HIV among young people (aged 15-24 years) in South Africa was estimated to be 22% (UNAIDS, 2012c) in 2010. (Fact, attribution) There is a notable gendered dimension to the HIV epidemic, with incidence rates among adolescent women in South Africa three times greater than among adolescent men (Pettifor et al., 2005). (New attribution, note the et al. This was a work from 3 or more authors) Simultaneously, while this generation of South African adolescents are the first to enter adulthood in the post-apartheid era, full of new possibilities, they are coping with the inherited post-colonial legacy of high rates of poverty, violence, gender inequality, familial fragmentation and an educational and employment system that is challenging to negotiate (Klot & Nguyen, 2011). (Long sentence. EVERYTHING IN IT is attributed) These conditions may be colluding to present significant challenges in the fight against preventing new HIV infections among young South Africans. (No citation, this is an original idea or statement, or doesn’t warrant a specific citation)

　　It is estimated that approximately three quarters of mental health comorbidities that affect adults across the life course emerge during the age period of adolescence and young adulthood (Lundet al., 2009). In the US, it is estimated that about 11% of adolescents overall may be struggling with depression at any given time, with young women at increased risk (National Institute of Mental Health, 2014)…

　　(Pattern: This is a lot of different details about the problem being presented, and each of these details can be attributed to someone else. Often this is more about giving particular statements weight to show that weren’t just made up.

　　You can also write a introduction first, and find citations to back in up after. If you can’t find a citation, maybe it isn’t true and you shouldn’t say it. This paper is in health sciences where there is a larger emphasis on citation than there is in mathematical statistics. The 70-80% of sentences being attributed is overkill, and t’s only for the introduction and part of the discussion anyways.

　　This is an example of a “Citation farm” which is put in to cover every possibility.) Citations are a lot more prevalent in the introduction because that’s where the context for your own work is.

　　*The final accepted title of this paper was "High prevalence of depression symptomology among adolescents in Soweto, South Africa associated with being female and cofactors relating to HIV transmission" by Bernard Barhafumwa, Janan Dietrich, Kalysha Closson, Hasina Samji, Angela Cescon, Busisiwe Nkala, Jack Davis, Robert S. Hogg, Angela Kaida, Glenda Gray, Cari L. Miller

　　Example: Citations in Statistical Methods

　　Paper

　　This manuscript, “On the minimum coverage probability of model averaged tail area confidence intervals”, by Paul Kabaila, is a statistical methods paper that was authored by one person. The paper is open access and can be found in the Canadian Journal in Statistics at the webpage http://onlinelibrary.wiley.com/doi/10.1002/cjs.11349/full

　　Example 2 Methodology/Theory paper: Introduction from “On the minimum coverage probability of model averaged tail area confidence intervals”, by Paul Kabaila in the Canadian Journal of Statistics

　　(2018)

　　http://onlinelibrary.wiley.com/doi/10.1002/cjs.11349/full

　　1 INTRODUCTION

　　Commonly in applied statistics, there is some uncertainty as to which explanatory variables should be included in the model. Frequentist model averaging has been proposed as a method for properly incorporating this “model uncertainty” into confidence interval construction. Such proposals have been of particular interest in the environmental and ecological statistics communities; see Fieberg & Johnson (2015) for a recent review.

　　The earliest approach to the construction of frequentist model averaged confidence intervals was to first construct a model averaged estimator of the parameter of interest as follows. This estimator is a data-based weighted average of the estimators of this parameter under the various models considered. In this approach, the model averaged confidence interval, with nominal coverage , is centered on this estimator and has width equal to the quantile of the standard normal distribution multiplied by an estimate of the standard deviation of this estimator (Buckland et al., 1997). However, Hjort & Claeskens (2003) showed that the distributional assumption on which this confidence interval is based is incorrect in large samples. This problem effectively rules out the use of such a confidence interval. Hjort & Claeskens (2003) then proposed a new frequentist model averaged confidence interval that has the desired minimum coverage probability in large samples. However, this interval is essentially the same as the standard confidence interval based on the full model (Kabaila & Leeb, 2006 and Wang & Zou, 2013).

　　An important conceptual advance was made by Fletcher & Turek (2011) and Turek & Fletcher (2012) who proposed using data-based weighted averages, across the models considered, of procedures for constructing confidence intervals. In this way the model averaged confidence interval is constructed in a single step, rather than first constructing a model averaged estimator (to be used as the center of this interval) and then seeking an appropriate formula for the width of this interval. However, some problems have been identified by Kabaila, Welsh, & Abeysekera (2016) with the method of Fletcher & Turek (2011). This leaves the model averaged tail area

　　(MATA) confidence interval of Turek & Fletcher (2012) as a promising method, particularly in the normal linear regression context as exact pivotal quantities for the parameter of interest can be specified for each model under consideration. As Turek & Fletcher (2012) noted, their method can also be applied when one has only approximately pivotal quantities, for the parameter of interest, for each model under consideration.

　　Turek & Fletcher (2012) considered a data-based weight on a model that is proportional to , , and , where AIC, , and BIC are the Akaike Information Criterion, the Akaike Information Criterion corrected for small samples and the Bayesian Information Criterion, respectively, for the model. The performance of the MATA confidence interval depends greatly on the data-based model weights on which it is based. It is helpful to applied statisticians who wish to use MATA intervals if we can narrow down the choice of data-based model weights by eliminating the worst performing model weights from further consideration.

　　A computationally convenient formula for the exact coverage probability of the MATA interval was provided by Kabaila, Welsh, & Abeysekera (2016) in the simple scenario of two nested normal linear regression models: the full model and a submodel specified by a linear constraint on the regression parameter vector. They considered a parameter of interest that is a given linear combination of the components of the regression parameter vector for the full model. Kabaila, Welsh, & Mainzer (2017) considered the same scenario in their evaluation of a MATA interval constructed using data-based weights based on Mallows’ . Of course, it is of interest to also evaluate the MATA interval in the more complicated situations that we average over more than two linear regression models (e.g., models are considered for the real life data in Section 5).

　　In the present article, the family of models that we average over is obtained as follows. For each of a given set of components of the regression parameter vector, we either set the component to zero or let it vary freely. For the MATA interval we consider quite general data-based weights on these models. These weights include, as special cases, the weights considered by Turek & Fletcher (2012) and the weights based on Mallows’ considered by Kabaila, Welsh, & Mainzer (2017). In Section 4 we show how the results of Kabaila, Welsh, & Abeysekera (2016) can be used to provide an easily computed upper bound on the minimum coverage probability of the MATA interval. This upper bound is analogous to the upper bounds of Kabaila & Leeb (2006) and Kabaila & Giri (2009) on the minimum coverage probability of the post-model-selection confidence interval in the context of the same family of models and is proved using the approach of Kabaila & Giri (2009).

　　The most important measure (in the form of a single number) of the performance of a confidence interval is its confidence coefficient, defined to be the infimum of the coverage probability of a confidence interval (e.g., Casella & Berger, 2002). If the confidence coefficient of a confidence interval is far below its nominal coverage then this confidence interval should not be used. Our new upper bound on the minimum coverage probability of the MATA interval can be used to help eliminate poorly performing model weights from further consideration.

　　In Section 2 we describe the MATA interval in detail for general data-based weights. In Section 3 we first prove the intuitively plausible result that the wider the class of models over which one averages using specified data-based model weights, the smaller is the minimum coverage probability of the MATA interval, with given nominal coverage. This result is used in Section 4 to derive an easily computed upper bound on the minimum coverage probability of the MATA interval. The application of this upper bound is illustrated using some air pollution data in Section 5. In Section 6 we use this upper bound to provide some finite sample evidence against the use of a data-based weight on the model that is proportional to Section 6 we use this upper bound to derive some large sample results which provide further evidence against the use of this data-based weight.

　　关于介绍：

　　(没有引用关于常客如何解释的。但是，有些作者在此处引用了有关模型平均或模型不确定性的引用，但这可能归因于许多不同的人。)会工作。

　　Fieberg&Johnson既是引文又是暗含的阅读推荐。

　　“基于数据的加权平均值”不一定需要引用，因为它只是数学定义。但是，稍后关于分位数的陈述不仅是事实，而且是先前书面陈述，因此值得引用Buckland等。

　　某些引用，例如Hjort&Claeskens的引用，可能先于来自该来源的信息，但需要格外小心才能确切显示来自外部来源的信息。另外，引用的是Hjort&Claeskens的作品，而不是研究人员Hjort&Claeskens自己。

　　请注意，“ Fletcher&Turek(2011)和Turek&Fletcher(2012)”是引用相同或相似信息的两个不同来源。

　　Akaike信息准则古老而无处不在(随处可见)，因此您无需引用即可摆脱困境。同样，像其他类似名字的概念(例如Poisson分布)一样，Akaike信息准则本身就是一个引文。同样，关于Mallows的Cp，在后两段中提到。

　　在以“当前文章”开头的段落中：这是“引文农场”，是许多引文的参考密集段落，同时涵盖了许多相关的思想。这是一个例子

　　如本章注释中所述，“显示您的作业”。

　　请注意，一些名称不断出现在引用中。在一篇理论论文中，尤其是在一个非常狭窄的领域中，可以多次引用同一个人的作品。没关系

　　来引用您自己在该领域的先前工作，并且在您进行长期发展的理论研究时很常见。

　　关于Casella和Berger的引用：Casella和Berger是统计方面的标准教科书。 有人对它的定义进行了引用。 即使您使用了维基百科，这也是引用维基百科的一个很好的选择。

　　另外，最起码的内容可能会被视为“常识”，但最好以防万一，特别是当存在多个定义时)

　　首先，看一下这个资源(BibTex主要为您服务)

　　http://www.scientificstyleandformat.org/Tools/SSF-Citation-Quick-Guide.html

　　Citing References in Scientific Research

　　Papers

　　Compiled by Timothy T. Allen, revised 2000. This paper greatly expands upon a handout originally prepared by an unknown author for distribution to students in introductory earth science courses at Dartmouth College. The work is presented here without copyright, although acknowledgement is (of course) appreciated. This document is also available in in Adobe

　　Acrobat Format

　　Contents

　　· Introduction

　　· When to Cite References in a Scientific Paper

　　· Details of Citing References

　　· Details of Formatting Reference Lists

　　· References Cited (in this document)

　　Introduction

　　It is important to properly and appropriately cite references in scientific research papers in order to acknowledge your sources and give credit where credit is due. Science moves forward only by building upon the work of others. There are, however, other reasons for citing references in scientific research papers. Citations to appropriate sources show that you've done your homework and are aware of the background and context into which your work fits, and they help lend validity to your arguments. Reference citations also provide avenues for interested readers to follow up on aspects of your work -- they help weave the web of science. You may wish to include citations for sources that add relevant information to your own work, or that present alternate views.

　　The reference citation style described here is a version of the "Author, Date" scientific style, adapted from Hansen (1991) and the Council of Biology Editors (1994). Harnack & Kleppinger (2000) have adapted "CBE style" to cite and document online sources.

　　When to Cite References in Scientific Research Papers

　　You should acknowledge a source any time (and every time) you use a fact or an idea that you obtained from that source. Thus, clearly, you need to cite sources for all direct quotations. But you also need to cite sources from which you paraphrase or summarize facts or ideas -- whether you've put the fact or idea into your own words or not, you got the fact or idea from somebody else and you need to give them proper acknowledgement (even if an idea might be considered "common knowledge," but you didn't know it until you found it in a particular source).

　　Sources that need to be acknowledged are not limited to books and journal articles, but include internet sites, computer software, written and e-mail correspondence, even verbal conversations with other people (in person or by telephone). All different kinds of sources must be acknowledged. Furthermore, if you use figures, illustrations, or graphical material, either directly or in modified form, that you did not yourself create or design, you need to acknowledge the sources of those figures.

　　Details of Citing References in your Text

　　When you cite a reference in your text you should use one of the following three formats:

　　(1) Mention the author by last name in the sentence and then give the year of the publication in parenthesis:

　　According to Rodgers (1983), the Appalachian mountains were formed in three events.

　　(2) Give the facts or ideas mentioned by the author and then attribute these facts or ideas by putting both his or her last name and the date in parenthesis:

　　The first of the three events occurred in the Ordovician, the second in the Devonian, and the third in the Carboniferous and Permian Periods (Rodgers, 1983).

　　(3) Quote the author exactly--be sure to put the quoted phrase between quotation marks--and then list the author's name, the date, and the page number in parenthesis:

　　"All the climaxes produced mountainous islands or highlands that shed vast amounts of debris westward to form clastic wedges or delta complexes on the continental margin." (Rodgers, 1983, p. 229).

　　You only need to include the page number in the citation if you are quoting directly, or if the source is very long and the specific fact or idea you are citing can only be found on a specific page. Direct quotations that are more than 4 lines long should be set off from the rest of your paper by use of narrower margins and single spaced lines.

　　If you have more than one source by the same author published in the same year, distinguish them both in the in-text citation and in the reference list, by appending the letters a, b, c... to the year, in the order in which the different references appear in your paper. (For example: Allen 1996a, 1996b.)

　　If the reference you are citing has two authors, use the following format:

　　Periods of glaciation have a large effect on sea level (Ingmanson and Wallace, 1985).

　　If the reference you are citing has more than two authors, use the following format:

　　Hot spots are formed by the drift of plates over mantle plumes (Vink et al., 1985).

　　If your source of information is from a personal verbal communication, you would use the following format for the first citation from that person:

　　It is possible to correct the raw dD values measured on the mass spectrometer (Mark Conrad, Lawrence-Berkeley National Lab, personal communication).

　　Later citations to the same person can be shortened, as in:

　　The reproducibility of dD determined by these methods is thought to be about +/- 2 per mil (Conrad, personal communication).

　　If your source of information is from written correspondence (a letter or e-mail), you would substitute the word "written" for the word "personal" above, and you would add the date of the letter (if dated). Personal communications are generally not included in the References Cited or Bibliography section, although unpublished papers, reports or manuscripts should be.

　　If your source of information has no individual identifiable author, use the name of the organization to which the work can be attributed in place of the author's name:The reference citation style described here is a version of the "Author, Date" scientific style, adapted from the Council of Biology Editors (1994).

　　For internet sources without any identifiable author or date, simply use the URL address as the in-text citation:

　　As New England is located at the convergence of several distinct storm tracks

　　(http://www.mountwashington.org/mtw_mtn.htm), we expect to find clear differences in isotopic composition among seasons and potentially among different rain storm events (Fig. 1).

　　Such a source would be omitted from your References Cited or Bibliography section.

　　Details of Formatting Reference Lists

　　Your list of References Cited should include all of the references you cited in your paper, and no more! It should be arranged in alphabetical order by the last name of the first author. If you have more than one entry by the same author, they should be further ordered by increasing publication date (more recent papers last). If you have multiple sources from a single author published in the same year, distinguish them both in the in-text citation and in the reference list, by appending the letters a, b, c... to the year, in the order in which the different references appear in your paper.

　　(For example: Allen 1996a, 1996b.) You should include enough information that your readers will be able to find these sources on their own. The exact format is not critical, but consistency and completeness is. Reference lists are generally reverse-indented--this just helps the reader to find references to specific authors that much faster. Follow the examples given below and you will be all set.

　　For Books

　　List all authors by last name and initials, separated by commas if there are more than two authors. Put an "and" before the last author in the list. Then put the year of publication, the title of the book (in italics if possible), the publisher, the city, and the number of pages in the book.

　　One author:

　　Gould, S. J., 1983, Hen's Teeth and Horse's Toes, W. W. Norton, New York City, 413 p.

　　Two or more authors:

　　Ingmanson, D. E. and Wallace, W. J., 1985, Oceanography: An Introduction, Wadsworth, Belmont, CA, 530 p.

　　For Articles or Chapters with separate authors from a Book or Compilation

　　List the author(s) of the article using the same format given above for books, then give the year, the title of the article or chapter (no quotes, italics or underlines), then the name(s) of the editor(s) of the book or compilation, followed by "ed." or "eds.". Then put the title of the book

　　(in italics if possible), the publisher, the city, and the page numbers where the article can be found:

　　Rodgers, J., 1983, The life history of a mountain range-- Appalachians, in Hsu, K. J., ed., Mountain Building Processes, Academic Press, Orlando, p. 229-243.

　　For an Article from a Journal or Magazine

　　List the author(s) of the article using the same format given above for books, then give the year, the title of the article or chapter (no quotes, italics or underlines), then the title of the journal or magazine (in italics if possible), the volume number of the journal (do not use the publication date), and page numbers where the article can be found:

　　One author:

　　Maddox, J., 1987, The great ozone controversy, Nature, v. 329, p. 101.

　　Two or more authors:

　　Vink, G. E., Morgan, W. J., and Vogt, P. R., 1985, The Earth's hot spots, Scientific American, v. 252, p. 50- 57.

　　For Internet sources

　　Give the author's last name and initials (if known) and the date of publication (or last modification). Next, list the full title of the work (e.g. the specific web page), and then the title of the complete work or site (if applicable) in italics (if possible). Include any version or file numbers, enclosed in parentheses. Most important, provide the full URL to the resource, including the protocol, host address, and the complete path or directories necessary to access the document. Be sure to spell this out exactly! (best to use an electronic "copy" from the

　　"location" box of your browser and "paste" into your word processor). Finally specify the date that you last accessed the site, enclosed in parentheses.

　　Focazio, M.J., Welch, A.H., Watkins, S.A., Helsel, D.R., and Horn, M.A., 1999, A retrospective analysis on the occurrence of arsenic in ground-water resources of the United States and limitations in drinking-water-supply characterizations, U.S. Geological Survey Water-Resources Investigation Report 99-4279, http://co.water.usgs.gov/trace/pubs/wrir-99-4279/ (August 1, 2000)

　　Adapt these formats as necessary for other types of sources, including unpublished reports or manuscripts -- just be sure to include sufficient information that your readers could find or obtain these sources themselves, if need be.

　　Further information can be found by consulting Hansen (1991), Council of Biology Editors

　　(1994), and Harnack & Kleppinger (2000), particularly their chapter on Using CBE Style to Cite and Document Sources.

　　References Cited (in this document)

　　Council of Biology Editors, 1994, Scientific style and format: the CBE manual for authors, editors, and publishers, 6th edition, Cambridge University Press, New York. 825 p.

　　Hansen, W. R. (editor), 1991, Suggestions to authors of the reports of the United States Geological Survey, 7th edition, U. S. Government Printing Office, Washington, D. C., 311 p.

　　Harnack, A., and Kleppinger, E., 2000, Online! A reference guide to using internet sources, Bedford/St. Martin's, http://www.bedfordstmartins.com/online/index.html (August 1, 2000).

　　Combining math and language

　　Math is shorthand

　　As abstract as they seem, all mathematics can be described in words. In fact, when we write something in mathematical symbols, we're just using a shorthand for those words. This is easy to forget because mathematical symbols are manipulated completely differently than how we manipulate words. However, this also an important fact because when we need to use both together, like in a mathematical publication, we write the non-formula material as if the formula material is in words, not symbols.

　　The simple guide for doing this is to treat every formula as if it is a clause in a (sometimes very long sentence).

　　First, to envision this try to convert the following mathematical entities (equation, symbol, formula) into words....部分省略，Tzessay提供最优质的research paper代写，我们为同学们提供24小时代写服务。