Collaborative Rearch in Sociology:Trends
and Contributing Factors
Laura Hunter &Erin Leahey
Published online:11May 2008#Springer Science +Business Media,LLC 2008
Abstract To what extent do sociologists collaborate?Has this changed over time?What factors contribute to rearch collaboration among sociologists?To answer the questions,we examine trends in collaboration over a 70year period and empirically test a variety of explanations for the increa in collaboration that we find.We analyze data collected from a stratified random sample of articles in two leading sociology journals between 1935and 2005(n =1274).Most of our analys are descriptive and display trends over time.However,we pool the data across all years and estimate logistic regression models to asss the relative contribution of various factors.We find that the importance of geographical location has been waning since the 1950s,although the growth in cross-place collaborations stagnated between 1980and 2005.We find that quantitative rearch is more likely to be collaborative,as are projects requiring data collection,though this may change becau the collaboration rate among condary data urs is increasing at a faster rate.We find no significant gen
der differences in rates of collaboration,although male sole-authorship remains the most common form of publication.We also find the institutional prestige of coauthors is typically higher than that of sole-authors.Our results elucidate the extent of collaboration in sociology and reveal how veral factors have contributed to this major shift in work organization.
男孩子乳名
Keywords Rearch .Collaboration .Sociology .Coauthorship
Anecdotal and empirical evidence suggest that within scientific fields,rearch collaboration is on the ri,and that collaboration can be beneficial for individual scholars (Presr 1980)and for scientific progress more generally (Hara et al.2003).While investigations of collaboration in other fields have been conducted (Davis and Wilson 2001;Endersby 1996;Fisher et al.1998),less is known about the field Am Soc (2008)39:290–306
DOI 10.1007/s12108-008-9042-1
L.Hunter (*):E.Leahey
Department of Sociology,University of Arizona,Social Sciences Building,Room 400,
1145E.South Campus Drive,Tucson,AZ 85721,USA上诉状范文
e-mail:lahunter@u.arizona.edu
sociology.In fact,in their1999American Sociologist article,Babchuk et al.(1999:6) stated that“inquiries to date have not systematically explored,either for purpos of description or explanation,the extent to which cooperation in collaborative rearch and publication is becoming more common ”In this paper,we take Babchuk,Keith,and Peters’work as a starting point.Like them,we examine trends in collaboration over time using empirical data from sociology journals.Then,we move the rearch forward by not only theorizing but also empirically asssing various explanations for the increa in collaboration,such as changes in the demographic composition of the field and increas in communication-enhancing technology.
There are important reasons to study scientific collaboration.Collaboration, which is typically operationalized through coauthor relationships,is now normative and ubiquitous in most scientific ,over95%of articles in major periodicals in physics,biochemistry,biology,and chemistry are coauthored (Babchuk et al.1999)).Moreover,this normative style of rearch constitutes a major shift in work organization(Wenger1998).This new mode of work organization is related to other important social trends,including:ties between science,policy,and economic activity(Hellstrom2003);increas in work-bad inequality and the importance of professional netwo
有神rks(Kanter1994);and the stimulation of innovation(McKelvey et al.2003).In this paper we asss the extent of this shift in work organization for the field of sociology and we examine how various factors—technology,rearch methods,data accessibility,authors’gender, and institutional prestige—contribute to the shift.
物理学习方法
What Promotes Collaboration?
Technology and Place
Most social scientists agree that even with globalization and rapid advancements in information and communication technology,physical place still matters(Oln and Oln2000).In recent years,social scientists have been exploring the impact that shared space(neighborhoods,workplaces,schools)has had on individual behavior, including scientific collaboration.Scholars have found that spatial proximity encourages collaboration becau it encourages informal communication(Bozeman and Corley2004;Katz and Martin1997:5).Collaboration decreas exponentially with distance,especially for kinds of rearch that are bound to a laboratory(Landry and Amara1998;Walsh and Bayma1996).
But the importance of physical place may be waning relative to virtual space. Over a century ago,Karl Marx predicted that time would annihilate space in the modern era;that distances would no l
onger matter given the technological advances in communication and transportation(Griswold and Wright2004).To what extent is this true for scientific collaboration?Several scholars have theorized about the role that information technology may have on collaboration,but few have examined it empirically.Anecdotal evidence regarding the importance of information technology abounds:track-change capabilities in Microsoft Word that facilitate co-writing and editing of papers;access to datats and rearch publications via the Internet,and
大背头发型ppt插入备注licens to common computer programs,which ea the division of labor associated with data analysis.One mathematician cited in Walsh and Bayma(1996:670)relayed his experience collaborating with a distant scholar:“We communicate via email.It’s a critical component of my work.I can’t remember what it was I u it to keep in touch Ask a question about the literature.Ask questions of collaborators.”Information technology,especially computer-mediated communica-tion,allows physically distant scholars to find each other easily,remain in touch inexpensively(even as they travel),and work interdependently,efficiently,and perhaps with greater trust given that email exchanges are automatically logged.As Beaver(2001)suggests,the Internet may encourage collaboration across rearch institutes,states,regions,and even countries.We hypothesize that information technology has provided scholars with the opportunity to expand their academic networks into distant geographic locales.
Method
Previous rearchers have claimed that methodological approach—quantitative or qualitative—has affected rates of collaboration(Babchuk et al.1999;Endersby 1996;Moody2004;Presr1980).Why may this be so?If,as the scholars have suggested,collaboration increas efficiency through the division of labor,then methodological approaches that can be divided easily should lend themlves best to collaboration.For example,the various tasks involved in ethnographic fieldwork may be difficult coordinate if more than one rearcher is involved.On the other hand,the tasks involved in quantitative rearch,such as coding,programming, cleaning,and analyzing,may be more easily divisible among veral scholars. Additionally,quantitative methods have become increasingly complex and may require more extensive knowledge and equipment;thus the experti of veral scholars may be needed(Endersby1996;Hudson1996).Previous findings indicate that coauthorship is not evenly distributed across sociological work and is more likely in specialties that admit to an easier division of labor,such as the Sociology of Health and Medicine(Moody2004).Additionally,lower collaboration trends in history(Endersby1996)and anthropology(Babchuk et al.1999)lend support to our working hypothesis:that papers employing quantitative methods,which are more amenable to the division of labor,are more likely to be collaborative.
Data Accessibility
Whereas veral rearchers have addresd the role of methodology in collaboration patterns,few scholars have addresd the role of data accessibility—that is,whether the data come from extant,condary sources,or whether the data were collected by the rearchers themlves.Moody(2004)suggests but does not empirically asss that the ri of large-scale data collection in the social sciences lends itlf well to a team-production model that collaboration implies.Fisher et al.(1998)also contend that both large-scale primary data collection efforts and the availability of large condary datats from archival ,ICPSR)increa collaboration in political science.To date,no empirical data have been available to test the conjectures.In this paper we are able to determine if a relationship between data
accessibility and collaboration exists and if the relationship has changed over time, possibly coinciding with technological developments,such as electronic data archives.We anticipate that papers using condary,rather than primary,data sources are more likely to be collaborative.
Gender
Becau women are less integrated into professional networks than men(Kyvik and Teigen1996),the
y may also be less involved in collaborative relationships(Cole 1981;Kyvik and Teigen1996;McDowell and Smith1992).Among economists, McDowell and Smith(1992)found that there is a significant propensity for rearchers to lect coauthors of the same x.Therefore,the disproportionately large number of men in certain academic fields—even sociology,especially among tenured faculty(Roos and Jones1993)—rais the possibility that women may have fewer opportunities to collaborate,especially if they work in small departments (McDowell and Smith1992).Men are more likely than women to rve as additional first authors),supporting the argument that women may not have as many opportunities to collaborate(Clemens et al.(1995).
三国演义书籍
However,other empirical evidence suggests that women are more likely to collaborate,possibly arising out of x differences in desire to collaborate.Fisher et al.(1998)found that coauthorship is the most prevalent form of publication for women,whereas sole authorship is the most common type for men.Bypassing concerns that proportions will em larger for the minority group(in this ca, women)in a minority–majority group comparison(Rytina and Morgan1982), Moody(2004)works with odds derived from multivariate analys,finding that women have a50%greater odds of collaborating compared to men.The discrepancies in desired and actual collaboration between the xes may aris
e out of differences in professional confidence.Reskin(1978)suggests that women may be less confident in their professional careers,making them more reliant on support from their work environment and colleagues.
Given the conflicting evidence,we refrain from stating directional hypothes about the relationship between gender and collaboration.We explore whether gender informs the likelihood of collaborating,and how gendered collaboration patterns have changed over time.And becau previous rearch suggests that institutional prestige may be more determinative of a first-authored publication in a top journal for women than for men(Clemens et al.1995),we also asss whether gender interacts with another factor of interest to us:institutional prestige.
Institutional Prestige
We expect authors’institutional prestige to be positively related to collaboration patterns,though the casual direction remains unclear.It may be the ca that rearchers houd in prestigious institutions receive more invitations to collaborate and accept them.Although institutional prestige is more than simply an amalgamation of members’productivity(Keith and Babchuk1994),institutional prestige continues to rve as an indicator of quality(Clemens et al.1995),such that members of prestigious departments may have greater success when attempting to风雨同舟是什么意思
publish in prominent journals(Keith et al.2002).For this reason,and perhaps becau they have special competencies that are in demand(Melin2000),members of prestigious departments may be solicited as prospective collaborators more frequently than their peers in less prestigious institutions.Converly,it may be the ca that greater levels of collaboration enhance institutional prestige through the career mobility process.If collaboration enhances the quantity(De Solla Price and Beaver1966)and quality(Stephan and Levin1987;Presr1980)of work as well as the likelihood of acceptance(Lawani1986),then rearchers who collaborate may disproportionately be offered positions at top institutions.
Data and Methods
To examine and explain trends in collaboration over time,we collected historical data using archival material.We focud on the discipline of sociology and collected data from scholarly journals.We limited our analysis to two historically and currently top sociology journals that have been at the core of the discipline since their founding in the late19th and early20th centuries.The are general sociology journals that circulate among sociologists more widely than specialized journals.Becau the are the top journals in the discipline,they have consistently published innovative,pioneering work. For the reasons,their opportunity to influence others’work and establish trends is probably e
nhanced(Turner and Turner1990).Although not reprentative of all sociological rearch,articles in leading journals tell us something about“disciplinary standards and ideals”(Platt1996:126),and have been ud widely in previous rearch on the discipline of sociology(Clemens et al.1995;Leahey2005).1 We collected data from a20%stratified random sample of articles published between1935and2005in the American Journal of Sociology and American Sociological Review,thus supplementing the sample ud by Leahey(2005).We stratified by journal and by issue in order to obtain a sufficient number of articles from each year.We cho not to stratify by year itlf,as that could have resulted in a ries of articles from the same thematic issue to the exclusion of other issues published in the same year.Stratifying by issue guarantees that oddities associated with particular issues will be fully reprented.After eliminating erroneous inclusions,1,274articles remained2,424of which were coauthored.
Dependent Variable
We u coauthorship to measure rearch collaboration.We focus on coauthorship as a measure of collaboration to ensure that our results will be compatible with previous rearch(Bozeman and Corley2004;De Solla Price and Beaver1966;Endersby 1996;Martin-Sempere et al.2002;Melin2000;Merton1973).In addition,output
1We thank an anonymous reviewer for pointing out that becau there may be a relationship between collaboration and the likelihood of publishing in a prestigious journal,rates of collaboration documented here may not adequately reprent all sociological work.
2From the sampling frame we excluded rearch and teaching notes,book reviews,comments and replies, bibliographies,obituaries,and all articles in supplemental issues.
