The Nuts‐and‐Bolts of Publishing in the BJIR II: Evaluation Criteria
本文详细阐述了英国工业关系杂志（BJIR）的论文评审标准，包括论文类型、主题契合度、质量要求及对作者的实用建议，适合向该刊投稿的研究者参考。
Do the topic and method of the paper fall within the journal's purview? Is the paper likely to be relevant to the journal's readership? Is the paper of sufficient quality? Is the paper's potential contribution likely to be significant? Overall, is the paper compelling and convincing? Below, the third of these questions receives the greatest attention. However, as is evident, it is only one of five questions. Often, it is the ability to adequately address one or more of the remaining questions, not just whether the paper is of sufficient quality, that turns out to be critical for acceptance or rejection. Before proceeding, it is important to stipulate that we accept two “types” of paper. The first, which can be referred to as a “type I” paper, is directed at addressing a specific research topic or theory through the original analysis of “data” (qualitative as well as quantitative) either collected by the author or obtained through secondary sources. The second, which can be referred to as a “type II” paper, may draw on available data, but is typically directed at critically reviewing, assessing, and advancing the literature in a topic area. It may also apply or develop an alternative lens or theory for understanding and researching this area, possibly even drawing on original research or the research of others to demonstrate the value of this lens or theory. In the ideal, it will accomplish some combination of both of these purposes. The BJIR does not receive enough type II papers and, we think, this is a significant shortcoming of both the journal and of industrial relations as a general area studies. We therefore encourage authors to consider submitting them. The “risk” is higher, because these papers must make a real intellectual contribution to the journal's subject matter to stand much chance of making it through our refereeing process. But the “payoff” can also be higher, as these papers can be highly influential. Although type II papers are encouraged, they receive less attention below than do their type I counterparts. This is largely because their form and content can vary considerably, as can the criteria for assessing them. As such, it is difficult to provide guidelines for their authorship. It is also the case that by far the largest portion of papers submitted to and published by the journal fall into the type I category. Accordingly, while type II papers are referred to where relevant (see, especially, section 4.), focus is on their type I counterparts. This should not be taken as an indication that the latter are preferred over the former. The BJIR tends to be pluralistic both methodologically and ideologically. This means that authors can draw from multiple areas and perspectives in the social sciences, and there are no predefined (especially positivist) canons as to what constitutes “good” research. The journal also generally publishes on all topics relevant to industrial relations, defined broadly to include all aspects of work and employment relations. However, it privileges papers with a focus on or sensitivity to the institutions and institutional contexts of these relations, and/or their implications for matters of economy and society (i.e., the relationships between the economy and society). This can include (but is not restricted to) work on: differing forms of work and employment, work organisation and employer practices, systems of representation and rights at work, trade unionism, state policies, international organizations, and “new actors”, especially as these topics intersect with emergent social and economic issues such as work in developing countries, migrant workers, and job quality, but also as they pertain to more traditional issues such as pay, gender, conflict, and injustice. Papers should be written for a general industrial relations audience (broadly defined as per the journals’ purview). Those written only for a specific disciplinary audience are, as such, discouraged. For example, while the journal accepts papers by labour economists, such papers need to be decipherable to more than just labour economists. Thus, it is fine to submit papers that conform to the general methods employed by labour economists, but papers with pages of equations and statistical checks, as one sometimes finds in economics journals, are generally not suited to the BJIR unless most of these equations and statistical checks can be relegated to an appendix or referenced in a footnote without undermining the paper's contribution. If they can, this should in any case be done prior to submission. Papers should also have at least some empirical grounding. This applies to type II as well as type I papers, the former of which should at least draw on and reference the empirical literature or empirical developments. Purely normative or ideological analysis is generally outside of the journal's purview. So is “pure” theory (e.g., hypothetical demonstrations of union interests) unless the paper makes a significant intellectual contribution to our understanding of the subject matter in question. Philosophical or conversational essays, containing unsupported assertions and arguments, are generally rejected. Exposés, or papers that are directed primarily at portraying one or more actors negatively, are also generally rejected. This question is closely related to the former one but has more to do with the geographical location and interests of readers. In the case of the BJIR, core readers and contributors are British, and so work that addresses industrial relations in Britain is central to its mandate. However, we now also have a strong European (mostly western and northern) readership, and so are also interested in publishing papers on these countries or on the European Union in general. The same is true of the United States, Canada, and Australia. Such papers should, however, address a topic that is likely to appeal to readers outside of the country in question. The journal (and its readership) has also always tended to be more international in scope than its North American counterparts, and so we seek papers that address international issues. Broad analyses of developments in major developing countries or regions are of particular interest in this regard. So are analyses of working conditions in developing countries, broader debates over globablization, or either government or corporate labour standards and policies in the developed or developing world. Work done within virtually any country is, however, welcome, provided that it is likely to be relevant to our core readership. In the case of countries unlikely to be of particular interest to this readership, this normally means that it should have findings or implications that generalize beyond the specific country in question. Implicit to this discussion is also the question of how “big” or “small” a topic is – that is, just how important it is likely to be to our readership. If a study has clear implications for broader issues of economy and society or for comparative (institutional) analysis, it is much more likely to qualify even if the country in question is of otherwise limited relevance to our readership. If the paper is on a smaller topic, of likely interest only within that country, it is not likely to qualify. But even in “core” countries, the topic may be so micro as to not be of sufficient relevance to our readership. For example, more “micro” analyses directed at establishing the implications of specific practices (e.g. recruitment) for a managerial outcome (e.g., subsequent performance evaluations) is better suited to a management journal unless some broader lesson about labour markets or employee relations can be drawn from it. The exact structure of a paper and the way various sections within it should be developed are far from cast in stone. Papers should, however, normally have as tight a narrative as possible, so that, in the end, they convey a “story” and do not significantly depart from this story throughout. In the case of type I papers, this normally entails a short introduction, followed by a literature review and a theory component, then a discussion of research methods and data, then a presentation and reporting of empirical results, and, finally, a discussion of the results and set of conclusions. In the case of type II papers, it is rare for authors to report original research, and so there is typically no discussion of methods and data or reporting of empirical results (except when assessing other studies). The main part of the paper is the literature review and theory component, both of which typically need to be much more thorough and in-depth than for a type I paper. However, these papers may (and probably should) also contain a discussion and conclusion section, addressing the implications of the theory section. They should also include some empirical grounding, perhaps drawing on secondary data sources or referring to the empirical findings of other papers. Whether it is a type I or a type II paper, the introduction of the paper should be clear and to the point. Ideally, it should be no more than a page or two in length, serving primarily to telegraph the story to follow and convince the reader that the paper and the question(s) it addresses are of value. This section should identify the over-arching research question or objective, why it matters, and both the theory and methods by which it will be addressed in the paper. It may also outline or describe subsequent sections and preview the main arguments or findings. However, the latter should be very brief, amounting to (roughly) 50 words or less. This essentially represents the main contribution for type II papers. However, it also plays a critical role for type I papers, serving as the “set-up” for the empirical analysis to follow. It is critical that type I papers contain clear and convincing arguments, questions, expectations, or hypotheses for this analysis, and that these are well grounded in the relevant literature. There is no necessity to develop formal hypotheses, as long as, by the end of this section, the reader knows what it is the author will be exploring, why, and what the main expectations or arguments are. If formal hypotheses are developed, they need to be convincing and to follow clearly from the discussion of the literature and research question at hand. Hypotheses that seem to be ad hoc and weakly grounded can create the impression of a fishing expedition with little real purpose, significantly undermining a paper's chances. In the ideal, all hypotheses or arguments should be fully developed prior to conducting one's data analysis, so that the latter is truly a “test” of the former. In practice, however, authors often develop general expectations and questions and explore their data to determine how well these seem to be borne out, before fully developing hypotheses or arguments. They might also develop further hypotheses that derive from their results after having reported these in the paper, doing so either as part of the results section or of a subsequent section extending the analysis. Either way, when writing the paper it is essential that the authors provide some sort of convincing narrative, establishing in advance of the analysis section that the associations or the questions they will be addressing are meaningful and in effect preparing the reader for the sections to follow. Analyses that seem to be poorly grounded in the literature or that jump about from one finding and insight to another are likely to be judged of insufficient quality for the journal. Finally, this section should discuss the literature only in terms of its relevance to the research or argument at hand, and it should do so critically, establishing limitations to the literature and how the paper will advance the literature by addressing these limitations. This section should not, as a general rule, review literature or introduce material that is not specifically relevant to the research question or topic at hand. It is important, however, that it draws from, and even builds on, past work published in the journal that is relevant. A methods section is generally found only in type I papers, addressing the nature of the data and the methods employed to collect and analyse this data. It is critical that this section be written so as to convince the reader that the data collection (qualitative or quantitative) has been systematic. If the author is using a government produced secondary data set, it is important to at least briefly describe this data set and to reference more detailed sources on how it has been collected. If the author is using his/her own data set or that of another researcher or research group, it is important to discuss how the sample was selected, what the response rate was, how representative the sample would appear to be of the population in question, and possible limitations to the data. Where the research is qualitative, these criteria may be relaxed somewhat depending on the research at hand. But they cannot be ignored. Qualitative work cannot seem as if it is based on ad hoc conversations with a handful of actors or a haphazard reading of archival material. To the extent possible, authors should adhere to the formal rules for data collection one often finds in methods texts. However, the collection of data is often much messier than formal methods books would have one believe, and often researchers have to rely on secondary data sets which may not even have been collected for academic research purposes. The critical question is therefore not whether the data set conforms to formal rules found in methods texts, but rather whether it is of sound enough quality to provide reasonably convincing results. The better the data set, the more convincing the results are likely to be, but most editors and referees are likely to tolerate various limitations, especially if they do not seem to be of the author's making (e.g., if the author is using a secondary data set or there are inherent sampling difficulties) and the author is honest about them. Efforts to hide limitations or that they do not when in they only create the impression that the author is or does not them. The same is true of one's the one hand, available may be based on questions. may be more than for if the data set that others do not, as may be the case for government (e.g., the So they need not be for the paper's provided that the author their and, where possible, makes some to how they are example, through sensitivity If the data have been collected by the there is a of it is to Often, methods for that various and but which often in that the statistical more than the they are to The also often include for This can either the of they contain and the scope of the research, or it can in long and response and The question is not whether a of but rather whether they are as representative and convincing as possible the they are to and various limitations to their arguments apply to research. There are now for and these are of some if not for one's analysis, they may create a of one that may much of the of doing This is especially so where this work in which case of a may from papers submitted to the BJIR to have such to and, where they it has often not been clear how doing so has a of the methods they should be as as For research, this might include of the questions of or and possibly some of For research, it normally entails a the of the or to it and how they are the an or some the of the and possibly the This can to a of so it is important that it be as and as possible, in a Although some reporting of only a sample of the for or reference to a paper in which the is generally the case at the If there are a of and of these are only as the author can make such available on on a or in an The critical is that it should be for readers the and to the content of various without the paper. It is to editors and referees to their on a paper if the to it is not as as In some papers, especially written by economists, the methods section either or is followed by an or of the to be for the data. sections often of equations that are to most Such equations generally do not in an industrial relations journal unless essential for or the analysis to follow. If they equations that might be found (e.g., in an there is no need to include in the paper, and even if not, they may be better in an appendix or referred to in one or more The need for a results section also generally applies to type I papers In the case of this section can be because the “data” it is reporting cannot normally be into and as such and analysis. This and analysis has to be convincing to the and it has to be clear that it is based on a analysis of the research reported in the methods section. If the research or conversations with a of it is important to include from The reader has to be as to the of the “story” that the author is and, most important, the drawn from it. It is also important to how the reported results conform to or on the expectations and issues developed in the results that do not address questions or hypotheses in the review and theory section should generally be analyses typically do not have the same as their counterparts, if only because the in question are to As such, the results section should normally be It should also be to the point. It may include and but if analysis is discussion of these should be very unless it on issues or expectations developed at an in the paper or about the data results should also be as as possible in no more than a It is for authors to have more than one for in to explore how the introduction of some the for that have been But this should be done only if it on questions or hypotheses in the it only the impression of a fishing Although analyses should for more reporting than their counterparts, the with the and reporting of methods to be more may have to do with the of the or or with alternative possible for the findings. The authors should and address all of these issues to the extent possible, reporting various or sensitivity where But the of to in the of the paper should normally be limited to a pages at with the authors only reporting and results that are important for addressing these analyses that do not this end should generally be and available either on or on a or in an any in either the of the paper or in an appendix should be as clear and as For example, in the case of an least analysis, the author should normally report only the at statistical and and, can be reported in either or The former is to be preferred if the are (e.g., union and the latter if they are (e.g., or (e.g., of The author should normally from reporting exact but should rely on by or or or and or less. if the sample is (e.g. over the of these are if the sample is the latter are In the is typically while the is typically a and only relevant if it other findings in the paper. Finally, these should be based on two unless there are strong hypotheses, If a is not significant at the in the analysis or then it should as a general be as But if it does this or the question not just whether the in question is significantly with the but also what the of the especially important as sample In can be significant even the of their effect is and even The reporting of results should include not just of which are significant and what the effect are, but also of whether these results are with If they are there is need to if they are not, the author may in some as to why not and report further or analyses that follow from this It may also make to include a reporting the analysis, but only if the findings are important to the story and cannot be in a or should discuss only results that pertain to the research question or hypotheses the analysis. There is typically no need to discuss the results for unless they on the main results. if there are a of in the analysis, it may make to from the in question and below the that these have been The with the for may be available in an on or on most the results section should be clear and if not to all possible then at least to the of a reasonably to or methods that little beyond only readers and have no in a methods journal such as the If more or statistical methods are results to their more counterparts, then the latter should be and reference to the findings from the former to a or Finally, and should be unless they are essential to the “story” and cannot reasonably be through more they may seem as to for A type II paper may have a discussion section, but this section is, especially important for a type I paper. this the author should have much the story set out to The of this section, is to provide a essentially the reader as to what the of the paper have how the authors have about (e.g., the data set, if and methods what the possible limitations to the data set or results have been and why they may or may not be and what the main findings have doing the authors should one or a to discuss how the findings advance the what their implications may be for work in the area, and possibly what their broader implications may be for matters of economy and society or of This section may also include some about what the results especially if they to a broader analysis than has been in the to the paper. However, to some or draw implications that do not follow from the analysis typically in The of this section should be to convince the and referees that the paper makes a contribution to the not to make or assertions that are only likely to the author and the in some is to include a section, normally after the discussion section, and addressing possible limitations to the This is not a normally followed in the If there are limitations to the these should be and, to the extent possible, addressed on in the methods or results and then briefly in the discussion section. when these limitations are the author should not make that they are or have been taken of unless these are It is important not to various as to the of such limitations, especially where these are not is the papers have limitations. The question is what the author has done to address and how the author has been about their possible a section of at most two what the authors have II or found I and what or implications can be drawn from them. Often, however, authors the discussion and section, because the form the or much of the content the discussion section. This is a matter of depending on the author's and what to A section should not, however, the discussion section or seem as if it is an or A paper may be well written and if judged on its own just not make a sufficient enough contribution to in the This that the paper be original and advance the literature in some meaningful of the author's in writing the paper is to convince the and referees that, it this Often, however, authors can to do either because the contribution is not strong or because they have on the analysis for its own and why they are conducting it in the if the topic is of potential relevance to readers of the the and referees the question. For type I papers, it is normally important that the arguments and findings in the paper be As this is especially so if the paper draws on data or research in a country that outside of the journal's main Where this is the the authors have to convince the and referees that the paper that beyond the country in question. However, the same also applies to analyses of specific This is a particular for may report the results from research into one or a or but to address and why these results might generalize to other or This is especially if the researchers have to develop or a specific argument or set of arguments, with the that their paper is largely Such a paper may be but do little to advance our or understanding of the journal's subject There may be if a paper is addressing a question likely to be of major to the readers of the but these are In the case of type II papers, the critical question is whether there is a significant intellectual one that to or at least how of their subject In the ideal, such papers include a critical analysis of a major topic and the literature addressing and then at a provide some sort of clear alternative for further work in the area. such papers may a more economy the of and/or its and the implications of this for a major topic within the journal's purview. they may develop or apply an alternative for doing They for example, draw on some of critical or theory or on some emergent of analysis. It is important, however, that such do not just to a literature review with implications for research and they need to include more arguments, and/or about the literature. This question is largely by much of what has been But it is important, in the end, it is one that the must after reading through a paper. As such, it can make or a paper's In the must whether the question addressed by the paper is a compelling one in the that there is a sound for it in the literature more in issues of the economy and In the case of research, it is critical that the paper not seem to be the of a that has to a significant In the and any hypotheses must not seem to be little more than largely to the analysis but with no real or In the case of research, it must not seem as if the research is primarily from an author's ability to to a or and little with this question is whether the paper is This that the for the hypotheses or research question in the to the paper is well developed, so that the reader is that there is a strong for the analysis to follow. The reader must however, be that the findings from the paper are and not just an of the or the data set and analysis in the paper, and that there are not or more possible for the reported findings than have been provided by the As there are to virtually any of research. The question is just how these are. For example, in work, it often is not possible to one's arguments or the to them. But it is possible to convince the reader that there are strong for results as these or and that the or employed generally out alternative it is rare that one can that a data set is representative of the population in question or that various are of the they are to But it may be possible to that the of sampling is or to the research question at hand, and that the employed provide There is clearly a of that editors are likely to of when a paper. It is also very rare for an to that these are all to at least at the the of typically have the of a paper with often with an to and This just how difficult it can be to a paper that for at the BJIR are well of As such, the question they is not so much whether the five criteria have all been but rather whether they can be and if to all of the specific with of these criteria need not be for part of the job is to identify which have not been adequately and work with the author There are, however, a of that can, either in the to that the paper should be rejected. are addressed in the