An understanding of the research process is central to fostering the development of innovative knowledge that can be used to tackle the complex global challenges faced by humanity (Hoolohan, Mclachlan, & Larkin, 2019). In the ongoing race for knowledge advancement and innovation, governments and organizations face pressure to make informed decisions about how best to allocate research funds, improve current policies and incentive frameworks, and promote quality research (Vaesen & Katzav, 2017). Increasing resources are now being invested in research endeavors, leading to more knowledge being produced, even though a managerialism drive has impaired knowledge breakthroughs (Horta & Santos, 2019a). Most of our understanding of knowledge production relies on studies of research productivity and performance, both descriptive and inferential; such studies are informative and useful but are focused mostly on finalized research products such as papers and related bibliometric information complemented by demographic and institutional characteristics (Young, 2015) and additional perspectives are needed.

An understanding of how researchers’ mindsets and beliefs are associated with their strategic research approaches is just as pivotal to comprehending the knowledge creation process as an understanding of how environmental frameworks and other determinants shape research production. However, although there is a substantial body of research on the latter (e.g. Bentley, 2014), minimal attention has been paid to the former. Researchers and scientists have not achieved consensus on researchers’ own conceptions of research (Fuller, 2012). Conceptions of research matter because the cultural and social ethos guiding the researcher’s thinking and beliefs about research—and science in general—influence the motivations, decisions, and aptitudes related to engagement in research choices and interests (Niiniluoto, 2020). This study addresses this research gap and advances recent studies on researchers’ preferences for specific research agendas, the majority of which have focused only on a single field or discipline (Foster, Rzhetsky, & Evans, 2015; Horta & Santos, 2016; Ying, Venkatramanan, & Chiu, 2015). This exploratory study aims to answer the following research question: what are the associations between researchers’ conceptions of research and their preferences for strategic research approaches? The study is based on two inventories: the Conceptions of Research Inventory (CoRI) (Meyer, Shanahan, & Laugksch, 2007) and the Multi-Dimensional Research Agenda Inventory—Revised (MDRAI-R) (Horta & Santos, 2020).

The results of the OLS analysis are summarized in Table 2.

Because of the large sample size, most of the effects were significant, and we therefore focused on the findings with the largest coefficients due to their expected practical significance.

The conception of research as the discovery of truth showed several significant effects, but the most notable was a negative association with a society-driven orientation (B = -0.114, p < 0.01). A possible interpretation is that research focused on societal issues is likely to be more applied in nature and thus not directly related to the idea of research as a “truth-seeking” process, which may be considered more theoretical and abstract. However, this seems to stand in opposition to the focus on truth-seeking from a citizen science perspective, where non-professionals participate in scientific research with a “truth-seeking” purpose and focus on a mix of applied societal and scientific challenges (Wynn, 2017). The dimension misconceptions about research also showed several interesting effects, with high scores on this dimension associated with a high level of mentor influence (B = 0.277, p < 0.01), increased tolerance of low funding (B = 0.078, p < 0.01), and a higher orientation toward an academia-driven (B = 0.165, p < 0.01) and society-driven (B = 0.150, p < 0.01) approach. The link with mentor influence suggests that a researcher following a specific approach may feel more confident with the presence of a mentor, who provides a sort of guarantee that the chosen approach is valid. A researcher’s misconception about research may derive from socialization with the mentor during the doctoral studies. The same rationale holds for the association of misconceptions about research and the academia-driven dimension of the MDRAI-R. A researcher may feel confident about a specific research approach or conceptualization if he or she perceives that this research process is supported, accepted, or promoted institutionally, such as by a scientific community, university, or industrial research laboratory, as these organizations’ backing confer legitimacy on one’s actions and may be perceived by the researcher as holding the ultimate responsibility (Milgram, 1974). The association with low available funding may reflect the researcher’s difficulties in obtaining research funds if the research approach is considered flawed by research funding agencies. Finally, engaging with peers outside academia through collaboration with laypersons may increase a researcher’s confidence that their research process will achieve successful results because the researcher assumes the role of an expert who is in control of the research process (this may be the case even if the conceptualization or methods are flawed but the researcher believes they are sound; Gorman & Gorman, 2017).

The conception of research as problem-solving had positive associations with divergence (B = 0.157, p < 0.01), collaboration (B = 0.127, p < 0.01), and academia-driven (B = 0.113, p < 0.01) and society-driven (B = 0.310, p < 0.01) approaches. These results are consistent with the expectation that problem-solving implies a need to engage in multidisciplinary and collaborative research processes that are aligned with the field and, more specifically, with institutional policies and incentives, and are focused on “real-life” problems that involve the participation of non-experts in the research process (Lyall, 2019). Conceiving of research as re-research shows a number of weaker effects, the most notable of which were on scientific ambition (B = 0.104, p < 0.01), mentor influence (B = 0.087, p < 0.01), and academia-driven (B = 0.085, p < 0.01), and society-driven (B = 0.030, p < 0.1) approaches. The association with scientific ambition can be explained by the need to assess the findings of others and drive them forward, thus reinforcing one’s field position within what Kuhn (2012) calls “normal science,” or attempting to create a disruptive paradigm. Both are known to be pathways to authority within a field and to career success in research. The influence of the mentor and orientation toward academia-driven dimensions may be associated with research agendas that revisit research work, and the orientation toward society-driven work might indicate a need to reassess and update past findings as societal challenges and knowledge evolve at different paces. Viewing research as an insightful process showed notable positive associations with scientific ambition (B = 0.258, p < 0.01), collaboration (B = 0.241, p < 0.01), and discovery (B = 0.312, p < 0.01), with an additional negative association with mentor influence (B = -0.127, p < 0.01). These results stress the continuous struggle to find new perspectives that can lead to knowledge breakthroughs in specific topics and indicate that this struggle demands determination (associated with scientific ambition), collaboration to deal with the complexities of science, and risk-taking attitudes rather than conformity (Santos & Horta, 2019a). Finally, it is relevant to note that all of the dimensions of the CoRI showed statistically significant and positive associations with one MDRAI-R dimension: scientific ambition. This means that no matter which conceptualizations of research one believes in—and some researchers’ beliefs may include an overlap of several, even if one is dominant (Brew, 2001)—striving for scientific recognition and authority in one’s field continues to be a key objective of most researchers. This finding underlines the strongly rooted Foucauldian and Mertonian dynamics in contemporary science systems (Stehr & Grundmann, 2011).

Some interesting findings were made in relation to the control variables. When strategizing on research agendas, scientific ambition, discovery, and tolerance for low funding played a greater role for male researchers, while female researchers paid more attention to environmental factors such as academia-driven and society-driven approaches. These findings align with expectations based on the literature that male and female researchers face different institutional and organizational constraints on their work, which affect their thinking and beliefs about research practices, aims, and productivity, ultimately undermining female researchers from attaining intellectual leadership roles in epistemic communities (Oleksiyenko & Ruan, 2019). Female and male researchers also differ in terms of research preferences and work strategies, as female researchers tend to be more focused on issues of social innovation and knowledge exchange and are more likely to abide by institutional norms and values, whereas male researchers assume a more individualistic approach and are more career-focused and interested in establishing themselves in scientific communities as leading scientists (Ramos et al., 2015). Time since completing the PhD was more relevant than age to the development of research agendas, highlighting the growing importance of professional age in research endeavors compared with life age (Gaughan, 2009). This is further reinforced by the fact that the findings concerning these two variables differed substantially, perhaps because researchers’ careers are increasingly nonlinear and diverse in terms of time of entry into the career (Jaeger et al., 2017). For example, older researchers may be less ambitious in their research agendas (B = -0.020, p < 0.01), but the more professional experience a researcher has, the more ambitious their research agenda will be (B = 0.010, p < 0.01). Those working in the most prestigious universities were more focused on potentially disruptive research processes and more willing to drive forward ambitious research agendas (B = 0.046, p < 0.01), and they placed considerably less importance on academic (B = -0.134, p < 0.01) and societal factors (B = -0.080, p < 0.01) when strategizing on their research agendas. These results suggest that researchers at research universities, perhaps because of their high scientific potential and strong research profiles, are more engaged with their own agendas (and have the power to resist external pressures and pursue them) than researchers working in less research-oriented universities, who may find themselves with few options other than to follow field community and institutional guidelines and to align their behaviors and beliefs accordingly (Ursin et al., 2020). Finally, researchers who are internationally mobile tend to be more ambitious (B = 0.098, p < 0.01) and divergent (B = 0.087, p < 0.01), more likely to pursue potentially disruptive research (B = 0.137, p < 0.01), and more willing to engage in research even with limited funding (B = 0.107, p < 0.01). They are also less influenced by their mentors (B = -0.136, p < 0.01) and less prone to shape their research to be society-driven (B = -0.074, p < 0.01). These results align with findings that internationally mobile researchers are more independent, disruptive, and ambitious but are not as engaged with societal research issues (which may partly be due to language, cultural, and other barriers to collaborating with local laymen or understanding localized societal challenges), and that they are not dependent on mentor influences, which may decrease or be lost due to the mobility process (Huang, Daizen, & Kim, 2019; Kuzhabekova & Lee, 2020).

This study suggests a close association between the conceptions of research that researchers hold and their approach to setting their research agendas. The findings indicate that research conceptions are particularly associated with five dimensions of the research agenda strategy: scientific ambition, collaboration, discovery, academia-driven approach, and society-driven approach. Other dimensions influencing research agendas were also associated with research conceptualizations, but to a lesser extent. Only scientific ambition had positive associations with all conceptions of research, indicating that variety of ways of thinking about the meaning of research all lead to research agendas with an underlying focus on gaining scientific recognition, prestige, and authority in a given scientific field. This suggests that scientific ambition, the positional goods associated with it (e.g. prestige, one of the most sought-after commodities in science and academia; Coate & Howson, 2016), and material rewards (e.g. the “cumulative advantage” hypothesis; Kwiek, 2016) continue to be major drivers for researchers within the social stratification that characterizes the current global scientific social system (Kwiek, 2019). This is the case regardless of whether researchers have a dominant conceptualization of research among an overlap of several or a single definite conceptualization of research (Brew, 2001).

Collaboration had a positive association with almost all of the research conceptualization dimensions (non-statistical significance with research as the discovery of truth). The only negative association was with the concept of misconceptions about research, which can be explained by the fact that it is difficult for a researcher with these misconceptions to collaborate with others, and such a researcher is unlikely to be invited to collaborate on research projects and agendas set by others. There are two main reasons for this. First, misconceptions of research underlines the preconceived ideas that a researcher brings to the research process, which undermine the researcher’s expected neutrality and impartiality concerning a research object. Starting a research project with such preconceived ideas is more akin to a political or opinionated view about a research object than to a more ethically neutral research inquiry. Researcher biases are known to jeopardize the trustworthiness and the credibility of the research process and are therefore not acceptable to most of the research community (Jorgensen, et al., 2016; Joseph & Baldwin, 2000). Researchers’ biases also raise ethical research issues, lead to pseudo-scientific findings, and are sometimes related to politically and economically sponsored research, which has caused substantial damage to the image of science, research, and researchers (Nestle, 2016). Second, two of the items of the misconceptions about research dimension, i.e. “If followed correctly, research procedures will always yield positive results” and “for an activity to be called ‘research,’ it must involve experimentation,” highlight a somewhat positivistic but highly rigid view of research processes that would exclude specific types of quantitative research and all qualitative research. This is not attuned with the evolution of contemporary research towards multidisciplinarity, the use of multiple theoretical and methodological paradigms, and nuanced forms of engaging in research processes while following legitimized values and norms of research conduct by relevant scientific communities (Corry, Porter, & McKenna, 2019).

Similar to collaboration, almost all of the conceptions of research have statistically significant associations with discovery—that is, one’s willingness to research topics that are riskier but have greater impact potential (no statistical significance was found with research viewed as re-search). In the same vein as collaboration, all of the associations are positively associated with discovery except misconceptions about research. For the reasons mentioned above, this is expected because if one begins research with a preconceived idea, the findings are likely to conform to an initial expectation about what the results will be. Furthermore, even if the purpose of the research is to achieve breakthroughs and innovative thinking, the theoretical and methodological rigidity associated with a high score on misconceptions about research will prevent major findings, as such rigidity is not attuned to the conceptual, methodological, and analytical needs of contemporary science and research breakthroughs (Nairn, 2019).

The relationship between the dimensions of the CoRI and academia-driven approach is statistically significant in all dimensions except research as an insightful process. The statistically significant associations are positive except the association with research as the discovery of truth, which is negative. This may be related to the attitudes of researchers who hold this conception of research (e.g. presenting research impersonally, as if the researcher was not part of the research process; Brew, 2001), which makes them less influenced by institutional pressures and directives from the scientific community and the organization where they work. Contrarily, researchers more influenced by other conceptualizations of research may feel more motivated by institutionally driven pressures and rewards (such as engaging in participatory research that universities push as part of their research and service missions) and more supported in their research activities (even if they are based on misconceptions). The different signs of the statistical effects concerning academia (including the non-significance of one conception) may point toward tension between the research agenda being driven by researchers’ own interests and the matter of personal choice (Kuhn, 2012) and the need to adapt, conform, and adjust research agendas in the face of disciplinary, national, and institutional policies and incentives (Horta & Santos, 2020).

Finally, society-driven agendas were positively associated with misconceptions about research, which might reflect the role of the researcher as an expert in a power relation with non-expert laypersons even when the researcher’s expertise is based on potentially biased scientific premises and beliefs. Society-driven agendas were also positively associated with problem-solving, which reflects the interaction with communities enabling researchers to contribute to solving particular social challenges, and with re-research, which reflects the need to constantly revisit societal challenges as they and the knowledge base co-evolve. However, perhaps related to the more applied nature of society-driven research, there were negative associations between society-driven research and both research as the discovery of truth and research as an insightful process. The explanation for these negative associations may involve the positioning of the researcher as one almost absent from awareness and simultaneously involved in a cognitive process focused more on understanding than on providing an applied solution for a specific problem (Brew, 2001). Such research may have relatively low academic interest (not challenging enough for example) despite its practical importance for the community. This is consistent with the “layer conception” proposed by Brew (2001), in which the researcher is interested and concerned with uncovering what lies beneath the reality and through this discovery process eventually provides a better explanation or creates a new paradigm for explaining this reality (Brew et al., 2016). This reflective, abstract, and theoretically driven type of research conception is not consonant with more applied research or with collaborative engagement with laypeople.

This work was supported by doctoral grant PD/BD/113999/2015 from the Fundação para a Ciência e Tecnologia and by the European Social Fund and the Portuguese Ministry of Science and Education. This research was also funded by the Research Grants Council (Hong Kong SAR, China) through a project entitled “Characterizing researchers’ research agenda-setting: an international perspective across fields of knowledge” (project number: 27608516).

Both authors contributed equally to this work.