What gets published and what doesn’t? Exploring optimal distinctiveness and diverse expectations in entrepreneurship articles

Abstract

The field of entrepreneurship has seen remarkable growth, increasing the expectations of academic audiences. Articles need to balance novelty with rigorous methodology, theoretical contributions, social implications, and coherent argumentation to succeed in the publication process. However, navigating these varied and sometimes conflicting expectations to achieve optimal distinctiveness in academic narratives is challenging for authors. To explore how authors can achieve optimal distinctiveness amidst these complex expectations, we studied academic narratives and related editorial decisions of two leading entrepreneurship journals, Entrepreneurship: Theory & Practice (ETP, 4,151 papers) and Small Business Economics Journal (SBEJ, 4,043 papers), using computer-aided text analysis. Our study debunks common assumptions about what makes a successful entrepreneurship paper, providing an empirical basis for understanding actual versus perceived publication requisites. Furthermore, we extend optimal distinctiveness theory by demonstrating that high distinctiveness is not uniformly advantageous, meeting numerous expectations is not necessarily beneficial, and clear language is crucial for complex narratives. Our study underscores that crafting narratives is more nuanced than traditionally believed.

Plain English Summary

Getting published in Entrepreneurship Journals: Less is more! How can entrepreneurship scholars increase their chances of getting published? Our study delves into scholarly articles in entrepreneurship journals, investigating which papers are published and which papers are not. We challenge the assumption that authors must fulfil as many expectations as possible and emphasize the importance of addressing specific audience expectations. By analyzing narratives and editorial decisions from Entrepreneurship Theory & Practice (ETP) and Small Business Economics Journal (SBEJ), we uncover the key to publication success: tailoring articles to meet the targeted audience’s most pronounced requirements. Focusing on theoretical contributions when submitting one’s work to ETP and focusing on empirical contributions when submitting one’s work to SBEJ can increase the chances of getting your work published.

1 Introduction

The scholarly field of entrepreneurship has advanced tremendously in terms of quality and academic relevance in recent years (Audretsch, 2012; Davidsson, 2016; McMullen, 2019; Thurik et al., 2023). This increase in quality coincides with a rapidly growing number of submissions to leading entrepreneurship journals and, thus, higher expectations of scientific audiences such as editors and reviewers (Maula & Stam, 2020). Therefore, academic publications in the field of entrepreneurship increasingly need to meet the expectations of editors and reviewers such as theoretical contribution, methodological contribution, and academic writing. Nevertheless, adhering solely to these contributions and language expectations may not be sufficient for capturing the attention and interest of scientific audiences, given the vast amount of research being created in the field of entrepreneurship. Hence, entrepreneurship researchers need to also confirm novelty expectations and present narratives that are distinct. Consequently, papers are expected to convey distinctiveness while simultaneously adhering to contribution and language expectations; they are expected to be optimally distinct to achieve publication success (Patriotta, 2017).

However, achieving optimal distinctiveness when there are many different expectations is no easy feat. Authors may feel compelled to include excessive methodological details or incorporate an overwhelming number of theoretical explanations. Focusing simultaneously on meeting several expectations might, therefore, lead to narratives that are overly dense and difficult to comprehend (Kuechler & Vaishnavi, 2006). Additionally, different scientific audiences have varied preferences that might even be conflicting (Fisher et al., 2017), and it can be challenging to satisfy every individual reviewer or editor. Therefore, the following research question arises: How can authors achieve optimally distinct narratives when facing many different expectations?

To answer this question, we use computer-aided text analysis (CATA), investigating the last version of paper submission abstracts and their related editorial decisions in two leading entrepreneurship journals, Entrepreneurship: Theory & Practice (ETP, 4,151 papers) and Small Business Economics Journal (SBEJ, 4,043 papers), between 2017 and March 2022. Abstracts provide a concise summary of the entire paper, typically including the research question, methodology, results, and conclusions. Thus, the quality of the abstract can be seen as a proxy for the quality of the paper as a whole. Additionally, abstracts typically follow a standardized format that includes a fixed number of words. This standardization makes it easier to compare abstracts across different papers, journals, and research areas. More precisely, it avoids subjectively giving weight to particular sections of the paper, such as the results section in qualitative research articles. Therefore, investigating abstracts submitted to (and not only published) in these journals provides a unique context to understand the conditions under which narratives of papers are legitimate to get published.

The findings of our study reveal that the importance of distinctiveness in journal submissions is contingent on whether the majority of submissions are distinct or similar (i.e., not distinct) in nature (Haans, 2019). This highlights the crucial role of audiences in evaluating narratives, particularly within journal environments (Fisher et al., 2017). Furthermore, we observe that contribution claims significantly increase the likelihood of publication in prominent entrepreneurship journals. While theoretical contributions are paramount in Entrepreneurship Theory and Practice (ETP), Small Business Economics Journal (SBEJ) places greater emphasis on empirical contributions. Our findings provide initial evidence suggesting that meeting a wide range of expectations may not necessarily lead to better outcomes. Finally, our study establishes that effective academic writing enhances publication legitimacy. This discovery offers the indication that linguistic proficiency plays a crucial role in communicating optimally distinct narratives.

Our study makes two major contributions to the entrepreneurship literature. First, we seek to validate prevalent assumptions disseminated in editorials, books, and workshops regarding the writing style and content of entrepreneurship papers (e.g., Audretsch et al., 2022; Fayolle & Wright, 2014). We rigorously examine whether widely held beliefs about style and content characteristics align with publication success. This empirical approach sheds light on the actual versus perceived requisites of crafting research papers in the entrepreneurship domain. Second, our study contributes to optimal distinctiveness theory of narratives (Navis & Glynn, 2011) by exploring mechanisms related to a variety of audience expectations. The first mechanism highlights that the positive effect of distinctiveness (Taeuscher et al., 2021) disappears in environments where many competing narratives also score high in distinctiveness (Haans, 2019). The second mechanism proposes that meeting more expectations does not lead to better results. The third mechanism emphasizes that language use can play an important role in making highly distinct and/or complex narratives more comprehensive (Pennebaker et al., 2014). The exploration of these mechanisms suggests that crafting narratives that gain legitimacy is more nuanced than previously understood.

2 Theoretical background

2.1 Optimal distinctiveness of entrepreneurship articles

Critical scientific audiences, such as editors and reviewers, point to academic narratives as a crucial antecedent of publication success in management and entrepreneurship journals (Brattström & Wennberg, 2022; Patriotta, 2017; Pollock, 2021; Shepherd & Wiklund, 2020). We broadly define an academic narrative as a purposefully crafted narrative that aims to contribute to a research field.

Within the literature on narratives, optimal distinctiveness theory is one of the most established theories that aims to explain why some narratives gain legitimacy from critical audiences while others do not (Navis & Glynn, 2011). Optimal distinctiveness theory assumes that critical audiences, such as editors and reviewers, have to deal with many narratives (in our context: manuscript submissions). Hence, authors have to compete for their attention by being distinct (Landström & Harirchi, 2019; Salvato & Aldrich, 2012). Simultaneously, optimal distinctiveness theory assumes that critical audiences have expectations regarding narratives that authors need to meet to gain legitimacy (in our context: manuscripts being accepted for publication) (Brattström & Wennberg, 2022; Pollock, 2021). For example, in the case of research paper publications, audiences have expectations around the study’s theoretical (e.g., what does the study add to the current state of knowledge in the field?) (Barney, 2018; Corley & Gioia, 2011; Grant & Pollock, 2011; Rynes, 2002; Welter, 2011; Whetten, 1989), and empirical contribution (e.g., does methodological rigor exist?) (Anderson et al., 2019; Parker, 2020; Wennberg & Anderson, 2020). These two assumptions of optimal distinctiveness theory “give rise to a ‘double bind’ whereby authors are somehow ‘instructed’ to be innovative and surprise the reader while at the same time being expected to abide by the normative boundaries” of expectations (Patriotta, 2017, p. 748).

These normative boundaries are particularly challenging when there are various audience expectations. Authors may need to carefully consider and balance multiple dimensions of their narrative to accommodate these various expectations (Fisher et al., 2017). For example, they may need to simultaneously address theoretical and social contributions (George et al., 2016; Olsen et al., 2016; Wiklund et al., 2019). In such cases, authors may find it challenging to reconcile and meet all the diverse expectations of audiences. They need to make strategic decisions about which expectations to prioritize, depending on the specific context, the goals of their work, and the perceived importance of different reviewers or editors (Soublière & Lockwood, 2022).

2.2 Various expectations of entrepreneurship research audiences

To identify the various expectations from critical audiences of entrepreneurship research articles, we performed a literature review of editorials and articles related to publishing entrepreneurship research (see Table 1). This encompassed a careful examination of editorials, research articles, and other relevant academic publications discussing the nuances and standards of publishing within the entrepreneurship research domain. The results show three major types of expectations: (1) contribution-related, (2) novelty-related, and (3) language expectations.

Table 1 Expectations overview

2.2.1 The role of contribution-related expectations

We define contributions as advancements of existing knowledge that move the field of entrepreneurship forward and provide new directions for future research (Corley & Gioia, 2011). We distinguish three major types of contributions: theoretical contributions (e.g., Barney, 2018; Rynes, 2002; Whetten, 1989), empirical contributions (e.g., Anderson et al., 2019; Haans et al., 2016; Maula & Stam, 2020), and social contributions (Chen et al., 2022; Wickert et al., 2021).

First, theoretical contributions encompass endeavors of theory building (Corley & Gioia, 2011; Gioia et al., 2013; Whetten, 1989) and theory testing (Anderson et al., 2019; Haans et al., 2016), which serve essential roles in advancing knowledge in entrepreneurship research. Theory building involves a comprehensive exploration of diverse perspectives, actively participating in ongoing scholarly discourse (Craig, 2010; Shepherd & Wiklund, 2020). Theory building aims to map knowledge (Kraus et al., 2020) and address gaps, neglected areas, or sources of conceptual confusion prevalent within the field (Alvesson & Sandberg, 2011, Sandberg & Alvesson, 2011). This pursuit is further bolstered by an emphasis on interdisciplinary approaches, enabling the assimilation of insights from various disciplinary realms to enrich conceptualization (Eden, 2002; Shepherd & Wiklund, 2020; Tranfield et al., 2003). Conversely, theory testing entails subjecting extant theories to empirical investigation, including those engendered through theory building (Anderson et al., 2019; Haans et al., 2016). The primary objective of theory testing is to scrutinize and validate the proposed theoretical underpinnings, as well as to challenge their tenability through empirical research. Thereby, researchers can assess its applicability and relevance, ultimately contributing to the refinement and advancement of theoretical constructs.

Second, empirical contributions refer to original research findings that add to existing knowledge by providing new evidence through economic and/or exploratory insights. Economic insights elucidate the economic implications and consequences of research findings, thereby enhancing our understanding of entrepreneurial phenomena (Anderson et al., 2019; Parker, 2020). On the other hand, exploratory insights contribute to the search for new avenues or relationships within entrepreneurship, pushing the boundaries of knowledge and stimulating further research (Wennberg & Anderson, 2020). To achieve economic and/or exploratory insights, various articles have pointed to the research design as the basis of such contributions (Aguinis & Bradley, 2014; De Massis & Kotlar, 2014; Hsu et al., 2017). A research design outlines the methods, procedures, and techniques that will be used to collect and analyze data to obtain meaningful and reliable findings. For example, researchers are expected to carefully consider the choice of data sources (Harvey, 2011; Maula & Stam, 2020) and sampling techniques (De Massis & Kotlar, 2014; Eisenhardt, 1989, 2021). Furthermore, robust and appropriate analysis methods should be employed to address the research question, and the presentation of results should be concise and supported by suitable statistical analyses (Aguinis & Bradley, 2014; Molina-Azorín, 2011).

Third, recent trends indicate the relevance of social contribution, referring to the positive impact of research on individuals, organizations, or initiatives becoming increasingly important (Chen et al., 2022; Wiklund et al., 2019). More precisely, top-tier entrepreneurship journals are increasingly emphasizing research that has a positive impact on society by tackling grand challenges (George, 2016; George et al., 2016; Wiklund et al., 2019), providing practical implications (Chen et al., 2022; Colquitt & George, 2011; Kuckertz, 2012), and promoting diversity (Welter et al., 2017). This encourages new perspectives that extend beyond the traditional emphasis on theories around profitability and economic growth (Colquitt & George, 2011). This shift in focus can foster interdisciplinary collaborations and open doors to innovative research avenues (Chen et al., 2022).

2.2.2 The role of novelty-related expectations

Novelty-related expectations refer to the extent to which an audience anticipates a narrative to be original (Corley & Gioia, 2011; Parker, 2020), bold (Bacq et al., 2021; George, 2016), interesting (Landström & Harirchi, 2019; Salvato & Aldrich, 2012; von Krogh et al., 2012), and relevant (Reinartz, 2016; Shepherd & Wiklund, 2020). Indeed, objective indicators of novelty are often absent in journal articles; thus, ‘distinctiveness’ serves as a crucial reference point for audiences’ perceptions of novelty (Taeuscher et al., 2021) The presence of distinctiveness can have a favorable impact on the credibility of a publication, as long as the advantages associated with these expectations outweigh the cognitive drawbacks that come with distinctiveness (Taeuscher et al., 2021). This observation carries significant implications for the concept of "optimal distinctiveness," as it questions the belief that distinctiveness inherently hampers the establishment of legitimacy (Taeuscher et al., 2021). Consequently, we deviate from the widely held notion that entrepreneurship research articles inevitably encounter a conflict between distinctiveness and meeting audience expectations.

2.2.3 The role of language expectations

Academic writing is crucial in conveying unique academic narratives (Patriotta, 2017). A high level of academic writing refers to terms and formulations related to “formal, logical, and hierarchical thinking” (Pennebaker et al., 2014, 2015, p. 21). In contrast, a low level of academic writing reflects that terms and formulations follow “informal, personal, here and now, and narrative thinking” (Pennebaker et al., 2015, p. 21). Academic writing involves several key elements, such as problematization (Alvesson & Sandberg, 2011, Sandberg & Alvesson, 2011), structured presentation (e.g., Cochrane, 2005; Geletkanycz & Tepper, 2011), clear writing (Craig, 2010; Huff, 1999; Johanson, 2007), explanatory logic (Sparrowe & Mayer, 2011), and recognition of boundaries (Post et al., 2020). The relevance of academic writing is often emphasized by scholarly audiences as an essential aspect of their credentials, as it aligns with the requirement to express scientific findings clearly (e.g., Craig, 2010; Huff, 1999; Johanson, 2007) and logically (Sparrowe & Mayer, 2011). Editors also underscore the necessity of academic writing in providing better guidance for academic papers (Patriotta, 2017). Moreover, empirical evidence indicates that employing academic writing corresponds to educational success since it can demonstrate expertise (Pennebaker et al., 2014).

3 Method

To assess the role of optimal distinctiveness in academic narratives, our study proceeds in an exploratory way. This exploratory focus enables us to delve deeper into the intricate relationship between diverse expectations and optimal distinctiveness. To this end, we use computer-aided text analysis (CATA), analyzing paper abstracts (academic narratives) submitted to two leading entrepreneurship journals. Like other entrepreneurship studies that use CATA (e.g., Fisch & Block, 2021; Moss et al., 2018), we adopt a stepwise methodology to assess a large amount of text. In particular, we study paper submissions in three steps (see Fig. 1). First, we collect data on academic narratives and related editorial decisions. To compare papers across editorial decisions, we also apply a systematic filter logic. Second, we operationalize constructs using two specific forms of CATA: dictionary-based approaches and topic modelling. Third, we analyze data using binary logistic regression and check for the validity of our results by applying various robustness checks.

Fig. 1

Method in three steps

3.1 Empirical context

We gather data from two leading entrepreneurship journals: Entrepreneurship: Theory & Practice (ETP) and Small Business Economics Journal (SBEJ).

Entrepreneurship Theory & Practice (ETP) is a leading scholarly journal in entrepreneurship published by SAGE. With an impact factor of 10.5 (2022) and a 5-year impact factor of 14.4 (2022), the journal aims to publish original conceptual and empirical research that contributes to the advancement of entrepreneurship (Entrepreneurship Theory and Practice, 2023). Between ETP’s first publication in 1976 and 2023, the journal published more than 1,100 papers (Entrepreneurship: Theory and Practice, 2023). ETP represents a critical audience for entrepreneurship scholars.

Small Business Economics Journal (SBEJ) is a leading entrepreneurship journal published by Springer Science. SBEJ covers research into the field of entrepreneurship from different disciplines, including economics, finance, management, psychology, and sociology. SBEJ has an impact factor of 6.4 (2022) and a 5-year impact factor of 7.4 (2022). Between SBEJ’s first publication in 1989 and 2023, the journal published more than 2,600 papers (Small Business Economics, 2023). SBEJ represents a critical target audience for scholars studying entrepreneurship.

For both journals, we investigate papers that were submitted to and not just published in the respective journal. This provides a unique context to understand the conditions under which academic narratives are legitimate to be published.

3.2 Data

We collected data from the editorial managers of ETP and SBEJ between 2017 and summer 2022, counting 4,151 final-round paper submissions for ETP and 4,043 final-round paper submissions for SBEJ. The data encompasses information on editorial decisions and includes text data (i.e., abstracts and titles). The editorial decision indicates the success of a paper (i.e., whether the paper was rejected or accepted in the final round). Consequently, the data enables us to understand outcomes for academic narratives, determined by critical audiences (i.e., editors) and often based on the recommendations of another critical audience (i.e., reviewers).

In a few cases, the editorial decision of a paper was unclear. For example, some submissions were still under revision, incomplete, or withdrawn by the authors. Removing these submissions resulted in a sample of 3,973 papers for ETP and 3,661 papers for SBEJ. Moreover, we use computer-aided text analysis (CATA) to capture optimal distinctiveness of academic narratives by investigating abstracts (Haans, 2019; Taeuscher et al., 2021). In line with good research practice, we eliminated outliers by excluding abstracts of less than 50 words (e.g., Fisch & Block, 2021). This filter ensures that text analysis does not overestimate specific terms due to a limited number of words. Additionally, we filtered abstracts with more than 300 words since they strictly disregarded the guidelines of both journals. The final sample comprises the abstracts of 3,704 studies submitted to ETP and 3,592 studies submitted to SBEJ. While 7% of those studies were accepted in ETP, 9% were accepted in SBEJ (see Fig. 2).

Fig. 2

Dataset overview

We then checked whether abstracts were representative of full paper journal articles. In particular, we calculated the similarity of abstracts and full papers for all articles published in ETP (see Fig. 3) and SBEJ (see Fig. 4) between 2018 and 2020 using cosine similarity. The distribution indicates a bifurcation in abstract representation. While many abstracts offer moderate – high insight into full articles, a significant number of them provide an almost perfect representation. In light of this, we argue that a considerable number of abstracts in ETP and SBEJ are representative of their corresponding full texts.

Fig. 3

Similarity abstracts vs. fullpapers (ETP)

Fig. 4

Similarity abstracts vs. full papers (SBEJ)

3.3 Dependent variable

Publication legitimacy is our dependent variable. We proxy this variable by using the editorial decision of a paper. Editors’ final decisions indicate either the rejection or acceptance of a paper. Therefore, we construct a dummy variable for publication legitimacy: 1 = accepted, 0 = rejected.

3.4 Independent variables

We use CATA to capture our independent variables – theoretical contribution claims, empirical contribution claims, social contribution claims, distinctiveness, and academic writing. CATA can be executed through dictionary-based, rule-based, topic modelling, and machine learning approaches (Humphreys & Wang, 2018). First, dictionary-focused approaches identify constructs based on vocabulary and word counting (e.g., Fisch & Block, 2021; Meurer et al., 2022; Schou et al., 2022). Second, rule-based approaches refer to criteria such as sentiment patterns that researchers apply to structure data (e.g., Courtney et al., 2017). Third, classification techniques enable scholars to categorize texts based on a subset or training set of the data (e.g., Williamson et al., 2021). Fourth, topic modelling recognizes patterns within the data without predefined categories (e.g., Haans, 2019; Taeuscher et al., 2021).

Which approach is selected depends on how clearly the construct is outlined (Humphreys & Wang, 2018). If the concept is explicit (e.g., academic writing), a dictionary- or rule-based approach can be used to quantify it (Humphreys & Wang, 2018). Often, standardized dictionaries, such as LIWC-22, exist that researchers can refer to when capturing a construct. In the absence of a standardized dictionary, it is also possible to systematically create a dictionary capturing a predefined construct (e.g., Pennebaker et al. 2007; Short et al., 2009, 2010). However, if the identification of the concept in linguistics is not yet obvious or scholars intend to create a posteriori operationalization findings, a classification strategy is commonly used, in which the scholar defines text categories and then examines repeating linguistic patterns within these categories (Humphreys & Wang, 2018). In other cases, if the categories are not specified, the researcher can also use topic modelling or unsupervised machine learning that recognize groups within the text and then define the differences between those groups using patterns (Humphreys & Wang, 2018).

3.4.1 Theoretical contribution claims

We operationalize theoretical contribution claims using a dictionary-based approach. More precisely, we inductively develop a dictionary, following steps suggested in prior research (Payne et al., 2011; Short et al., 2009, 2010). First, we use CAT Scanner to explore exclusive words in both samples (ETP & SBEJ), resulting in 13,079 words. Second, one of the authors read through the word list and kept only the words related to theoretical writing conventions. Third, we sent the prefinal list of words to two independent entrepreneurship researchers that evaluated the dictionary. In this way, we ensured that the dictionary captures the construct sufficiently (Short et al., 2010). The final dictionary comprises 195 terms (see Appendix). After the approval of the dictionary, we use the R function LIWCalike to measure how many theoretical contribution claims are included per inquiry.

3.4.2 Empirical contribution claims

We capture empirical contribution claims using linguistic inquiry word count (LIWC). LIWC is one of the most established dictionary-based tools in management and entrepreneurship research (Fisch & Block, 2021). It can either be applied for self-developed dictionaries (see theoretical contribution claims) or is executed through the software LIWC-22. LIWC-22 includes a variety of linguistic (e.g., word count) and psychological measures (e.g., affection) (Boyd et al., 2022). Applying LIWC-22, we use the degree to which quantitative language is used (‘quantity’) to proxy empirical contribution claims. ‘Quantity’ is a linguistic variable and suitable to measure empirical contribution claims since it indicates a focus on economic and statistical insights.

3.4.3 Social contribution claims

We operationalize social contribution claims using a dictionary-based approach, specifically employing the self-developed dictionary of Moss et al. (2018). The choice of this dictionary was influenced by its comprehensive representation of both economic and social themes. While it was originally curated for a crowdfunding context, the wordlist effectively captures a diverse spectrum of economic and social themes that are prevalent in the broader academic discourse, especially within entrepreneurship literature. By using Moss et al. (2018) dictionary, we could leverage a tool that has undergone prior validation and application, providing our research with a foundation built upon peer-reviewed and recognized methodologies. To capture and quantify the presence of these themes in our inquiries, we utilized the R function LIWCalike, ensuring a consistent and systematic analysis. This approach not only facilitated our analysis by providing a ready-made and refined tool but also aligned perfectly with our research's aim to delve deep into the economic and social themes in the entrepreneurship literature.

3.4.4 Distinctiveness

We use a topic modelling approach to capture distinctiveness. Topic modelling enables distinctiveness to be reliably assessed as a multidimensional variable (Haans, 2019; Hannigan et al., 2019; Kaplan & Vakili, 2015; Taeuscher et al., 2021). We identified prevalent topics in academic narratives using Latent Dirichlet Allocation (LDA), the most widely used topic modelling approach (Antons et al., 2019; Banks et al., 2019). We balanced the trade-off between topic variety and simplicity of interpretation by limiting the number of topics to 100, which is in line with previous research (Haans, 2019; Kaplan & Vakili, 2015; Taeuscher et al., 2021). Next, we validated our topic models using techniques suggested by DiMaggio (2015). We then identified each paper as a probabilistic representation of the topics (Haans, 2019; Taeuscher et al., 2021). As a result, the discovered topics enabled us to quantify the extent to which the content of a paper differs from the content of prototypical submissions (Taeuscher et al., 2021). Thus, we follow Haans (2019) and calculate distinctiveness as

$$\sum_{T = 1}^{100}abs({\Theta }_{T,i}-{\overline{\Theta } }_{T})$$

where ${\Theta }_{T,i}$ indicates the weight of topic $T$ in paper $i$ and where ${\overline{\Theta } }_{T}$ indicates the average weight of topic $T$ in all submitted papers. The distinctiveness of a specific paper is thus measured as the total of the absolute deviances between the topic weights of paper $i$ and the average topic weight across 100 topics in all submitted papers. If a paper employs the same topic proportions as the average of all papers, its distinctiveness is zero.

3.4.5 Academic writing

Finally, we apply a dictionary-based approach to capture academic writing. Following Pennebaker et al. (2014), we use the summary variable ‘analytical thinking’ in LIWC-22 to proxy academic writing. A high score in analytical thinking indicates formal, logical, and hierarchical thinking, whereas a low number suggests more informal, personal, present focus, and narration in texts (Pennebaker et al., 2015).

3.4.6 Construct validity checks

We undertook manual checks on a subset of our data to validate the accuracy of our dictionary-based constructs. This helped us ensure that the instances where these words were flagged corresponded to actual claims of contribution and not mere descriptions.

3.5 Control variables

In line with other papers using CATA, we controled for linguistics and psychological language use (Fisch & Block, 2021). In particular, we controled for all summary variables included in LIWC-22, which has become a common practice in entrepreneurship research (Fisch & Block, 2021). To avoid overestimating the impact of certain terms, studies using LIWC commonly included word count as a control variable. Moreover, we included the number of words per sentence to capture whether authors use short sentences or long, complex sentences that are difficult to understand. Similarly, longer words indicate the percentage of words that are longer than six letters. We also controled for the ratio of dictionary terms included in a text, since some papers might incorporate unusual terms that are not captured in broad, standardized dictionaries. Last, we included the psychological summary variables authenticity, clout, and emotional tone to assure that the obtained results are not confounded by individual differences in these aspects and to control for potential biases they may introduce into the study. All controled variables were extracted from LIWC-22.

3.6 Quantitative analysis

We designed our research with the intent of delving into exploratory insights rather than merely testing for specific effects. Our study’s exploratory nature corresponds with recent calls for more exploratory research in the entrepreneurship field (e.g., Anderson et al., 2019). Our aim is to shed light on the intricate dynamics between diverse expectations and optimal distinctiveness. To achieve this objective, we examined the expectations from critical audiences of entrepreneurship research articles through an extensive literature review of editorials and articles related to publishing entrepreneurship research (see Table 1). This comprehensive review illuminated three major categories of expectations: contribution-related, novelty-related, and language-related expectations.

In response to these multifaceted expectations, we adopted a binary logistic regression approach that involves contrasting rejected papers with accepted papers, utilizing six distinct models to capture various facets of academic narratives. These models are selected to provide a comprehensive perspective on the determinants of publication outcomes and how they relate to the expectations of entrepreneurship research. Model 1 serves as a baseline, incorporating only control variables. Model 2 introduces contribution claims, while Model 3 focuses on distinctiveness claims. Model 4 delves into the realm of academic writing. In a bid to understand the interplay, Model 5 sheds light on both contribution claims and distinctiveness. Model 6 is an all-encompassing model, including all independent variables.

To ensure robustness in our exploratory findings, we first examine the sensitivity of our results to the operationalization of independent variables. We investigate whether employing alternative measures for these variables would yield congruent results. Additionally, in validating our regression models, we test for the goodness of fit of our models using the Hosmer–Lemeshow test, pseudo R-square, and variance inflation factors. Additionally, we check for the practical relevance of our findings using data plots and investigating odds ratios.

4 Results

4.1 Descriptive statistics

Table 2 shows the descriptive statistics. The last column indicates the p values obtained from an analysis of variance (ANOVA). The ANOVA findings reveal substantial differences between the two journals for several variables. For example, Table 2 shows that academic narratives submitted to ETP score lower in academic writing (p = 0.000), empirical contribution claims (p = 0.000), and distinctiveness (p = 0.000) than submissions to SBEJ. Furthermore, academic narratives submitted to ETP score higher in terms of theoretical (p = 0.000) and social contribution claims (p = 0.000).

Table 2 Descriptive statistics by journal (ETP and SBEJ)

Because our measure of publication legitimacy is binary, we can explore differences regarding editorial decisions (i.e., rejection or acceptance). Table 3 shows the mean value and standard deviation for each variable and editorial decision across both journals. The last column indicates whether there is a significant difference between rejected and accepted papers across both journals (ANOVA). The ANOVA findings reveal substantial differences regarding publication legitimacy and show that accepted articles are more distinct in their narratives than rejected articles (p = 0.000). Similarly, accepted articles show higher levels of analytical thinking (p = 0.000) and empirical contribution claims (p = 0.000).

Table 3 Descriptive statistics by editorial decision (acceptance and rejection)

Table 4 shows the bivariate correlations for all variables using the ETP dataset. The correlations range from -0.246 to.418. Similarly, Table 5 shows correlations for all variables using the SBEJ dataset. The correlations range from -0.320 to.266. To provide a more rigorous assessment, we computed the variance inflation factors (VIFs) for all variables (see Appendix) and created kernel density charts (see Appendix). We found that the VIFs range between 1 and 1.619. This confirms that multicollinearity is not a significant concern in our model.

Table 4 Correlations between all variables (ETP)

Table 5 Correlations between all variables (SBEJ)

4.2 Main analysis

4.2.1 Binary logistic regressions

Our main analysis employs binary logistic regressions to explore how meeting audience expectations impacts publication legitimacy. When examining the results presented in Tables 6 and 7, distinct nuances emerge between ETP and SBEJ academic narratives. Submissions to SBEJ typically showcase higher distinctiveness scores. However, for ETP, every one-point increase in this distinctiveness score is associated with a 22.8% rise in the odds of achieving publication legitimacy (p = 0.000). This suggests that while SBEJ more generally appreciates distinct content, ETP particularly rewards submissions that stand out from the crowd.

Table 6 Regression results with publication legitimacy (probability of paper acceptance) as the dependent variable (ETP)

Table 7 Regression results with publication legitimacy (probability of paper acceptance) as the dependent variable (SBEJ)

Turning our attention to contribution claims, the data reveals that for ETP, every additional word out of 100 that denotes a theoretical contribution correspond to a 0.2% increase in the odds of publication legitimacy (p = 0.021). In contrast, for SBEJ, each word out of 100 emphasizing an empirical contribution leads to a 0.8% increase in the odds of publication legitimacy (p = 0.000). Hence, academic narratives submitted to ETP have a greater emphasis on theoretical contributions, which seem to have an increasingly important role in achieving publication success in ETP compared to SBEJ. In SBEJ, theoretical contribution claims do not significantly increase the probability of publishing an article. While academic narratives submitted to SBEJ score lower in terms of theoretical contribution claims, they have a higher degree of empirical contribution claims, which is likely to be associated with SBEJ’s link to economic research.

Additionally, academic writing plays a pivotal role in both journals. In ETP, every one-point increase in the odds of academic writing results in a 0.2% growth in the odds of publication legitimacy (p = 0.000). Similarly, in SBEJ, each additional point in the score enhances the odds of achieving legitimacy by 0.3% (p = 0.000). These findings underscore the relevance of strong academic writing in achieving publication legitimacy. This suggests that authors aiming for publication in ETP and SBEJ should focus on academic writing to enhance the likelihood of their work being accepted for publication.

4.2.2 Alternative measurements

We test whether our results are sensitive to the operationalization of independent variables (see Appendix). We rerun our models replacing one of the independent variables with an alternative measure. First, we replace distinctiveness with distinctiveness square, assuming a quadratic relationship. Although Taeuscher et al. (2021) show that the effect of distinctiveness on legitimacy is linear for novelty-expecting audiences, several empirical papers on the optimal distinctiveness of narratives portray the relationship as an inverted U-shape (e.g., Haans, 2019). When we employ this different operationalization of distinctiveness, our main results remain similar for both journals (see Appendix), supporting Taeuscher et al. (2021) findings. Second, we replace social contribution claims by using the variable ‘Social’ of the LIWC-22 tool to capture expressed social processes. This variable is suitable to replace social contribution claims because the score indicates the importance of social contributions in the text. While results remain the same for SBEJ, ETP shows a very small positive effect (1.001) for social contribution claims. Third, we replace academic writing with the variable ‘Cognition’. Cognition is suitable to replace academic writing because it more broadly captures “different ways people think or refer to their thinking” (Boyd et al., 2022, p. 17). Replacing academic writing, our results remain largely the same. While all effects for SBEJ are stable, ETP shows a lower and no longer significant effect for academic writing.

4.3 Post hoc analysis of distinctiveness

To understand the different findings regarding the relevance of distinctiveness in both journals, we conduct an additional analysis in three steps. First, we test whether ETP and SBEJ are different by merging both datasets and creating a dummy variable for the journal (1 = ETP, 0 = SBEJ) as well as interaction effects with theoretical contribution claims and distinctiveness (see Tables 8, 9, 10). Second, we compare the distribution of distinctiveness across both journals (see Fig. 5). Third, we plot the predicted value of publication legitimacy across the whole range of distinctiveness for both ETP and SBEJ (see Fig. 6).

Table 8 Marginal effects overview (ETP)

Table 9 Marginal effects overview (SBEJ)

Table 10 Regression Results with publication legitimacy (probability of paper acceptance) as the dependent variable (ETP and SBEJ)

Fig. 5

Distinctiveness by journals

Fig. 6

Average marginal effects distinctiveness

The results in Table 10 suggest differences in the competitive environments of ETP and SBEJ. Notably, the distinctiveness scores between the two journals exhibit minimal overlap. This suggests that submissions to SBEJ, compared to ETP, are more likely to face competition of many papers that score high on distinctiveness. Hence, distinctiveness seems to play a more prominent role in ETP. The marginal effects indicate that for ETP, distinctiveness has a notable impact (i.e., an increase in distinctiveness is associated with an increase in the predicted probability of a paper’s acceptance). In contrast, for SBEJ, the marginal effect of distinctiveness on paper acceptance is not pronounced, indicating that distinctiveness’s effect on publication propensity differs between the two journals.

5 Discussion

5.1 Expectations of entrepreneurship research audiences

Our exploration into the expectations of entrepreneurship research audiences provides valuable insights into the academic publication process. When contextualized within the broader literature, our findings challenge some established norms, reinforce others, and, most importantly, pave the way for a more reflective academic discourse on publishing in entrepreneurship research.

In particular, the differences between ETP’s inclination for theoretical contributions and SBEJ’s affinity for empirical work, as supported by the extant literature (Carlile & Christiensen, 2005; Anderson et al., 2019), underscore the evolving nature of academic preferences. This suggests that journals, in their quest for niche positioning and domain expertise, tend to gravitate towards a particular type of contribution. This distinction offers authors a strategic advantage in tailoring their manuscripts to align more closely with a journal’s predisposition.

However, the divergence in the preference for theoretical versus empirical work between journals such as ETP and SBEJ raises questions about the broader entrepreneurship journal landscape. Does this division promote a more comprehensive understanding of the field, or does it risk creating silos where theoretical and empirical work seldom intersect? While specialization allows for in-depth exploration, it is equally crucial for disciplines to maintain a balance and encourage interdisciplinary engagement (Eden, 2002; Shepherd & Wiklund, 2020; Tranfield et al., 2003). The interplay between theory and empiricism is essential for the robust evolution of any academic field, as both offer complementary insights (Anderson et al., 2019; Haans et al., 2016; Hambrick, 2007).

Another interesting dimension that emerges from our study is the role of distinctiveness in academic publishing. The prevailing discourse around distinctiveness in academic work (von Krogh et al., 2012; Landström & Harirchi, 2019; Salvato & Aldrich, 2012) finds resonance with our observations regarding ETP’s preference for standout submissions. However, our findings go a step further, suggesting that distinctiveness is not beneficial in all journal environments. In journal environments such as SBEJ that receive a broad range of submissions, distinctiveness does not seem to be a promising publication strategy. In such contexts, what may be perceived as unique in one domain could be commonplace in another. Furthermore, with a multitude of voices, perspectives, and methodologies vying for attention, the threshold for what constitutes ‘distinctive’ becomes higher.

5.2 Optimal distinctiveness and expectation variety

The burgeoning interest in optimal distinctiveness within narrative research (Lounsbury & Glynn, 2001, 2019; Navis & Glynn, 2011) emphasizes the importance of a balance between crafting a distinctive narrative and meeting audience expectations for achieving narrative legitimacy. Our study contributes to this literature stream by suggesting that legitimacy is contingent on fulfilling a particular set of expectations. While previous research highlights that narratives achieve legitimacy by meeting a diverse array of expectations (Zimmerman & Zeitz, 2002), thus fitting within an audience's acceptable boundaries (Deephouse, 1999), our study suggests that catering to a variety of expectations may inadvertently blur the focus of narratives and potentially incite interpretive inconsistencies among audiences. Our evidence implies that centering on particular expectations may be more advantageous than meeting various broader expectations of the audience (Zhao et al., 2017).

Building on this finding, we explore three primary mechanisms related to the unfolding of optimal distinctiveness when a large variety of audience expectations exist. First, while the optimal distinctiveness literature generally emphasizes the importance of distinctiveness for legitimacy (Navis & Glynn, 2011; Taeuscher et al., 2021), Haans (2019) suggests that this role is subject to the competitive environment of the narratives. To broaden this dialogue, we provide insights that challenge the assumption that distinctiveness is invariably necessary for legitimacy, illustrating that its importance is conditional (Haans, 2019). Specifically, in our analysis of academic narratives, we observed stark contrasts in the reception of distinctive submissions. While distinctive papers in ETP often carved a niche for themselves amidst more traditional studies, similar submissions to SBEJ found themselves vying for attention in a sea of standout works, each striving for uniqueness in various forms. Furthermore, in contrast to Haans (2019), our findings do not reveal an (inverted) U-shaped pattern. Instead, aligning with Taeuscher et al. (2021), we assert that distinctiveness positively influences legitimacy when the audience, such as editors and reviewers, is open to novelty. However, this advantage diminishes in situations where most competing narratives are already highly distinctive (Haans, 2019).

The second mechanism emphasizes that while contributions can significantly enhance a paper's appeal, it may not be beneficial to increase the number of contributions. While recognizing the relevance of various types of contributions, audiences might only focus on one major expectation that they have. Thus, it becomes paramount for authors to discern which contributions resonate most profoundly with their target audience (Fisher et al., 2017). Overloading a paper with numerous contributions could undermine its main message, leaving readers overwhelmed or unclear about its primary significance. Given the exploratory nature of this finding, we encourage future research to delve deeper into this phenomenon and test the impact of multiple contributions on a paper's reception and legitimacy within the academic community.

Third, the present paper uncovers specific conditions affecting the balance between audience expectations and narrative distinctiveness. We argue that language use is important in garnering audience acceptance (e.g., Clarke & Cornelissen, 2011; Moss et al., 2018). The nuances in linguistic choices, tone, and framing can significantly influence how a narrative is perceived. Hence, the nuances in language can either enhance or detract from the paper's distinctiveness, shaping its perceived value and relevance. With these insights in mind, the literature on optimal distinctiveness (e.g., Barlow et al., 2019; Taeuscher & Rothe, 2021; Zhao et al., 2017) may need to consider language elements alongside content-based audience expectations and distinctiveness.

5.3 Limitations and implications for future research

While our study brings a novel perspective to the literature on optimal distinctiveness of narratives, it also opens multiple pathways for future research.

First, editorial and reviewer bias could affect our research. Recognizing that editorial perspectives and expertise could inadvertently favor certain elements such as empirical contributions or theoretical robustness, we see a valuable opportunity for future research. Subsequent studies could expand our knowledge base by collecting and analyzing data on the professional backgrounds, fields of expertise, and experience levels of editors and reviewers. Such an endeavor could provide insights into the decision-making processes and selection criteria behind academic publications. Furthermore, we encourage subsequent research to consider the influence of changes in editorship and the diversity of editorial boards on the content and themes of journals. An in-depth analysis of the strategies used by editors-in-chief for promoting their journals could provide important insights into their preferred academic narratives.

Second, while ETP and SBEJ are representative in certain aspects of the entrepreneurship research landscape, these journals may not capture the full breadth and diversity of audience expectations and preferences across the entire field. Relying on two journals means that our insights are inevitably influenced by the editorial directions, reviewer preferences, and historical trajectories of these particular audiences. Other journals may have different orientations, priorities, and biases, which could lead to divergent findings. Furthermore, the choice of only two journals may not account for variations within subdomains of entrepreneurship research or between more established journals and newer entrants in the field. Consequently, our conclusions, while suggestive, should be interpreted with caution and may not be generalizable across all entrepreneurship research journals. Future research could benefit from expanding the sample size to include a broader spectrum of journals, both in terms of reputation and thematic focus. This would improve our understanding of the interplay between narrative distinctiveness and audience expectations in academic publishing.

Third, our research primarily focused on academic papers, which inherently possess their own set of norms, expectations, and structures. This context may differ considerably from other narrative forms. For instance, when examining the literature on optimal distinctiveness narratives, many studies predominantly address the distinctiveness of entrepreneurial narratives. These narratives, typically crafted for business pitches, investor relations, or marketing purposes, have different objectives and are subject to different pressures and expectations compared to academic writings. Furthermore, the mechanisms that drive acceptance or rejection in academic journals may be distinct from those in the entrepreneurial domain. While academic papers emphasize rigor, clarity, and contribution to existing knowledge, entrepreneurial narratives might prioritize persuasion, vision, and feasibility. Such differences could manifest in the way distinctiveness is perceived and valued. Hence, while our findings provide valuable insights into the academic setting, they may not be directly transferable to other contexts. Future research should consider exploring these mechanisms in diverse settings to determine the universality or specificity of our findings.

Fourth, our study is exploratory, revealing potential patterns and interconnections rather than drawing concrete and definite conclusions. While it broadens our understanding of the interplay between the optimal distinctiveness of narratives and expectation variety, it also emphasizes the importance of using more holistic and detailed approaches. We therefore suggest a neo-configurational approach such as qualitative comparative analysis (QCA) to unlock possibly complex insights into the optimal distinctiveness of narratives. QCA, a method that combines the strengths of qualitative and quantitative research, creates new possibilities for examining the intricate interplay of diverse factors affecting publication legitimacy. We propose utilizing QCA to explore how differing configurations of factors such as theoretical contribution and academic writing impact pubication legitimacy. Given QCA's foundations in set theory, it allows for an exploration of how combinations of elements intersect and collectively influence an outcome. By illuminating these configurations, future researchers can uncover a broader range of successful combinations for publication. This method can also pinpoint potential synergies between factors that boost the distinctiveness and allure of a paper. Furthermore, applying QCA could help reveal how these configuration variances across different journals and fields shape publication legitimacy.

Appendices

Appendix A: Additional insights on measurements

1.1 Theoretical contribution claims

11

Table 11 Dictionary on theoretical contribution claims

1.2 Additional information on topic modelling

Our text corpus for identifying prevalent topics in the entrepreneurship literature to calculate distinctiveness of academic narratives encompasses 7,295 paper abstracts. The abstracts in our sample contain at least 50 words and 160.1 words on average. In line with research standards, we prepared the text in five steps. First, we changed the text to lower case. Second, we removed all nonalphabetic signs, such as punctuations or hyperlinks. Third, we deleted all English stop words in the text. Namely, we deleted all words that are included in the stopwords dictionary of the R text mining package tm. A full list of the words can be shown through the command: stopwords("en") in R. Fourth, we trim whitespace from text. Fifth, we removed words that occurred less than 5 and more than 75 times in the corpus. The resulting corpus was the base for three subsets - topics across journals (ETP and SBEJ), topics within ETP, topics within SBEJ.

In the next step, we had to identify a) the number of topics, b) the sampling algorithm, and c) the number of iterations. In all three cases, we followed Haans (2019); Kaplan and Vakili (2015), and Taeuscher et al. (2021). Therefore, we selected 100 topics. “100 topics can satisfy the demand for sufficient variance while ensuring that meaningful human interpretation is still possible” (Taeuscher et al., 2021, p. 169). Additionally, we used the Gibbs algorithm and ran 400 iterations.

Last, to validate our topic models, we followed DiMaggio (2015), who identifies a topic model as valid if it distributes identical words into subjects that have different meanings. For instance, the word “solution” occurs in topics 36, 42, 71, and 96 across both journals. However, while topic 36 centers around technological solutions, topic 42 focuses on political solutions, topic 71 on financial solutions, and topic 96 on lean startup solutions. As a result, we conclude that the proposed topic model has adequate validity (DiMaggio, 2015).

1.3 Data examples

Legend: theoretical contribution claims, empirical contribution claims, social contribution claims.

1.3.1 Entrepreneurship theory & practice examples

“Crowdfunded microlending research implies that both communal and agentic characteristics are valued. These characteristics, however, are often viewed as being at odds with one another due to their association with gender stereotypes. Drawing upon expectancy violation theory and research on gender stereotypes, we theorize that gender-counterstereotypical facial expressions of emotion provide a means for entrepreneurs to project “missing” agentic or communal characteristics. Leveraging computer-aided facial expression analysis to analyze entrepreneur photographs from 43,210 microloan appeals, we show that women benefit from stereotypically masculine facial expressions of anger and disgust, whereas men benefit from stereotypically feminine facial expressions of sadness and happiness.” (Davis et al., 2021).

Distinctiveness Score: 1.15

Analytical Thinking Score: 83.03

“This editorial draws attention to time to advance entrepreneurship research by focusing on two aspects of timetime perspective and time management. We initiate a deeper conversation on time in entrepreneurship and illustrate the value of a time-based lens for entrepreneurship research through discussing examples at the individual, firm and context levels. These examples consider underdog and portfolio entrepreneurs;  well-being; social and unethical entrepreneurial behavior; entrepreneurial teams and entrepreneurinvestor dyads; firm strategy; industry and cultural contexts. We review promising methods for time-conscious entrepreneurship research: process, true longitudinal, diary, experience sampling, observational, work-shadowing and time-use studies; historical approaches; experiments; and simulations.” (Lévesque & Stephan, 2020, p. 163).

Distinctiveness Score: 0.84

Analytical Thinking Score: 93.12

1.3.2 Small business economics examples

“This paper contributes to explain the persistence of differences in levels of entrepreneurship within and across countries. We provide an explanation based on the dynamic interplay between purposeful intergenerational transmission of preferences for entrepreneurship and public Administration efficiency. Individuals vote on taxes and the collected taxes fund the civil servants' wages. The performance of the administration generating an efficient normative and regulatory environment, affects the success of entrepreneurship. We show that an economy can reach two different long-run equilibria: a traditional equilibrium, with a low proportion of entrepreneurs, high taxes and an inefficient Administration and, an entrepreneurial equilibrium with a high proportion of entrepreneurs and, lower taxes but enough to implement an efficient Administration. The equilibrium achieved depends on the tax policy followed by the different generations. If decisions are made by majority voting in a myopic way, then the initial conditions of the society become crucial. This result explains persistence: an economy evolves around similar levels of entrepreneurship unless some reforms are implemented.” (Olcina et al., 2020).

Distinctiveness Score: 1.84

Analytical Thinking Score: 93.9

“Initial coin offerings (ICOs) are a rapidly growing phenomenon wherein entrepreneurial ventures raise funds for the development of blockchain-based businesses. Although they have recently sprouted up all over the world, raising millions of dollars for early-stage firms, few empirical studies are available to help understand the emergence of ICOs across countries. Based on the population of 915 ICOs issued in 187 countries between January 2017 and March 2018, our study reveals that ICOs take place more frequently in countries with developed financial systems, public equity markets, and advanced digital technologies. The availability of investment-based crowdfunding platforms is also positively associated with the emergence of ICOs, while debt and private equity markets do not provide similar effects. Countries with ICO-friendly regulations have more ICOs, whereas tax regimes are not clearly related to ICOs.” (Huang et al., 2020, p. 77).

Distinctiveness Score: 1.79

Analytical Thinking Score: 92.32

Figures 78

Fig. 7

Data plots (ETP)

Fig. 8

Data plots (SBEJ)

Appendix B: Additional insights on regressions

Tables 12, 13

Table 12 Regression results—alternative measures

Table 13 Variance inflation factors