Facets of Specialization and Its Relation to Career Success: An Analysis of U.S. Sociology,1980 to 2015

Abstract

We investigate how sociology students garner recognition from niche field audiences through specialization. Our dataset comprises over 80,000 sociology-related dissertations completed at U.S. universities, as well as data on graduates’ pursuant publications. We analyze different facets of how students specialize—topic choice, focus, novelty, and consistency. To measure specialization types within a consistent methodological frame, we utilize structural topic modeling. These measures capture specialization strategies used at an early career stage. We connect them to a crucial long-term outcome in academia: becoming an advisor. Event-history models reveal that specific topic choices and novel combinations exhibit a positive influence, whereas focused theses make no substantial difference. In particular, theses related to the cultural turn, methods, or race are tied to academic careers that lead to mentorship. Thematic consistency of students’ publication track also has a strong positive effect on the chances of becoming an advisor. Yet, there are diminishing returns to consistency for highly productive scholars, adding important nuance to the well-known imperative of publish or perish in academic careers.

Keywords

field change sociology of science scientific careers structural topic modeling

Sociology operates on the verge of its own undoing, seemingly without respite. Throughout the twentieth century, many scholars sounded the alarm around the fragile state of the discipline: sociology is disintegrating (Znaniecki 1945:514), amid a clash of theories (Giddens 1976:715), compartmentalized into “congeries of outsiders to each other” (Collins 1986:1340–41), with subspecialties drifting apart from each other (Hand and Judkins 1999:18). As early as 1920, we find calls to map the field (Committee on Standardization of Research 1920, cited in Rhoades 1981:16). It seems the discipline can only be pursued, if at all, against its own tendency for self-subversion (Holmwood 2010:650).

Yet, after a century of sociology’s alleged disintegration, students continue to pursue the discipline. In fact, the number of awarded PhDs in sociology has increased steadily since the 1990s, and in recent years, it is close to the prior peak of the late 1970s (see Figure 1), which is known as a period of critical overcrowding (Collins 1986:1338; Crane and Small 1992:199). The field’s demographics changed too. Since the late-1980s, women have been awarded the majority of sociology doctorates, and the share of ethnic minority graduates climbed from approximately 10 percent in the 1970s to around one third in 2016. In contrast, the number of faculty positions has slightly declined since 2000 (ASA 2021a). These antagonistic trends might increase the pressure to publish early in one’s career, which, in turn, raises the bar that young scholars must clear to secure a permanent position in sociology (Warren 2019).

Figure 1.

Demographic Developments in U.S. Sociology

Specialization has long been recognized (Collins 1986; Weeber 2006), maligned (Coser 1990; Gitlin 1990), and studied (Leahey, Keith, and Crockett 2010; Leahey and Moody 2014) as a publication strategy of extant sociology faculty, but little attention is given to how graduate students adopt this strategy. This is surprising, given that young scholars are charged with reproducing and transforming the field (Bourdieu 1988:94; Kuhn 1962:151). Our objective here is therefore to answer, how do sociology students specialize and to what professional returns? To observe students’ specialization, we utilize an understudied genre of writing: the dissertation, which, we argue, enacts a student’s appeal for recognition in one of the field’s research areas.

The academic specialization process entails adopting a particular research focus, and sustaining this focus consistently over time to become a recognized expert. This focus can vary in degree and targeted content. It also occurs within a larger context of available themes in the field, with respect to which it can be described as novel or traditional specialization. We synthesize these facets of specialization into three antinomies: tradition versus innovation (Foster, Rzhetsky, and Evans 2015; Kuhn 1959; Leahey and Moody 2014), focus versus breadth (Leahey 2007; Leahey et al. 2010; Teodoridis, Bikard, and Vakili 2019), and consistency versus change (Koppman and Leahey 2019; Trapido 2015). With regard to student dissertations, the first two facets concern students’ topical choices: Does their thesis deal with a narrow set of topics? Is it within an identifiable research tradition, or does it span across them? The last facet concerns the extended publication track of students after graduation: Do they continue to publish in the same thematic area as their dissertation or do they change their focus? In addition to these three facets, we also consider specialization in the sense of students’ choice of specific topics, which we call targeted specialization.

To study research specialization, we focus on the language used in student theses and ensuing published works. We fit a structural topic model (STM) (Roberts, Stewart, and Airoldi 2016; Roberts et al. 2014) to the abstracts of student dissertations. We recover topics, or “word systems” (Astley 1985:499), reflective of the research areas directly expressed in young scholars’ writings, and then use these topics to develop measures for each facet of specialization. By linking this information to students’ ensuing careers, we are able to ascertain the relation that different forms of specialization have on an important long-term career outcome in academia: time to becoming a primary advisor. Such attainment is a conservative proxy for success—only around 3 percent of PhDs in our broad sample become advisors of sociology-related theses—but it reflects if and when a student occupies an influential position in field reproduction. Furthermore, by focusing on cohorts of graduate students, our approach allows us to illuminate specialization strategies used at the start of a career, whereas prior studies concentrating on salaries (Leahey 2007) or citations of extant faculty (Sinatra et al. 2016) may suffer from survival bias.

We connect facets of specialization to the time it takes until first mentorship via event-history models. We find students in 1980 to 2015 became advisors of sociology-related PhDs via distinct specialization strategies. Students were most likely to achieve influential field positions when they targeted certain topics reflective of the cultural turn or statistical models rather than adopting a thematic focus in general. They were also likely to secure such positions when they adopted novel topic-combinations rather than traditional forms. Finally, students were more likely to become advisors when they established a consistent research identity early in their careers. Such consistency need not be large, however. Only a small number of publications consistent with the dissertation provides considerably better chances to achieve a position as an advisor. This finding adds important nuance to the well-known imperative of publish or perish.

Specialization as Career Strategy in Sociology

During the plenary sessions of the 1988 annual meeting of the American Sociological Association, Gitlin (1990:216) argued that sociology had lost public value because students with ever more specialized interests invested in obscure scholastic writing rather than engage in debates of public import. Coser (1990:208–9) added that in this kind of highly specialized writing, the “methodological tail wags the substantial dog,” and concepts lack critical bite. Moore’s (1990) rejoinder in this debate argued, on the contrary, that African American, Latinx, and women sociologists, in particular, play a significant, albeit specialized, role in advancing the discipline’s critical tradition, often in their teaching and in organizing for regional professional associations.

The 1988 ASA polemic occurred during a period in which young sociologists adopted high degrees of specialization to avoid downward mobility into contingent academic work (Weeber 2006:59). More formally, in the 1980s and 1990s, young sociologists pursued specialization as an effective adaptation to an overcrowded academic market, in which they could only compete for the attention of their peers within narrow disciplinary niches (Collins 1986).

The narrative of fragmentation alluded to in the introductory paragraph, and echoed by Gitlin and Coser, is grounded in the true state of affairs of a growing field with many loosely connected themes and an impoverished corps—a discipline that is increasingly subject to the pressures of specialization and market-oriented higher education (Holmwood 2010). What this narrative obscures is that fragmentation is an emergent property of the way novice scholars vie for a place in the field; namely, through specialization. It is no surprise then, that previous efforts to investigate sociology—as a field (Leahey and Moody 2014) and career (Koppman and Leahey 2019; Leahey 2007; Leahey et al. 2010)—focus on the issue of specialization.

Our own contribution to a better understanding of specialization as a career strategy in sociology is threefold: we investigate an understudied site for specialization—the dissertation; a critical career outcome—becoming an advisor; and a suite of specialization strategies—topic choice, focus, novelty, and consistency.

The Dissertation Genre

The dissertation is one of the most decisive documents in a scholar’s life.¹ It can be thought of as a claim to a career and to a research program worth pursuing (Bourdieu 1988:94). A thesis can also be thought of as a rite of passage (Collins 2002:43), whereby a young scholar shows deference to a field’s canon (Paré, Starke-Meyerring, and McAlpine 2011:220). Of course, young scholars have also been described as central agents of field transformation (Kuhn 1962:151), a role that may also be expressed in the dissertation genre.

Dissertations constitute a site for the specialization strategies of young sociologists; this is the field constituency variously charged with reproducing or transforming the field. Also, this constituency’s efforts to capture the attention of a niche audience fuel the disciplinary lament of fragmentation. In addition, unlike any other genre of academic writing, the dissertation does not suffer from publication bias, a pressing issue in the sociology of science (Fortunato et al. 2018).

Becoming an Advisor

Becoming an advisor encodes a crucial aspect of the success of a specialization strategy. It is through relations of mentorship that scholars recruit allies into intellectual movements and legitimate novel academic identities (Frickel and Gross 2005). Which graduates do or do not become advisors, and the time it takes to become one, reflects which forms of specialization are in most immediate demand. Scholars who achieve this stage become gatekeepers of the specialized disciplinary niche. Also, intergenerational networks of mentor-mentee relationships can reinforce the status of different methodologies in sociology subfields (Bandelj 2019). These are also central to the transmission of highly codified research practices (Leahey 2006).

Despite the critical importance of advisorship as a career achievement, particularly in relation to specialization strategies, this is not an outcome that receives much attention in the literature. Many scholars emphasize rewards granted to esteemed faculty as important points in academic careers—for instance, prestigious prizes (Zuckermann 1977), higher salaries (Leahey 2007), or high-impact papers (Sinatra et al. 2016). Just a handful of studies focus on the acquisition of long-term positions, like (tenured) professorships (Balsmeier and Pellens 2014; Leahey et al. 2010; Lutter and Schröder 2016). This work finds that besides the quantity and quality of the basic unit of scholarly achievement (i.e., publications), other non-meritocratic factors, especially gender (van den Besselaar and Sandström 2016) and affiliations at elite institutes (Burris 2004), influence the chances of becoming faculty. Instead of using appointments to professorships or getting tenure as a proxy for academic success, we focus on all graduates of sociology-related theses and identify which go on to become advisors of sociology-related theses themselves and the time to such a position. We argue this is a more conservative and substantive approximation of achieving an influential and successful scientific career.

There is a considerable gap between the site in which we will study specialization—the dissertation—and the career stage in which we will measure the success of this strategy—becoming an advisor. To connect the two, we conceptualize and measure different facets of specialization, some of them narrowly connected to the document of the dissertation; others extending outward from the dissertation to a scholar’s entire documented research output.

Facets of Specialization

Students write about specific methodologies, theories, and subject matters that appeal to distinct audiences in the field. Through the selection of topics, students can display varying degrees of thematic latitude (focus versus breadth); furthermore, they can consolidate prior links and traditions, or combine topics that have rarely been combined (tradition versus novelty). In their future work, students can persist in the same vein of their dissertation or venture out to other themes (consistency versus change). These different facets of specialization have rarely been studied together, or with an approach that allows the specialized subject matter to emerge directly from students’ writing.

Specializing on Target Topics

A dissertation draws from multiple topics; that is, multiple methods, theories, and subject matters expressed in careful word choices that elicit recognition from peers. This community of peers may be organized in specialized sections of a professional organization (Leahey 2007; Leahey and Moody 2014); or it may be more loosely structured as a collective that shares a set of methods and ontologies (Knorr Cetina 2009) and “similar views on contested scientific topics” (Teplitskiy et al. 2018:1828). In either case, the topic structure expressed in dissertation writing displays students’ aptitude to elicit their community’s signature thought style (Antons, Joshi, and Salge 2019:3044–5; Fleck 1979). This crucial act of position-taking through topic choice will have differential professional returns, as students will vary in their success dealing with a topic, and the communities students appeal to through their topic choices may have more or less to offer them in terms of career prospects.

The effect of targeted specialization in specific topics on career outcomes is not well understood (Gieryn 1978). Among the few studies that exist, Hoppe and colleagues (2019) find that in medicine, selecting certain topics contributes to lower success rates of African American scientists at the NIH. Munoz-Najar Galvez, Heiberger, and McFarland (2020) find that in education, targeting topics associated with the interpretive approach improves PhD students’ chances of an academic career. And Moody (2004) finds that for sociology, specialty areas shape collaboration networks, and quantitative work is more likely to be coauthored. Based on the extant literature, we formulate a baseline hypothesis that specializing more than others in specific topics will likely shape career prospects:

Hypothesis 0: Targeting certain topics will affect PhD students’ chances of becoming an advisor.

Focus versus Breadth

In contrast to the effect of targeted specialization, much effort has been put into disentangling whether specialists or generalists have more successful careers. This distinction captures the trade-off between spending limited resources to gain an in-depth understanding of a specific domain versus gathering more superficial knowledge across several domains. Previous research describes affordances and penalties for both strategies. For instance, specialized scholars who master the literature of a subfield can leverage this mastery to write more papers (Leahey 2007) or produce breakthrough inventions (Conti, Gambardella, and Mariani 2014). In the sociology of concepts and categories, Hannan and colleagues (2019:4) argue that, in a variety of contexts, “a clear position in one category tends to increase an offer’s appeal”; they call the obverse of this affordance the “category-spanning discount.”

In contrast, Teodoridis and colleagues (2019) find that affordances for both specialist and generalist strategies apply in creative work, depending on the pace of change in a given field, with generalists being the more desirable collaborators in slower fields. A related stream of work argues that creative workers should seek boundary-spanning combinations that achieve greater recognition (Burt 2004; Fleming 2001; Fleming, Mingo, and Chen 2007; Schilling and Green 2011). Acknowledging the competing evidence, as well as the lack of pertinent case studies dealing with the dissertation genre or academic field novices, we hypothesize two competing claims:²

Hypothesis 1a: A topically focused thesis leads to higher chances of becoming an advisor.

Hypothesis 1b: A topically broad thesis leads to higher chances of becoming an advisor.

Novelty versus Tradition

Several studies find that novel combinations of prior knowledge increase the recognition of academic papers. This line of work describes unusual combinations of chemical substances (Foster et al. 2015), bibliographic references (Uzzi et al. 2013), disciplines (Leahey, Beckman, and Stanko 2017), subfields (Leahey and Moody 2014), concepts (Hofstra et al. 2020), and topics (Antons et al. 2019) as garnering outsized recognition. However, there is also evidence that papers benefit from closeness to established knowledge. Foster and colleagues (2015) make clear that the “bread and butter” business of academia (i.e., job talks and grants) rewards the production of traditional work, and Shi, Leskovec, and McFarland (2010) show that work squarely situated within and consistent with a research community has the likeliest chance of recognition. Fleming (2001), Uzzi and colleagues (2013), Trapido (2015), and Leahey and colleagues (2017) all point out that novel recombination of ideas and categories represents a high-risk high-reward endeavor that often fails, but sometimes yields great success.

It is unclear how novelty reflects on young scholars and their prospects to become advisors. Younger scholars are more likely to pursue novel research (Bourdieu 1988; Kuhn 1962), but they often face precarious job prospects. Thus, we posit competing claims between innovative and traditional combinations (Kuhn 1959):

Hypothesis 2a: Novel topic combinations in a thesis increase the odds of becoming an advisor.

Hypothesis 2b: Traditional topic combinations in a thesis increase the odds of becoming an advisor.

Consistency versus Change

Our hypotheses thus far connect crucial features of specialization in students’ dissertation writing to a long-term outcome of their academic careers: becoming an advisor. We reviewed a broad literature showing the relevance of specialization to career outcomes. But the link between an initial state of specialization and a long-term outcome relies on the assumption that entry conditions into an academic field can credibly affect how a student’s career unfolds. This link has been theorized in classical works in the sociology of science, where scholars’ initial standing in a field is said to express and reproduce stable aspects of their institutional and social status (Bourdieu 1988). Of course, this assumption needs to be tested empirically. To do that, we introduce another facet of specialization, which we refer to as consistency versus change. This facet describes the link between the topic structure of a student’s dissertation and topic choices in ensuing work, leading up to the career stage when the person may be considered to become an advisor.

Patterns of consistency (versus change) affect the unfolding of professional careers. For a new scholar, persisting in the same research area is a safe bet because changing specializations requires starting over, and chances of failure are high (Gieryn 1978:105–7). Yet, there are times when a significant pivot might pay off. For instance, movie actors increase their chances of being hired by playing similar roles early in their careers, and then by playing against type later on (Zuckerman et al. 2003). By contrast, Indian bureaucrats are rewarded for being a generalist early and adopting a more specialized and consistent identity later (Ferguson and Hasan 2013).

Closer to our case, Koppman and Leahey (2019) find that in sociology, long-standing commitments to a relevant identity, such as interdisciplinarity, increase a scholar’s likelihood of using unconventional methods. The authors also find that affiliations to high-ranking universities add to the effect of a consistent identity; a result that is consistent with cumulative advantage (Correll et al. 2017). Similarly, for academic engineers, Trapido (2015) shows that audiences are more likely to deem authors’ novel contributions valuable when they are consistent with previously developed identities as innovators. Synthesizing previous findings on the effects of consistency leads to a third pair of competing hypotheses:

Hypothesis 3a: A consistent research identity (persisting with topics expressed in one’s thesis) improves the odds of becoming an advisor.

Hypothesis 3b: A changing research identity (departing from topics expressed in one’s thesis) improves the odds of becoming an advisor.

Analytic Strategy

To study how student theses reflect strategies of specialization, we draw on approximately all doctoral theses from 1980 to 2015 that conduct research related to sociology. We apply structural topic models (STM) (Roberts et al. 2016; Roberts et al. 2014) to those texts to identify the themes used in students’ writings. In so doing, we adopt a methodological frame that uses the actual language in students’ theses. The identified themes cut across fixed schemes of prelabeled categories often used in prior work. We argue that students’ specialization strategies are reflected in how they use these topics, and we develop metrics indicative of these strategies’ usage. Finally, we use event-history models to test hypotheses as to which specialization strategies enable graduates to quickly reach sociology advising positions.³

Sample

Our analyses draw on a comprehensive database of doctoral theses written in the United States provided by ProQuest. We selected all dissertations in which the stem “sociol” was present in the title, abstract, keywords, or department (N = 82,363).⁴ In doing so, we cast a wide net for theses related to sociological “thought styles” (Fleck 1979), representing the disciplinary structure as fragmented and influenced by neighboring fields. This relatively broad definition allows us to trace graduates from multiple disciplinary backgrounds who also contribute to the sociological field (Lamont and Molnár 2002). Knowing the department of graduation also allows us to distinguish between students at sociology’s “institutional heartland,” that is, students graduating from sociology departments, and authors whom we label as sociology-related. The data comprise each dissertation’s abstract, title, name of the graduate student, advisor’s name, institutional affiliation, and National Research Council (NRC) discipline. In total, the records contain theses from 12 disciplines,⁵ cover 275 different U.S. institutions, and stretch from 1980 to 2015.

Identifying Thought Styles via Structural Topic Models

To discover thematic structures in dissertations, we use topic modeling (Blei 2012; Evans and Aceves 2016). Scholars have used topic models successfully to describe, for instance, the methodological divide of sociology (Schwemmer and Wieczorek 2020), themes of management research (Antons et al. 2019), math education (Inglis and Foster 2018), computer science (Anderson, McFarland, and Jurafsky 2012), and education research (Munoz-Najar Galvez et al. 2020). A conventional topic model takes as input a collection of texts formatted as a document-term matrix—a table in which every document is represented as a vector of word counts over the vocabulary of the document collection. The model learns the mixture of topics that make up each document and the words that belong to each topic via successive updates to two sets of prior multinomial distributions, (1) one of documents over topics and (2) one of topics over words.

We fit a structural topic model (STM) to the abstracts of sociology dissertations. STMs were developed by Roberts and colleagues (2014) to enable the use of document covariates (e.g., publication year) to improve the consistent estimation of topics. One of the ways an STM can include document covariates is in the estimation of a document-specific distribution over topics based on a document’s covariates. In other words, in an STM, topic proportions can be modeled as a function of covariates.⁶ Our model uses a spline on year of publication to allow the average topic proportions to vary over time. We will describe the resulting trends in the first section of our results.

Like other unsupervised learning procedures (Molina and Garip 2019), however, researchers have to decide the number of topics (k) the model should identify. Note that this is problematic because the relevant topics are not known a priori. We selected a solution with 60 topics after using a careful three-step approach to select k (see Part A of the online supplement).

Two output matrices are essential for us. The first contains a list of topics, each consisting of words ordered by their probability of being associated with the respective topic. Table 1 presents these topics using FREX values (Roberts et al. 2016), where a word is weighted by the harmonic mean of its rank in terms of frequency (FR) and exclusivity (EX) for a topic. For instance, among the most descriptive words for topic40 (T40) are “qualitative,” “theme_emerge,” “ground_theory,” “phenomenological,” and “live_experience.” It seems intuitive to assume theses with high loads of T40 are engaged in qualitative methods. Thus, labeling topics resembles qualitative coding approaches, because topics allow multiple interpretations depending on the perspective we use to look at them (Inglis and Foster 2018).

Table 1.

Topics in Sociological Dissertations, 1980 to 2015

ID	Label	FREX terms	Prevalence
1	French/Can Soc	french, france, canada, canadian, ghana, haitian, aboriginal, quebec	.006
2	Identity	identity, negotiate, discursive, construction, understanding, discourse, ethnography, everyday_life	.038
3	Religion	church, congregation, evangelical, denomination, pastor, worship, mormon, catholic	.010
4	Race: Latinx	mexican, mexican_american, los, angeles, puerto_rican, latina, chicano, puerto_rico	.008
5	Self-perception	perception, satisfaction, self_esteem, self_concept, self_perception, expectation, self, congruence	.019
6	Violence	violence, abuse, domestic_violence, victimization, rape, sexual_abuse, sexual_harassment, batter_woman	.011
7	School	teacher, public_school, school_district, teacher_student, charter_school, school_choice, teacher_perception, school_reform	.019
8	Child Welfare	foster_care, child_maltreatment, adoptive, adoptee, foster_child, adoptive_parent, permanency, child_welfare_system	.011
9	Medical	physician, hospitalization, mental_illness, medical, hmo, chronic_pain, medication, asthma	.012
10	Rural/Environ	forest, rural_community, climate_change, community, conservation, local_community, environmental_justice, earthquake	.016
11	Method: Experiments	experimental_group, treatment_group, pretest_posttest, posttest, program, intervention, program_evaluation, effectiveness	.018
12	Econ Soc	expenditure, income_inequality, household, saving, income, tax, estimate, cost	.025
13	College Students	freshman, grade_point_average, international_student, college_students, service_learn, campus, semester, gpa	.014
14	Method: Survey/Scales	score, subscale, locus_control, item, scale, anova, significant_difference_find, chi_square	.033
15	Social Movement	social_movement, protest, electoral, election, mobilization, activism, democracy, political	.025
16	Internet	internet, online, digital, gamble, technology, information, addict, gaming, facebook	.008
17	Labor	agricultural, farm, industrial, crop, irrigation, farmer, factory, capitalist	.018
18	Gender: Sexuality	gay, lesbian, homosexuality, gay_man, lgbt, transgender, sexual_identity, homophobia	.011
19	Career	career, mentor, career_development, career_choice, success, recruitment, professionalism, mentor_relationship	.012
20	Participation	participation, activity, recreation, volunteer, leisure, civic_engagement, camp, disability	.011
21	Sociolinguistics	linguistic, speaker, dialect, sociolinguistics, vowel, grammar, phonological, code_switch	.015
22	Social Services	service, board, agency, social_service, service_delivery, personnel, referral, fund	.016
23	Status/Groups	group, attribution, identification, group_member, cognitive, trait, preference, judgment	.021
24	Illness	nurse, grief, nurse_home, hospice, bereavement, nurse_care, grieve, end_life	.006
25	Theory	philosophical, moral, habermas, rationality, theorist, weber, sociology_knowledge, epistemological	.034
26	Science	science, graph, algorithm, error, cluster, dna, parameter, simulation	.017
27	Organizations	collaboration, organizational_structure, stakeholder, participatory, nonprofit, organization, partnership, advocacy	.018
28	Marriage	couple, marital, intimacy, husband_wife, marital_quality, marriage, marital_adjustment, relationship_satisfaction	.012
29	Gender: Feminist	woman, infertility, feminist, gender_equality, abortion, patriarchy, woman_color, gender_relation	.011
30	Ethnic Studies	muslim, taiwanese, hindu, indigenous_people, arab, post_soviet, indigenous, ethnic_conflict	.013
31	Sport/Consumption	sport, tourism, athlete, athletic, brand, restaurant, football, student_athlete	.008
32	Work	work, employee, job_satisfaction, organizational_commitment, transformational_leadership, organizational_climate, emotional_intelligence, organizational_performance	.015
33	Law	law, legal, juror, attorney, legislation, lawyer, litigation, supreme_court	.016
34	Criminology	prison, inmate, juvenile, incarceration, offense, recidivism, prisoner, correctional	.013
35	Fatherhood/ Divorce	father, stepfamily, single_mother, father_involvement, fatherhood, coparenting, single_parent_family, paternal_involvement, family_structure	.012
36	Method: Survey/Opinion	attitude, respondent, attitude_toward, orientation, survey, belief, acceptance, willingness	.025
37	Arts	music, theatre, cinema, jazz, film, hip_hop, artist, art	.013
38	Demography	fertility, earnings, labor_force_participation, contraceptive, employment, labor_market, birth_interval, job_search	.016
39	Social Work	social_work, therapist, trainee, social_work_education, msw, school_social_work, field_instructor, work_alliance	.009
40	Method: Qualitative	qualitative, theme_emerge, ground_theory, phenomenological, live_experience, interview, semi_structure_interview	.037
41	Social Capital	network, social_capital, social_network, trust, exchange, cooperative, interorganizational, network_structure	.017
42	Higher Education	faculty_member, ged, college_choice, graduate_school, professor, postsecondary_education, postsecondary, aspiration	.012
43	Poverty	homeless, homelessness, house, poverty, welfare_reform, homeownership, dog, relocation	.011
44	Globalization	global, south_africa, globalization, ngo, brazil, human_right, world_bank, cross_national	.022
45	Conflict	friendship, boy, conflict, behavior, aggression, mediation, aggressive, interaction	.014
46	Family Disorder	sibling, family_function, alcoholic, adhd, sibling_relationship, family_environment, autism, eat_disorder	.012
47	Noise	fiction, victorian, poem, jewish_identity, literary, shakespeare, american_jewish, writer	–
48	Adolescence	adolescence, among_adolescent, teen, youth, suicidal_ideation, protective_factor, parental_monitor, african_american_adolescent	.017
49	Parenting	parent_involvement, parenting_style, parent_perception, head_start, school_readiness, parent, parent_child, parent_education_program	.012
50	Method: Modeling	model, contextual, theoretical_model, causal, explanation, hypothesis, indicator, individual_level	.043
51	Urban	neighborhood, police_officer, suburb, community_police, city, gentrification, gang, redevelopment	.012
52	Aging	elderly, retirement, old_adult, grandparent, grandchild, caregiver_burden, grandmother, old_people	.012
53	Immigration	immigrant, migrant, refugee, vietnamese, chinese_american, filipino, navajo, asian_indian	.017
54	Motherhood	maternal, toddler, maternal_depression, mother_infant, emotion_regulation, attachment, adult_attachment, mother_child_interaction	.011
55	Stress	cope, stressor, cope_strategy, stress, cope_behavior, hardiness, psychological_distress, veteran	.013
56	Mass Media	message, television, news, broadcast, news_medium, mass_medium, media, journalist	.011
57	Knowledge	knowledge, phase, plan, learn, strategy, approach, development, expert	.029
58	Race: African American	black, white, racial, racism, black_white, racial_identity, multiracial, race_relation	.015
59	Health	hiv, hiv_aid, obesity, health_behavior, nutrition, immune_deficiency, prenatal_care, bmi	.013
60	Historical Soc	change, society, shift, today, period, transformation, traditional, modern	.028

Note: We interviewed experts in various sociological research fields and asked them to derive labels for each topic based on the most probable and distinguishing terms of the STM. The “Prevalence” column represents a topic’s popularity, that is, the sum of its loads across all theses and years (in percent of all loads). Part A of the online supplement provides further information on why T47 is considered noise (see Figure A3 in the online supplement). We also provide an extended topic-term matrix including prob-scores and additional terms in the repository (https://github.com/RapHei/Facets-of-specialization).

The second important output for our undertaking is the document-topic matrix, often denoted θ. This matrix contains each document summarized as a vector of the proportions of its words associated with each topic (Roberts et al. 2014:1069). These topic proportions represent the strength of association between each topic and a given dissertation. We use the θ matrix to construct metrics for students’ topic choices, focus, novelty, and consistency.

Following Antons and colleagues (2019:3044–5), we argue that classification of scholarly work resulting from a topic model resembles Fleck’s (1979) “thought styles” because it provides “a snapshot of the word choices representing the thought collective’s encoding of its signature thought style.” Topics then “bear a more or less distinctive coloring conforming to a given thought style” (Fleck 1979:109), and we treat a dissertation described in terms of topics as a proxy for students’ position in the sociological discourse—sociology’s “word system” (Astley 1985:499)—at a crucial early career stage. These topics will later become a means to measure facets of specialization (targeted specialization, focus, novelty, consistency) that we relate to students’ career outcomes.

Dependent Variable

Our dependent variable is the time to becoming a primary advisor of a sociology-related PhD student. We view “sociology-related” theses as those that are defended in a sociology department or that contain the stem “sociol,” as described in the Sample subsection.⁷ The broad sample of dissertations related to sociology represents the permeability of disciplinary boundaries (Lamont and Molnár 2002). Our dependent variable, which follows the same inclusion criteria as our sample, reflects a student’s success at becoming a gatekeeper for future students aspiring to enter the sociological field.⁸ We interpret the time it took a student to first become an advisor as a conservative approximation of reaching an influential position in the sociological field.

To construct this variable, we matched all unique PhD student names to the names of primary advisors in our sample. This matching procedure captures students’ transition from student to advisor within the U.S. graduate education system. Note that former students in our sample who move on to advise students on dissertations with no reference to sociology would not appear as advisors in our data.

Independent Variables

In this section, we describe the way we measure each facet of specialization: targeted specialization, degree of focus, degree of novelty, and consistency with ensuing published work. All our measures are built on top of the document-topic matrix θ, which represents documents as mixtures of topics.

A topic of targeted specialization is distinctive of a student’s dissertation relative to the extent to which peers use the same topic. We want to capture that distinction among peers, because most topics (e.g., theory or methods) are used to a similar extent in many dissertations—that is, most topics in a dissertation cannot be discerned as targets of specialization. Therefore, we develop a metric that ascertains the extent to which students target a topic above and beyond most of their peers. Formally, the metric for targeted specialization measures if the topic load of a topic i in a document d (θ_di,t) is larger than the 75th to 90th percentiles (in steps of 5) of that topic in the previous year across all theses $(θ_{d i, t} > θ_{i, t - 1}^{p e r c e n t i l e s}) .$ This suite of threshold-based metrics captures targeted specialization with different levels of strictness. In the Results section, we depict only coefficients that are statistically significant across all thresholds, which provides the most conservative estimates. Results for other thresholds are available in Part B of the online supplement.

The degree of focus in a thesis (Hypothesis 1) is measured by calculating the Herfindahl index for the discrete probability distribution over topics θ_d that summarizes each document d as a mixture of topics. Here we are following Ferguson and Hasan (2013), who used the Herfindahl index to measure the focus of prior experience in specific types of work and the effect on Indian administrative officers’ careers. The Herfindahl index is calculated by summing the squared topic proportions over all topics k in a dissertation, that is, $\sum_{i}^{k} θ_{i}^{2}$ . This metric ranges from 0 to 1; larger values indicate a dissertation is more focused in a narrower set of topics.

Our measure of novelty (Hypothesis 2) captures the extent to which a dissertation introduces rare or unique combinations of topics with respect to prior dissertations. For that purpose, we adapt the relational measurement of “novel integration” used by Leahey and Moody (2014:236). We begin by converting the document-topic matrix, θ, into an occurrence matrix, where each dissertation receives a value of 1 on the two topics with highest proportions, and a value of 0 elsewhere. This matrix records the combination of the two most descriptive topics in each dissertation. Then, we subset this matrix to consider only the dissertations published three years prior to the document for which we want to measure novelty. Call this dissertation A, and imagine it combines the topics higher education and violence. We compute a ratio r of the number of times we observe the combination of higher education and violence to the expected occurrence of this combination. The novelty of dissertation A is calculated as 1 – r, with higher values indicating greater novelty of its distinctive combination. We provide empirical examples in Part C of the online supplement.

Students’ thematic consistency captures their continued focus on the topics first discussed in their dissertations. We gathered each student’s publications as part of a larger project in which we matched articles published by students in the ProQuest sample with papers recorded in the Web of Science database.⁹ In reconstructing students’ publication records, we only consider articles in journals classified under the social sciences and humanities. In total, we match 140,746 articles to 18,548 students over a period of 35 years.

In a second step, we use our structural topic model to represent the abstracts of the journal publications as mixtures of the 60 topics that best describe U.S. dissertations related to sociology. We compare students’ dissertation to their ensuing publications using cosine similarity as a measure of thematic distance—a conventional procedure in natural language processing (Manning, Raghavan, and Schütze 2008). These comparisons allow us to keep a running average of how consistent students remain with respect to the topics of their dissertation over time. In so doing, consistency provides a longitudinal measure to answer Hypothesis 3 on how coherent students’ publication records appear in light of their dissertation, or, put differently, how consistent a research identity is when measured from the start of an academic career.

To close this section, we discuss the control variables of our models. We use three controls for academic standing: productivity, institutional prestige, and insider status. Productivity is the number of papers a student published, which we introduce to our models in log scale. For institutional prestige, we use an indicator variable that takes a value of 1 when the student’s institution appears in the top 10 of the U.S. News and World Report rankings for graduate departments. This indicator is discipline and year specific when available. However, we had no recourse to comparable rankings before 1992, so we use that year as reference for all previous years.¹⁰ We also control for students’ status as field insiders and outsiders. We define insiders as young scholars graduating from sociology departments; conversely, field outsiders are PhD students writing a thesis related to sociology but receiving their doctoral degree from other departments.

We also control for students’ gender and race, which are both covariates known to influence scientific careers (Cech et al. 2011; Huang et al. 2020; van den Besselaar and Sandström 2016). Gender has also been shown to influence the choice of research topics (Key and Sumner 2019), in general, and methods, in particular (Grant, Ward, and Rong 1987). Gender and race are derived from students’ names by matching first names to Social Security Administration data to predict gender, and matching surnames to U.S. Census data to predict race (Hofstra et al. 2017; Hofstra and de Schipper 2018).¹¹ All our variables are described in Table 2.

Table 2.

Overview of Variables

Variable	Description	Operationalization	Mean	SD	Range
Advisor	Becoming an advisor	Dichotomous. PhD students who mentor advisees at sociology departments or advise theses that contain “sociol” (further details in notes 4 and 7).	.031	.174	[0,1]
Elite	Prestige of alma mater	Dichotomous. University of PhD students within top 10 of US News ranking in their department’s discipline.	.113	.316	[0,1]
Female	PhD student’s gender	Dichotomous. Estimated by using first name and Social Security Administration data (further details in note 11).	.545	.498	[0,1]
White	PhD student’s race	Dichotomous. Estimated by using full name and U.S. Census data (further details in note 11).	.708	.455	[0,1]
Insider	PhD at sociology or other department	Dichotomous. “Insider” (PhD granted by a sociology department) or “outsider” (PhD granted by other department).	.505	.500	[0,1]
Pub_Cum	Cumulated number of PhD student’s publications	Numeric. Number of publications of a PhD student as given in Web of Science (further details in note 9).	1.542	6.830	[0,359]
Targeted Specialization	Specialization in each topic relative to cohort	Dichotomous. 1 if the proportion of a topic in a thesis is higher than 75 to 90 percent of other students graduating in the previous year (H0). Details in Part B of the online supplement.	.101	.301	[0,1]
Focus	Focus versus breadth in topic proportions	Numeric. Degree of focus/breadth in a thesis calculated using the Herfindahl index (H1). Details in Part C of the online supplement.	.153	.077	[0,.951]
Novelty	Traditional versus novel topical combinations	Numeric. Degree a thesis integrates unusual topic combinations (H2). Details in Part C of the online supplement.	–.126	.949	[–8.27, 1]
Consistency	Consistent versus changing research identity	Numeric. Degree for how far or close students’ publications are in reference to one’s thesis (H3). Details in Part C of the online supplement.	.089	.213	[0,.99]

Event-History Analysis

Our data provide information not only on who becomes an advisor of sociology PhDs, but how quickly they reach that position. On average, it takes 10 years between PhD and first serving as an advisor (see Part D of the online supplement). We take advantage of the time dependencies in our data by drawing on event-history analysis (EHA). This modeling strategy gives us two estimates of interest: the likelihood of becoming an advisor, and time to first advisorship.

EHA assumes a collection of units moves among a finite number of states that may occur at any point in time and are influenced by some observable factors (Coleman 1981:6). Eligible data should therefore include events, time elapsed until events happen, and time-constant or time-varying covariates. In our case, this setup models the time from PhD graduation to first becoming a PhD advisor, and how this process is influenced by facets of specialization reflected in one’s thesis (topic choice, focus, novelty) and publication record (consistency), controlling for social attributes and performance (gender, race, prestige of alma mater, publications).¹²

The language of “hazard rate” of an event points to origins in health studies (Amorim and Cai 2015), but EHA has many applications in social sciences and offers a variety of models (Blossfeld, Rohwer, and Schneider 2019; Box-Steffensmeier and Jones 2004; Singer and Willett 2003). We chose a parametric model with Weibull distribution to measure the time until first advisorship (see model choices in Part D of the online supplement).

Results

The Relation between Topic Choice and Becoming an Advisor in Sociology

In this and the following subsection, we test our hypotheses regarding facets of specialization and their effects on the time it takes students to become advisors in sociology. We begin by discussing targeted specialization, then we turn our attention to focus, novelty, and consistency. To test our hypotheses, we present an event-history model that measures the time between thesis and first advisorship (see Table 3).¹⁶ In our sample, 2,345 PhD students advised at least one student with a thesis related to sociology or at a sociology department. Thus, only around 3 percent of all PhDs in our sample end up as a primary advisor for doctoral students later in their career. This underlines that our outcome represents a conservative proxy of academic success, that is, advising assumes having a faculty position as well as responsibility over graduate training.

Table 3.

Event-History Models for Time to First Advisorship in Sociology

	(1)	(2)	(3)	(4)	(5)
Elite	2.060^*** (.110)	1.887^*** (.107)	1.791^*** (.105)	1.782^*** (.104)	1.720^*** (.102)
Female	.882^** (.043)	.883^* (.045)	.890^* (.049)	.888^* (.049)	.906 (.050)
White	1.434^*** (.081)	1.418^*** (.083)	1.431^*** (.087)	1.433^*** (.087)	1.454^*** (.088)
Insider	1.665^*** (.084)	1.612^*** (.085)	1.434^*** (.082)	1.427^*** (.082)	1.427^*** (.082)
Pub_Cum	1.550^*** (.029)	1.275^*** (.030)	1.237^*** (.031)	1.061 (.036)	.664^*** (.053)
Focus		.979(.025)	.952(.032)	.972(.032)	.974(.036)
Novelty		1.094^*** (.027)	1.099^*** (.028)	1.096^*** (.028)	1.089^** (.034)
Consistency		1.388^*** (.025)	1.362^*** (.026)	2.112^*** (.141)	1.752^*** (.165)
T2 Identity			1.584^*** (.122)	1.574^*** (.121)	1.569^*** (.120)
T14 Method: Survey/Scales			.666^*** (.072)	.662^*** (.072)	.670^*** (.073)
T50 Method: Modeling			1.352^*** (.104)	1.342^*** (.103)	1.293^*** (.100)
T58 Race: African American			1.356^*** (.102)	1.359^*** (.103)	1.387^*** (.105)
T59 Health			.730^*** (.071)	.721^*** (.070)	.716^*** (.070)
Consistency_Sqr				.865^*** (.018)	.868^*** (.032)
Consistency × Pub_Cum					1.762^*** (.158)
Consistency_Sqr × Pub_Cum					.909^*** (.026)
Focus × Pub_Cum					.996(.023)
Novelty × Pub_Cum					1.008(.024)
N	82,363	79,137	79,137	79,137	79,137
AIC	27,960	25,417	25,018	24,960	24,801

Note: All models use weights as given by official SED numbers (see Part D of the online supplement), with exponentiated coefficients (hazard ratios) and standard errors in parentheses. Models presented here use only the strictest threshold for targeted specialization (i.e., students who focus more on a topic than 90 percent of their peers). To facilitate readability, we depict the effect of specific topics only if it is significant (p < .001) across all thresholds (see Part B of the online supplement for further details on targeted specialization). Note that the reduced number of observations in all models except (1) is due to the construction of the novelty metric (the window comprises three years). Robustness checks for this and other choices are reported in Part F of the online supplement.

p < .05; ^**p < .01; ^***p < .001 (two-tailed).

We can confirm our baseline hypothesis (Hypothesis 0) and affirm that the topic structure of a thesis—the particular combination of subjects, methods, and theories—affects career chances. Figure 4 depicts five targeted specializations that are strongly associated with becoming an advisor. Each of these topics is highly significant (p < .001) across all thresholds of the targeted specialization variable.¹⁷

Figure 4.

Hazard Ratios for Becoming an Advisor in Sociology

Methods matter

Specializing in a specific topic reveals mixed effects. Three topics affect a student’s chances of becoming an advisor positively, and two decrease the likelihood of advisorship. T14 survey scales had the strongest negative effect. T14 was also among the most declining topics, which we linked to severe contemporary challenges with developing surveys (Meyer et al. 2015). However, the negative relationship to advisorship may also stem from the wide range of applications of this technical skill. Future research should explore the extent to which specialization in this topic, and others with similar technical content, may provide differential job opportunities in academia and industry.

In contrast with surveys, T50 modeling has a strong positive effect. This topic represents general statistical approaches applied in dissertations and is semantically close to a diverse set of subject matters (see Figure 2). The effects of T14 survey scales and T50 modeling should be read together as suggesting a possible shift in dominant quantitative research practices and may even represent a turn toward “data science.”

Cultural turn, identity, and race

T2 identity is among the most popular and trending topics for sociology PhD students. This topic reflects the cultural turn, one of the major currents in social sciences in the past decades (Alexander 1988; Jacobs and Spillman 2005). Our results reveal that the probability of becoming an advisor rises considerably when a PhD student writes about identity, culture, or discourse. The effect is stronger the more a student focuses on T2.

T58 race: African American also significantly increased a student’s chance of becoming an advisor. The effect of T58 also grows with greater focus on it. Note that the specific topic about race our model captures tends to be combined with T2 identity (Figure 2). However, the social construction of race as a sociological object of study in U.S. scholarship predates the cultural turn (cf. Du Bois [1897] 2020). Sociological debates about race in the United States span far more varied intellectual histories than those captured in the cultural turn. These histories are to this day in a state of flux, revision, and recovery (Magubane 2016; Morris 2015).

Neighboring fields

Finally, we find that T59 health is negatively associated with the likelihood of becoming an advisor of sociology-related dissertations. Across all thresholds of the targeted specialization covariate, focusing on T59, which has as distinctive terms “HIV,” “obesity,” “health_behavior,” and “immune_deficiency,” reduces chances to advise PhD students by almost 30 percent. Despite its growing popularity in sociology-related dissertations (see Figure 3), T59 is not at the core of the sociological field (see Figure 2). At the same time, social theories seldom inform public health studies (Pound and Campbell 2015). T59 may represent a changing academic environment with more fluid research demands, to which sociology—with its permeable boundaries—is particularly sensitive. Yet, for the observed time period, students specializing in T59 health have lower chances of becoming advisors of sociology-related PhDs.¹⁸

To assess the magnitude of the reported effects, Figure 5 compares targeted specialization to the influence of race and gender. Race and gender are two of the most formative dimensions in U.S. society, stratifying social hierarchy in general (McCall 2002), and academe in particular (van den Besselaar and Sandström 2016). Our models confirm that female, minority PhD graduates have a significantly lower likelihood of becoming an advisor, net of an individual’s merit in terms of publications (see Table 3). Thus, racial, and to a lesser degree, gender inequalities leave their footprint in sociology departments, despite the field’s growing diversity (see Figure 1). Comparing these to T50 modeling, Figure 5 shows T50’s effect has a similar magnitude.¹⁹ Of course, this does not mean these inequalities can be compensated for through topic choice in a dissertation alone. Instead, this result points to the importance of studying the interplay of topic choice, status, and inequality in academia (even more for a scholar’s entire record, see Bandelj 2019; Key and Sumner 2019; Koppman and Leahey 2019).

Figure 5.

Cumulative Hazards for Time to First Advisorship, by Gender, Race, and Targeted Specialization in T50 Methods: Modeling

Summing up our results on targeted specialization, we find that topics related to identity, race, and quantitative modeling positively contribute to students’ chances of becoming an advisor in sociology. In contrast, students specializing in survey methods and health in their theses are less likely to advise sociology-related PhDs. This sheds new light on which sociological topics garner influential field positions.

The Influence of Focus, Novelty, and Consistency on Becoming an Advisor in Sociology

In addition to the influence of targeted specialization, we explore three facets of specialization discussed prominently in prior literature on academics’, and more generally, creative workers’, success: focus, novelty, and consistency. By breaking specialization into these analytic dimensions, we are able to identify which facet of specialization is most related to becoming an advisor, and net of the specific topic choices made by PhD students.

Regarding Hypothesis 1, we find no significant effect (p < .05) of focus in our models (Table 3). The effect of focus vanishes after considering the influence of targeted specialization and other facets of specialization that are missing controls in previous studies (e.g., Conti et al. 2014; Leahey 2007; Schilling and Green 2011; Teodoridis et al. 2019).²⁰

Hypothesis 2 asks whether novel or traditional combinations of topics in a thesis promote PhD graduates’ career chances. Across all models, students integrating novel combinations of topics in their thesis have close to a 10 percent higher chance of advising sociology PhDs themselves. Again, this is net of other forms of specialization. The sociological field thus seems to reward unusual blends of thought styles expressed in early stages of a scholar’s career. Of course, the specific chains of mechanisms that connect early-stage innovation and later-stage career success will require further study. However, our results are in keeping with Leahey and Moody’s (2014) findings that sociology rewards innovations, such as subfield integration.

Our final hypothesis (Hypothesis 3) concerns consistency, and switches the text source from the dissertation to a student’s publication track. We find scholars who publish articles consistent with their dissertation have a better chance of becoming advisors. The size of this effect is comparable to the largest effects of specific topic choices. However, the effect is particularly strong for moderate values of consistency and marginally decreases for higher values (i.e., has a negative squared effect). This supports prior work showing that a consistent identity is essential to promote novel work (Trapido 2015) or unusual methods (Koppman and Leahey 2019). Our finding adds further nuance, however, by suggesting there are diminishing returns to identity consistency.

The consistency metric rests on students’ number of publications during their careers. Therefore, we investigate the relationship between publication output over students’ careers and their identity consistency.²¹ Figure 6 demonstrates that high productivity can decrease the likelihood of mentorship at low levels of consistency; in other words, a dramatically changing research identity is not rewarded. In contrast, only a small number of publications with a consistent thematic angle (in reference to the thesis) provides better chances to achieve a position as an advisor. The relation between identity consistency and productivity, therefore, can describe significant differences in career outcomes.

Figure 6.

Marginal Effects of Interaction between Consistency and Number of Publications

Discussion

Even the most experienced scientists perceive their discipline from a necessarily limited individual perspective. The sociology of science (Merton 1973), and, more recently, science of science (Fortunato et al. 2018; Wang and Barabási 2021), provide a more comprehensive bird’s eye view on academe’s inner-workings. A key concern in these areas of study is the fierce selection process inherent to all scientific fields, that is, who achieves an academic career and by what strategy? In this article, we disentangle distinct facets of specialization as potential career strategies (targeted topics, focus, novelty, and consistency), provide a consistent methodological frame to measure them (structural topic modeling), and use an understudied population (PhD students) to capture how specialization strategies used at an early career stage relate to long-term success in academia.

Our results show across a number of robustness checks that it is not topical focus of a thesis that provides career advantages, but targeted specialization and novel combinations. In particular, theses closely related to the cultural turn (T2 identity), statistical approaches (T50 modeling), and race (T58 race: African American) correspond with rare ascents to mentorship. Dissertations targeting survey methods (T14) are, in contrast, less likely to bring the graduate an advisorship. In addition, we find that departures from tradition and embracing novel combinations of topics help graduates become advisors. Last, and most important, we find that students whose ensuing publications echo and elaborate their dissertations exhibit a consistent research identity and have a considerably higher chance of becoming an advisor. Whereas there are diminishing returns to consistency for highly productive scholars, a relatively small number of consistent publications significantly improves the likelihood of achieving an advisor position, adding important nuance to the publish or perish imperative.

The social sciences are not a favored source of case studies in the sociology of science (Leahey 2008). Yet, we chose to study sociology because of the long-standing polemic in this field regarding specialization. On the one hand, specialization has been recognized as a necessary strategy in a massified field with narrowing job opportunities (Collins 1986). On the other hand, specialization is said to be responsible for the discipline’s fragmentation (Coser 1990; Gitlin 1990). In our field, specialization is a multifaceted, contentious, high-stakes career strategy. Our approach was to analytically and empirically distinguish these facets of specialization. With the exception of targeted specialization, the facets are abstract enough to provide a common analytic framework that can be applied to multiple academic fields.

There are many compelling prospects for such a comparative study of academic specialization. The first one is that facets of specialization may vary in relevance across fields that participate in similar intellectual movements. For instance, academic shifts like the cultural turn (Jacobs and Spillman 2005) encompass the social sciences as a whole, with fits and starts specific to each discipline. Yet, we do not know if underneath this shared thematic surface lie different relations between focus, innovation, consistency, and academic success. Does the combination of novel and consistent specialization yield similar advantages for career prospects in other disciplines? And what about specific topic choices? We have yet to uncover how common themes come to express distinct ways of vying for recognition across academic fields.

A second prospect of a comparative sociology of specialization is the study of migration. Our interest in the reproduction of the sociology field led us to narrowly focus on who becomes an advisor in a sociology department or of a sociology-related dissertation. However, one could imagine other outcomes of interest, such as becoming an advisor in a neighboring or distal field, or becoming an advisor in a focused or eclectic set of topics. On this point, there is not only an opportunity to better understand academics’ individual projects, but also the way the academic system distributes rewards and risks in the form of migration patterns.

A third opportunity of this prospective research program would be the comparative study of field relations to government and industry. In a prior study of education research, we found that graduate students grew more interested in interpretive research approaches just as pressure mounted from the government to norm the field according to the standard of randomized control trials (Munoz-Najar Galvez et al. 2020). These results were suggestive of the emergence of a particular configuration of field autonomy, expressed in how education graduate students chose to specialize. Fields with different relations to industry and government may achieve autonomy in varied ways, which, in turn, should lead students to seek different forms of specialization.

In focusing on a long-term career outcome of specialization (i.e., advisorship), we omitted the lived conditions of scholars that may lead them to choose and sustain this strategy. Future research will do well to embed specialization in a broader system of career checkpoints—spanning from undergraduate to tenure (Meyers et al. 2018)—working arrangements (Cech and Blair-Loy 2014), and family formations (Wolfinger, Mason, and Goulden 2009). In addition, linking PhD students’ initial thematic tracks with alternative outcomes could prove fruitful. People who have gained a sure footing in academia as an advisor might also occupy other important positions, for example, in councils, committees, or media discourse. Thus, connecting the dots of scientists’ professional life course before and after the PhD seems a worthwhile step over and above our findings.

A key limitation of our study is that we only have access to primary advisors to construct our outcome variable.²² In addition, we have no information on student drop-out rates, so we cannot control for selection biases happening before graduation. Nevertheless, we believe the focus on graduating students and their timing to becoming a (primary) advisor is a conservative proxy for academic success, given the resources necessary to achieve doctoral mentorship. We also note that our results are limited to the time period and set of students we study. The stability of our results over 30 years suggests much of this pattern may still hold, such as the extant divisions in the field and specific topics and strategies that secure advising roles. However, we also acknowledge that the period from 2015 to 2021 has been especially tumultuous socially and has likely resulted in further change. It may be that the extant political context has resulted in a sea change for hiring and what matters.

Supplemental Material

sj-pdf-1-asr-10.1177_00031224211056267 – Supplemental material for Facets of Specialization and Its Relation to Career Success: An Analysis of U.S. Sociology, 1980 to 2015

Supplemental material, sj-pdf-1-asr-10.1177_00031224211056267 for Facets of Specialization and Its Relation to Career Success: An Analysis of U.S. Sociology, 1980 to 2015 by Raphael H. Heiberger, Sebastian Munoz-Najar Galvez and Daniel A. McFarland in American Sociological Review

Footnotes

Acknowledgements

Special thanks to David Jurgens, Bas Hofstra, and Vamshi Krishna for help with data collection and architecture. Our gratitude as well to Joshua Gagne, Charles Gomez, Patrick Kaminski, E. Nadia Kline, Benjamin Keep, Elina Mäkinen, Richard Münch, Peter McMahan, and Michael Windzio. These data were collected by the Mimir Project, conducted at Stanford University.

Funding

The data collection on this project was partly funded by the Office of the President at Stanford University and other aspects were partly funded by the National Science Foundation [Awards 1827477, SMA1829240, and 1633036]. We are also grateful for funding received from the Fulbright Program, the Fritz Thyssen Foundation, and the German Federal Ministry of Education and Research [Award 01PU17021C].

ORCID iDs

Raphael H. Heiberger

Sebastian Munoz-Najar Galvez

Daniel A. McFarland

Notes

Raphael H. Heiberger is a tenure-track professor at the University of Stuttgart and head of a lab on computational social science. He graduated from the Otto-Friedrich University of Bamberg and was a Fulbright Fellow at UC-Berkeley. His research interests revolve around statistical models, scientific careers, political polarization, and financial markets.

Sebastian Munoz-Najar Galvez is the Bluhm Family Assistant Professor of Data Science and Education at the Harvard Graduate School of Education. He studies how research agendas form, and how they accrue public value.

Daniel A. McFarland is a professor of education, and by courtesy, sociology and organizational behavior at Stanford University. His current research is focused on the sociology of science and knowledge innovation.

References

Abbott

Andrew

. 2001. Chaos of Disciplines. Chicago: University of Chicago Press.

Add Health. 2018. “Publications” (http://www.cpc.unc.edu/projects/addhealth/publications).

Alexander

Jeffrey C.

1987. Sociological Theory since 1945. London, UK: Routledge.

Alexander

Jeffrey C.

1988. “The New Theoretical Movement.” Pp. 77–101 in Handbook of Sociology, edited by Smelser

N. J.

Beverly Hills, CA: Sage Publications.

American Sociological Association (ASA). 2021a. “Mean Number of Tenured and Tenure-Track Sociology Faculty” (https://www.asanet.org/academic-professional-resources/data-about-disipline/data-dashboard/sociology-programs/mean-number-tenured-and-tenure-track-sociology-faculty-institution-type).

American Sociological Association (ASA). 2021b. “Doctorates Awarded in Sociology, by Gender” (https://www.asanet.org/academic-professional-resources/data-about-disipline/data-dashboard/degrees-awarded/doctorates-awarded-sociology-gender).

American Sociological Association (ASA). 2021c. “Doctorates Awarded in Sociology, by Race or Ethnicity” (https://www.asanet.org/academic-professional-resources/data-about-disipline/data-dashboard/degrees-awarded/doctorates-awarded-sociology-race-or-ethnicity).

Amorim

Leila DAF

Cai

Jianwen

. 2015. “Modelling Recurrent Events: A Tutorial for Analysis in Epidemiology.” International Journal of Epidemiology 44(1):324–33.

Anderson

Ashton

McFarland

Daniel A.

Jurafsky

Dan

. 2012. “Towards a Computational History of the ACL: 1980–2008.” Pp. 13–21 in Proceedings of the ACL-2012 Special Workshop on Rediscovering 50 Years of Discoveries. Stroudsburg, PA: Association for Computational Linguistics.

10.

Antons

David

Joshi

Amol M.

Salge

Torsten Oliver

. 2019. “Content, Contribution, and Knowledge Consumption: Uncovering Hidden Topic Structure and Rhetorical Signals in Scientific Texts.” Journal of Management 45(7):3035–76.

11.

Astley

W. Graham

. 1985. “Administrative Science as Socially Constructed Truth.” Administrative Science Quarterly 30(4):497–513.

12.

Balsmeier

Benjamin

Pellens

Maikel

. 2014. “Who Makes, Who Breaks: Which Scientists Stay in Academe?” Economics Letters 122(2):229–32.

13.

Bandelj

Nina

. 2019. “Academic Familism, Spillover Prestige and Gender Segregation in Sociology Subfields: The Trajectory of Economic Sociology.” The American Sociologist 50(4):488–508.

14.

Besbris

Max

Khan

Shamus

. 2017. “Less Theory. More Description.” Sociological Theory 35(2):147–53.

15.

Best

Steven

. 2007. “Culture Turn.” The Blackwell Encyclopedia of Sociology. Hoboken, NJ: Wiley-Blackwell.

16.

Blei

David M.

2012. “Probabilistic Topic Models.” Communications of the ACM 55(4):77–84.

17.

Blei

David M.

Andrew Y.

Jordan

Michael I.

2003. “Latent Dirichlet Allocation.” The Journal of Machine Learning Research 3:993–1022.

18.

Blossfeld

Hans-Peter

Rohwer

Götz

Schneider

Thorsten

. 2019. Event History Analysis with Stata. London, UK: Routledge.

19.

Bourdieu

Pierre

. 1988. Homo Academicus. Stanford, CA: Stanford University Press.

20.

Box-Steffensmeier

Janet M.

Jones

Bradford S.

1997. “Time Is of the Essence: Event History Models in Political Science.” American Journal of Political Science 41(4):1414–61.

21.

Box-Steffensmeier

Janet M.

Jones

Bradford S.

2004. Event History Modeling: A Guide for Social Scientists. Cambridge, UK: Cambridge University Press.

22.

Burris

Val

. 2004. “The Academic Caste System: Prestige Hierarchies in PhD Exchange Networks.” American Sociological Review 69(2):239–64.

23.

Burt

Ronald S.

2004. “Structural Holes and Good Ideas.” American Journal of Sociology 110(2):349–99.

24.

Cech

Erin A.

Blair-Loy

Mary

. 2014. “Consequences of Flexibility Stigma among Academic Scientists and Engineers.” Work and Occupations 41(1):86–110.

25.

Cech

Erin A.

Rubineau

Brian

Silbey

Susan

Seron

Caroll

. 2011. “Professional Role Confidence and Gendered Persistence in Engineering.” American Sociological Review 76(5):641–66.

26.

Coleman

James S.

1981. Longitudinal Data Analysis. New York: Basic Books.

27.

Collins

Randall

. 1986. “Is 1980s Sociology in the Doldrums?” American Journal of Sociology 91(6):1336–55.

28.

Collins

Randall

. 2002. The Sociology of Philosophies: A Global Theory of Intellectual Change, rev. ed. Cambridge, MA: Belknap Press of Harvard University Press.

29.

Conti

Raffaele

Gambardella

Alfonso

Mariani

Myriam

. 2014. “Learning to Be Edison: Inventors, Organizations, and Breakthrough Inventions.” Organization Science 25(3):833–49.

30.

Correll

Shelley J.

Ridgeway

Cecilia L.

Zuckerman

Ezra W.

Jank

Sharon

Jordan-Bloch

Sara

Nakagawa

Sandra

. 2017. “It’s the Conventional Thought That Counts: How Third-Order Inference Produces Status Advantage.” American Sociological Review 82(2):297–327.

31.

Coser

Lewis A.

1990. “The Virtues of Dissent in Sociology.” Pp. 207–13 in Sociology in America. London, UK: Sage Publications.

32.

Crane

Diana

Small

Henry

. 1992. “American Sociology since the Seventies: The Emerging Identity Crisis in the Discipline.” Pp. 197–234 in Sociology and Its Publics: The Forms and Fates of Disciplinary Organization, edited by Halliday

T. C.

Janowitz

Chicago: University of Chicago Press.

33.

Du Bois

W. E. B.

[1897] 2020. The Conservation of Races. New York: Good Press.

34.

Evans

James A.

Aceves

Pedro

. 2016. “Machine Translation: Mining Text for Social Theory.” Annual Review of Sociology 42(1):21–50.

35.

Ferguson

John-Paul

Hasan

Sharique

. 2013. “Specialization and Career Dynamics: Evidence from the Indian Administrative Service.” Administrative Science Quarterly 58(2):233–56.

36.

Fleck

Ludwik

. 1979. Genesis and Development of a Scientific Fact. Chicago: University of Chicago Press.

37.

Fleming

Lee

. 2001. “Recombinant Uncertainty in Technological Search.” Management Science 47(1):117–32.

38.

Fleming

Lee

Mingo

Santiago

Chen

David

. 2007. “Collaborative Brokerage, Generative Creativity, and Creative Success.” Administrative Science Quarterly 52(3):443–75.

39.

Fortunato

Santo

Bergstrom

Carl T.

Börner

Katy

Evans

James A.

Helbing

Dirk

Milojević

Staša

Petersen

Alexander M.

Radicchi

Filippo

Sinatra

Roberta

Uzzi

Brian

Vespignani

Alessandro

Waltman

Ludo

Wang

Dashun

Barabási

Albert-László

. 2018. “Science of Science.” Science 359(6379):eaao0185.

40.

Foster

Jacob G.

Rzhetsky

Andrey

Evans

James A.

2015. “Tradition and Innovation in Scientists’ Research Strategies.” American Sociological Review 80(5):875–908.

41.

Frickel

Scott

Gross

Neil

. 2005. “A General Theory of Scientific/Intellectual Movements.” American Sociological Review 70(2):204–32.

42.

Giddens

Anthony

. 1976. “Classical Social Theory and the Origins of Modern Sociology.” American Journal of Sociology 81(4):703–29.

43.

Gieryn

Thomas F.

1978. “Problem Retention and Problem Change in Science.” Sociological Inquiry 48(3–4):96–115.

44.

Gitlin

Todd

. 1990. “Sociology for Whom? Criticism for Whom?” Pp. 214–26 in Sociology in America. London, UK: SAGE Publications.

45.

Grant

Linda

Ward

Kathryn B.

Rong

Xue Lan

. 1987. “Is There an Association between Gender and Methods in Sociological Research?” American Sociological Review 52(6):856–62.

46.

Griffiths

Thomas L.

Steyvers

Mark

. 2004. “Finding Scientific Topics.” Proceedings of the National Academy of Sciences 101:5228–35.

47.

Hand

Carl M.

Judkins

Ben

. 1999. “Disciplinary Schisms: Subspecialty ‘Drift’ and the Fragmentation of Sociology.” The American Sociologist 30(1):18–36.

48.

Hannan

Michael T.

Le Mens

Gaël

Hsu

Greta

Kovács

Balázs

Negro

Giacomo

Pólos

László

Pontikes

Elizabeth

Sharkey

Amanda J.

2019. Concepts and Categories: Foundations for Sociological and Cultural Analysis. New York: Columbia University Press.

49.

Hofstra

Bas

Corten

Rense

van Tubergen

Frank

Ellison

Nicole B.

2017. “Sources of Segregation in Social Networks: A Novel Approach Using Facebook.” American Sociological Review 82(3):625–56.

50.

Hofstra

Bas

Kulkarni

Vivek V.

Galvez

Sebastian Munoz-Najar

Bryan

Jurafsky

Dan

McFarland

Daniel A.

2020. “The Diversity–Innovation Paradox in Science.” Proceedings of the National Academy of Sciences 117(17):9284–91.

51.

Hofstra

Bas

de Schipper

Niek C.

2018. “Predicting Ethnicity with First Names in Online Social Media Networks.” Big Data & Society (https://doi.org/10.1177/2053951718761141).

52.

Holmwood

John

. 2010. “Sociology’s Misfortune: Disciplines, Interdisciplinarity and the Impact of Audit Culture.” The British Journal of Sociology 61(4):639–58.

53.

Hoppe

Travis A.

Litovitz

Aviva

Willis

Kristine A.

Meseroll

Rebecca A.

Perkins

Matthew J.

Hutchins

B. Ian

Davis

Alison F.

Lauer

Michael S.

Valantine

Hannah A.

Anderson

James M.

Santangelo

George M.

2019. “Topic Choice Contributes to the Lower Rate of NIH Awards to African-American/Black Scientists.” Science Advances 5(10) (https://doi.org/10.1126/sciadv.aaw7238).

54.

Huang

Junming

Gates

Alexander J.

Sinatra

Roberta

Barabási

Albert-László

. 2020. “Historical Comparison of Gender Inequality in Scientific Careers across Countries and Disciplines.” Proceedings of the National Academy of Sciences 117(9):4609–16.

55.

Inglis

Matthew

Foster

Colin

. 2018. “Five Decades of Mathematics Education Research.” Journal for Research in Mathematics Education 49(4):462–500.

56.

Jacobs

Mark D.

Spillman

Lyn

. 2005. “Cultural Sociology at the Crossroads of the Discipline.” Poetics 33(1):1–14.

57.

Jacomy

Mathieu

Venturini

Tommaso

Heymann

Sebastien

Bastian

Mathieu

. 2014. “ForceAtlas2, a Continuous Graph Layout Algorithm for Handy Network Visualization Designed for the Gephi Software.” PLoS ONE 9(6):e98679.

58.

Key

Ellen M.

Sumner

Jane Lawrence

. 2019. “You Research Like a Girl: Gendered Research Agendas and Their Implications.” PS: Political Science & Politics 52(4):663–68.

59.

Knorr Cetina

Karin

. 2009. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press.

60.

Koppman

Sharon

Leahey

Erin E.

2019. “Who Moves to the Methodological Edge? Factors That Encourage Scientists to Use Unconventional Methods.” Research Policy 48(9):103807.

61.

Kuhn

Thomas S.

1959. “The Essential Tension: Tradition and Innovation in Scientific Research.” Pp. 162–74 in The Third University of Utah Research Conference on the Identification of Scientific Talent, edited by Taylor

C. W.

Salt Lake City: University of Utah Press.

62.

Kuhn

Thomas S.

1962. Structure of Scientific Revolutions. Chicago: University of Chicago Press.

63.

Lamont

Michèle

Molnár

Virág

. 2002. “The Study of Boundaries in the Social Sciences.” Annual Review of Sociology 28:167–95.

64.

Leahey

Erin

. 2006. “Transmitting Tricks of the Trade: Advisors and the Development of Research Knowledge.” Teaching Sociology 34(2):93–110.

65.

Leahey

Erin

. 2007. “Not by Productivity Alone: How Visibility and Specialization Contribute to Academic Earnings.” American Sociological Review 72(4):533–61.

66.

Leahey

Erin

. 2008. “Methodological Memes and Mores: Toward a Sociology of Social Research.” Annual Review of Sociology 34(1):33–53.

67.

Leahey

Erin

Beckman

Christine M.

Stanko

Taryn L.

2017. “Prominent but Less Productive: The Impact of Interdisciplinarity on Scientists’ Research.” Administrative Science Quarterly 62(1):105–39.

68.

Leahey

Erin

Keith

Bruce

Crockett

Jason

. 2010. “Specialization and Promotion in an Academic Discipline.” Research in Social Stratification and Mobility 28(2):135–55.

69.

Leahey

Erin

Moody

James

. 2014. “Sociological Innovation through Subfield Integration.” Social Currents 1(3):228–56.

70.

Levin

Michael

Krawczyk

Stefan

Bethard

Steven

Jurafsky

Dan

. 2012. “Citation-Based Bootstrapping for Large-Scale Author Disambiguation.” Journal of the American Society for Information Science and Technology 63(5):1030–47.

71.

Lutter

Mark

Schröder

Martin

. 2016. “Who Becomes a Tenured Professor, and Why? Panel Data Evidence from German Sociology, 1980–2013.” Research Policy 45(5):999–1013.

72.

Magubane

Zine

. 2016. “American Sociology’s Racial Ontology: Remembering Slavery, Deconstructing Modernity, and Charting the Future of Global Historical Sociology.” Cultural Sociology 10(3):369–84.

73.

Manning

Christopher D.

Raghavan

Prabhakar

Schütze

Hinrich

. 2008. Introduction to Information Retrieval. Cambridge, UK: Cambridge University Press.

74.

McCall

Leslie

. 2002. Complex Inequality: Gender, Class and Race in the New Economy. London, UK: Routledge.

75.

Merton

Robert K.

1973. The Sociology of Science. Chicago: The University of Chicago Press.

76.

Meyer

Bruce D.

Mok

Wallace K. C.

Sullivan

James X.

2015. “Household Surveys in Crisis.” Journal of Economic Perspectives 29(4):199–226.

77.

Meyers

Lindsay

Brown

Abigail

Moneta-Koehler

Liane

Chalkley

Roger

. 2018. “Survey of Checkpoints along the Pathway to Diverse Biomedical Research Faculty.” PLoS One 13(1):e0190606.

78.

Molina

Mario

Garip

Filiz

. 2019. “Machine Learning for Sociology.” Annual Review of Sociology 45(1):27–45.

79.

Moody

James

. 2004. “The Structure of a Social Science Collaboration Network: Disciplinary Cohesion from 1963 to 1999.” American Sociological Review 69(2):213–38.

80.

Moore

Joan

. 1990. “Criticism for Whom? Social Criticism and Sociological Elitism.” Pp. 227–30 in Sociology in America. London, UK: Sage Publications.

81.

Morgan

Stephen L.

Gelbgiser

Dafna

Weeden

Kim A.

2013. “Feeding the Pipeline: Gender, Occupational Plans, and College Major Selection.” Social Science Research 42(4):989–1005.

82.

Morris

Aldon

. 2015. The Scholar Denied: W. E. B. Du Bois and the Birth of Modern Sociology. Oakland: University of California Press.

83.

Munoz-Najar Galvez

Sebastian

Heiberger

Raphael H.

McFarland

Daniel

. 2020. “Paradigm Wars Revisited: A Cartography of Graduate Research in the Field of Education (1980–2010).” American Educational Research Journal 57(2):612–52.

84.

National Science Foundation. 2020. “Survey of Earned Doctorates” (https://ncsesdata.nsf.gov/home/).

85.

Paré

Anthony

Starke-Meyerring

Doreen

McAlpine

Lynn

. 2011. “Knowledge and Identity Work in the Supervision of Doctoral Student Writing: Shaping Rhetorical Subjects.” Pp. 215–36 in Writing in Knowledge Societies, edited by Starke-Meyerring

Paré

Artemeva

Horne

Yousoubova

WAC Clearinghouse (https://doi.org/10.37514/PER-B.2011.2379.2.11).

86.

Platt

Jennifer

. 1998. A History of Sociological Research Methods in America, 1920–1960. Cambridge, UK: Cambridge University Press.

87.

Pound

Pandora

Campbell

Rona

. 2015. “Locating and Applying Sociological Theories of Risk-Taking to Develop Public Health Interventions for Adolescents.” Health Sociology Review 24(1):64–80.

88.

Rawlings

Craig M.

McFarland

Daniel

Dahlander

Linus

Wang

Dan

. 2015. “Streams of Thought: Knowledge Flows and Intellectual Cohesion in a Multidisciplinary Era.” Social Forces 93(4):1687–722.

89.

Rhoades

Lawrence J.

1981. A History of the American Sociological Association, 1905–1980. Washington, DC: American Sociological Association.

90.

Roberts

Margaret E.

Stewart

Brandon M.

Airoldi

Edoardo M.

2016. “A Model of Text for Experimentation in the Social Sciences.” Journal of the American Statistical Association 111(515):988–1003.

91.

Roberts

Margaret E.

Stewart

Brandon M.

Tingley

Dustin

Lucas

Christopher

Leder-Luis

Jetson

Gadarian

Shana Kushner

Albertson

Bethany

Rand

David G.

2014. “Structural Topic Models for Open-Ended Survey Responses.” American Journal of Political Science 58(4):1064–82.

92.

Schilling

Melissa A.

Green

Elad

. 2011. “Recombinant Search and Breakthrough Idea Generation: An Analysis of High Impact Papers in the Social Sciences.” Research Policy 40(10):1321–31.

93.

Schwemmer

Carsten

Wieczorek

Oliver

. 2020. “The Methodological Divide of Sociology: Evidence from Two Decades of Journal Publications.” Sociology 54(1):3–21.

94.

Shi

Xiaolin

Leskovec

Jure

McFarland

Daniel A.

2010. “Citing for High Impact.” Pp. 49–58 in Joint Conference on Digital Libraries (JCDL 2010). Brisbane, Australia (https://doi.org/10.1145/1816123.1816131).

95.

Sinatra

Roberta

Wang

Dashun

Deville

Pierre

Song

Chaoming

Barabási

Albert-László

. 2016. “Quantifying the Evolution of Individual Scientific Impact.” Science 354(6312) (https://doi.org/10.1126/science.aaf5239).

96.

Singer

Judith D.

Willett

John B.

2003. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. Oxford, UK: Oxford University Press.

97.

Swedberg

Richard

. 2017. “Theorizing in Sociological Research: A New Perspective, a New Departure?” Annual Review of Sociology 43:189–206.

98.

Teodoridis

Florenta

Bikard

Michaël

Vakili

Keyvan

. 2019. “Creativity at the Knowledge Frontier: The Impact of Specialization in Fast- and Slow-Paced Domains.” Administrative Science Quarterly 64(4):894–927.

99.

Teplitskiy

Misha

Acuna

Daniel

Elamrani-Raoult

Aïda

Körding

Konrad

Evans

James

. 2018. “The Sociology of Scientific Validity: How Professional Networks Shape Judgement in Peer Review.” Research Policy 47(9):1825–41.

100.

Trapido

Denis

. 2015. “How Novelty in Knowledge Earns Recognition: The Role of Consistent Identities.” Research Policy 44(8):1488–500.

101.

Uzzi

Brian

Mukherjee

Satyam

Stringer

Michael

Jones

Ben

. 2013. “Atypical Combinations and Scientific Impact.” Science 342(6157):468–72.

102.

van den Besselaar

Peter

Sandström

Ulf

. 2016. “Gender Differences in Research Performance and Its Impact on Careers: A Longitudinal Case Study.” Scientometrics 106(1):143–62.

103.

Wang

Dashun

Barabási

Albert-László

. 2021. The Science of Science. Cambridge, UK: Cambridge University Press.

104.

Warren

John Robert

. 2019. “How Much Do You Have to Publish to Get a Job in a Top Sociology Department? Or to Get Tenure? Trends over a Generation.” Sociological Science 6:172–96.

105.

Weeber

Stan C.

2006. “Elite versus Mass Sociology: An Elaboration on Sociology’s Academic Caste System.” The American Sociologist 37(4):50–67.

106.

Wolfinger

Nicholas

Mason

Mary A.

Goulden

Marc

. 2009. “Stay in the Game: Gender, Family Formation and Alternative Trajectories in the Academic Life Course.” Social Forces 87(3):1591–621.

107.

Znaniecki

Florian

. 1945. “Controversies in Doctrine and Method.” American Journal of Sociology 50(6):514–21.

108.

Zuckerman

Ezra W.

Kim

Tai-Young

Ukanwa

Kalinda

von Rittmann

James

. 2003. “Robust Identities or Nonentities? Typecasting in the Feature-Film Labor Market.” American Journal of Sociology 108(5):1018–74.

109.

Zuckermann

Harriet

. 1977. Scientific Elite: Nobel Laureates in the United States. New York: Free Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.54 MB

Facets of Specialization and Its Relation to Career Success: An Analysis of U.S. Sociology,1980 to 2015

Abstract

Keywords

Specialization as Career Strategy in Sociology

The Dissertation Genre

Becoming an Advisor

Facets of Specialization

Specializing on Target Topics

Focus versus Breadth

Novelty versus Tradition

Consistency versus Change

Analytic Strategy

Sample

Identifying Thought Styles via Structural Topic Models

Dependent Variable

Independent Variables

Event-History Analysis

Results

Popular and Trending Topics in Sociological Dissertations

The Relation between Topic Choice and Becoming an Advisor in Sociology

Methods matter

Cultural turn, identity, and race

Neighboring fields

The Influence of Focus, Novelty, and Consistency on Becoming an Advisor in Sociology

Discussion

Supplemental Material

sj-pdf-1-asr-10.1177_00031224211056267 – Supplemental material for Facets of Specialization and Its Relation to Career Success: An Analysis of U.S. Sociology, 1980 to 2015

Footnotes

Acknowledgements

Funding

ORCID iDs

Notes

References

Supplementary Material