2.1. Use of Theory in LIS

The definition of theory varies depending on the academic field and researcher. Buckland (1991, p. 18) argues for a broad definition of theory, stating that a theory is “in the broad sense of a description or extension of the nature of things, not in the more restricted sense, used in some sciences, of denoting fundamental laws formally stated and falsifiable.” To understand theory, it is necessary to compare and examine the definitions of similarly used analytical or interpretive tools: models. As one of the definitions of a model in the field of literature and informatics, Wilson (1999, p. 250) defined a model as a “framework for thinking about a problem… [that] may even contribute to a statement of the relationship.” When comparing Buckland’s definition of a theory with Wilson’s definition of a model, one can see that the two are closely related together, but difficult to differentiate. Many studies use theory and model interchangeably. However, Frankfort-Nachmias and Nachmias (1996) distinguish a theory from a model by the fact that a theory provides an explanation of the phenomenon, while the model describes the phenomenon. This emphasizes that the description of the phenomenon alone cannot show academic growth, and the maturity of the study can be measured through theories that can explain the phenomenon. Therefore, this study aims to see the degree of maturity by examining how theory is used in research.

Several studies have been conducted on how theories are used in the field of library and information science. Peritz (1980) reported that only 14% of sample articles from 1950 to 1975 could be theoretical research. Similar results have been reported in the 1980s, with Nour (1985) identifying theory in 21.2% of the literature published in 1980; Feehan et al. (1987) identifying 13% in articles published in 1984; and Järvelin and Vakkari (1990) finding theory in 10% of journal articles published in 1985. In addition, the ratio used for information behavior research was very low at 6% to 8% (Järvelin & Vakkari, 1990). They speculated the low usage of theory in information behavior research was due to the professional or practitioner origins of the field. In contrast, Pettigrew and McKechnie’s 2001 study found that 34.1% of studies applied theory. Pettigrew and McKechnie (2001) argue that while there were differences in which journals were sampled and when in previous studies, the use of theory, whether from LIS or elsewhere, in LIS research appeared to be increasing at the time. Kim (2004) also disagreed that theoretical research was insufficient in LIS, finding that 41.4% of the studies deal with theoretical development and utilization. However, Vakkari (2008) found a decreasing trend of theory use in papers within the Information Seeking in Context (ISIC) conference proceedings. Julien et al. (2011) found that the use of theory in research articles published by both scholars and practitioners is trending down, with 22.7% of articles using theory. These studies suggest that while theory use has increased over time, it may have reached a plateau and begun to trend down more recently. However, more current research is needed to investigate this trend. Variations in theory use may also be regional. In Korean LIS research journals, Jeong and Kim (2005) found only 10% of studies apply a theory. Wu et al. (2017) found 49.9% of articles published in Taiwanese LIS journals between 2010 and 2015 use theory.

Lund (2019) analyzed how often an information theory was cited using Google Scholar and found that Kuhlthau’s (1991) ISP was the most frequently cited theory to date. Lund (2019) also forecasted ISP to be the most frequently cited theory in the next 25 years. In a follow-up study, Lund (2021) analyzed dissertations studying information behavior and found that the ISP model was the top theory in information behavior studies, with TIW ranked in eighth place, being identified in 8 dissertations.

Theory can be used in different ways and for different purposes in research. Jeong and Kim (2005) suggested the degree of theory use in studies can be classified across five levels: spot citing, background review, theory discussion, theory application, and analytic evaluation. The five levels they proposed were used by Wu et al. (2017) to analyze the status of theoretical use in the Taiwanese LIS journals previously mentioned. Kumasi et al. (2013) subdivided the five levels and classified them into seven levels, and the classification is as follows:

Kumasi et al. (2013) developed these categories of theory talk by looking where and how theories are used within all articles from seven LIS journals published between 2009 and 2011. These categories ranged in intensity, with TD and positioning being minimal, theory diversification and conversation being moderate, and TA, testing, and generation being the highest major intensity. As discussed in Methods Section 3, the authors of this study modified the Kumasi et al. (2013) categories of theory talk to focus on a singular theory.

2.2. Theory of Information Worlds

The TIW was introduced in 2008 (Burnett & Jaeger, 2008) and then formalized and more fully presented in a book published in 2010 (Jaeger & Burnett, 2010). TIW is a revision and expansion of Chatman’s theories of small worlds and normative behavior (Burnett et al., 2001; Chatman, 1991, 1992, 1999) influenced by Habermas’ (1992) concepts of the public sphere and lifeworld. Based on these concepts and theories, TIW establishes a scalable framework consisting of five core concepts (Jaeger & Burnett, 2010):

The five core concepts of TIW interact with each other, rather than acting independently. Until now, many studies applying TIW have revealed that by applying the five core concepts of theory to phenomena, individuals follow the norms and value systems of the small world(s) to which they belong. TIW has been criticized by some scholars and some have combined it with other theories or proposed new theories to overcome the limitations of the theory. As an example of the former, Kitzie (2017) combined TIW with Goffman’s (1963) stigma theory, McKenzie’s (2003) information practice, Chatman’s (1996) information poverty, and de Certeau’s (1984) concept of spatial practice to analyze information behaviors within the LGBTQ+ community. In addition, Post (2020) used Becker (1982)’s concept of art worlds in combination with TIW in designing theories and methodologies for analyzing the information world of artists and curators. In an example of the latter, Yu (2012) critiqued TIW for its sole focus on social contexts of information behavior and a perceived dismissal of individual factors such as emotions and agency and instead proposes a concept for an individual’s information world based on Popper’s (1972) time, space, and world perspectives. Yu’s (2012) critique and insights from his research with hermeneutics and textuality led Burnett (2015) to build upon information worlds with an expanded theory of information domains to provide a more holistic analytical framework through which to examine, describe, and interpret information behavior. Through the framework of information domains, Burnett (2015) repositions the existing TIW as the social domain, and then adds the domain of the individual to interpret individual factors that may impact information behavior and the domain of signification to understand and interpret how signs, symbols, languages, and other tools may influence or add context to information behaviors. Despite the introduction of information domains, TIW still functions as an independent theory.

4.1. Theory Citation & Use Levels

From the third screening and initial coding of manuscripts, 68 peer-reviewed articles were identified as relevant to this study. Of these, 35 (51.47%) use TIW for TD; 12 (17.65%) for TP; 3 (4.41%) for TC; 18 (26.47%) for TA; and 11 (16.18%) for TG; again, noting that TG is not mutually exclusive with other codes. As such, a subsample of 26 peer-reviewed research articles uses TIW at the higher levels of TA, TC, and TG. In the findings, figures, and tables herein, the overall sample of all codes is designed as All Codes (n=68), the subsample of higher-level uses of theory is designated as TA, TC, or TG (n=26), and a subsample of only articles that applied the theory in empirical studies is designated as TA (n=18). As grouped, research articles with higher levels of theory use (TA, TC, or TG) account for 38.24% of the overall sample, and articles specifically applying the theory for empirical research account for 26.47% of the overall sample.

As depicted in Fig. 3, both overall citation (n=68) and use as well as higher usage levels (TA, TC, or TG; n=26) of the TIW are trending upward. Overall citation and use of the theory have a trendline with a linear R-squared value of 0.525 and higher-level usage has a linear trendline R-squared value of 0.468. TA only has the lowest upward trendline with an R-squared value of 0.285.

Fig. 3

Theory of information worlds citation and use count in articles, 2008 to March 2022. TA, theory application; TC, theory conversation; TG, theory generation.

Additionally, as depicted in Fig. 4, the number of research articles at each theory use level of the single theory talk model is also trending upward. TD has a linear trendline R-squared value of 0.459, TP has an R-squared of 0.269, TC has an R-squared of 0.159, TA has an R-squared of 0.286, and TG has an R-squared of 0.272. The trendlines and corresponding R-squared values⁶ in Figs. 3 and 4 are calculated using the function built into Google Sheets ( https://www.google.com/sheets/about/).

Fig. 4

Frequency of theory use levels per year, 2008 to March 2022.

4.2. Publication Trends

4.3. Research Contexts

Table 7 briefly summarizes the contents of articles coded for TA, TC, or TG (n=26). Due to scope and space limitations, a more in-depth analysis and discussion of the contents of articles coded as TA, TC, or TG will be conducted and shared in a follow-up study. Full citations for the articles referenced in Table 7 can be found in the appendix. These summaries provide context for the findings shared.

5.1. Research Question 1 - Theory Citation & Use

Based on the citation and use trends reported in Findings Section 4.1, it appears that the overall awareness and use of TIW are trending upward, albeit at a low rate. Higher levels of theory use, such as TC, TA, and TG are also trending upward, but at even lower rates. However, it is important to note that the goal of the figures and R-squared values presented in Section 4.1 is to identify potential trends and not to make predictions, which would not be possible based on the limitations of this study’s design, data, and sample size. While McKechnie et al. (2001), Pettigrew and McKechnie (2001), and Kim and Jeong (2006) found overall increases in theory use in the 1990s and early 2000s, Vakkari (2008) and Julien et al. (2011) reported downward trends in theory use and development. While the citation and use of TIW may be trending upward across all the theory talk categories, this study did not collect data related to other LIS theories, and as such, may not be indicative of the use of theory in LIS writ large.

The percentages of higher-level theory use (38.24%, TA, TC, or TG) are between the findings of Pettigrew and McKechnie (2001), who found that 34.1% of articles used theory, and Kim and Jung (2006), who found 41.4% of articles in their sample. Of course, those studies did not focus on a single theory, and applied different analyses on different samples at different times. However, if using the smaller subsample of only articles in the TA, which accounted for 26.47% of the overall sample, this rate is lower than the aforementioned studies but higher than historical studies from the 1980s and 1990s. While Kim and Jeong (2006) noted increases in theory use, articles focused on theory development were trending down. Aside from TA, the codes of TC and TG from the single theory talk model are closely associated with theory development, and both TC and TG articles trended upward, albeit slowly. TG articles were identified 11 times (16.18%). However, TC papers were the least common in the sample, with just 3 articles (4.41%) and the lowest R-squared value of all codes. While not focusing on TIW, Budd (2011) argues that, while challenging, continued research and debate about concepts and theories is important for LIS. Based on the findings, the authors would also encourage more theory and concept papers. TG papers were co-coded with 5 TP articles, 5 TA articles, and 1 TC article, suggesting that TG or development of TIW or related theories was or is not limited to empirical research. Again, these findings speak for the sample presented here and are not representative of the broader field.

5.2. Research Question 2 - Publication Trends, Affiliations, & Collaborations

Within the overall sample of any type or level of theory use, Huvila, a Professor at Uppsala University, appears to reference TIW the most. However, Huvila is not included among the other authors in the higher levels of theory use at all. The overwhelming majority of work related to TIW across all samples can be linked to Gary Burnett, one of the theory’s creators, or to his current or former students, whether he was a direct dissertation advisor or committee member. As described in Hollister et al. (2022), several doctoral candidates worked with Dr. Gary Burnett to develop a general codebook for the theory based on three methodologically and topically different dissertations. The general codebook was shared with other students and scholars, both at Florida State University as well as at other institutions. Of course, other students and scholars have used TIW in their own way, without the codebook. However, this finding suggests that effective mentorship and collaboration during doctoral studies can promote the use of theory (Hollister et al., 2022).

Intuitively, the most active institutions and countries using TIW are associated with these same scholars, who are mostly situated at institutions in North America, primarily at Florida State University and the University of Maryland. However, there are some noticeable efforts in Europe and in the East Asia and Pacific/Oceanic regions. As data collection was limited to articles in English and Korean, expanding the scope to other languages and using databases that have broader coverage of international journals and journals published in other languages may have identified more relevant or related works using TIW. As the findings stand, however, greater efforts to collaborate and support scholarship in the Global South and other areas outside of North America, Europe, and East Asia are needed.

While most authors were based in LIS, collaborations spanned 7 other disciplines in the overall sample. Business and journalism/communications were represented in both of the narrower subsamples of higher levels of theory use. Additionally, lead authors from other disciplines were identified in both the TA, TC, or TG subsample, with 1 in business and 1 in journalism/communication, and 1 in journalism/communication in the TA-only subsamples. While author order is not always indicative of an individual author’s level of contribution to a manuscript and author order norms vary across disciplines, these findings may suggest that TIW is making its way into other disciplines, primarily through interdisciplinary collaborations.

Given the authors’ discipline breakdown, it is unsurprising that the top journal venues across all samples are based in LIS. The top publication venues are mostly highly ranked journals in reputable indexes, suggesting that TIW is gaining exposure and use in highly visible journals with high impact factors. By identifying the publication venues as well as authors, affiliations, and disciplines associated with TIW, the authors hope to highlight the potential versatility of the theory, spotlight prominent and emerging scholars who have adopted it, and encourage further collaboration around TIW both within and outside of LIS.

5.3. Research Question 3 - Research Contexts

Aside from TIW being used as its own descriptive keyword in the research articles, the keyword analysis found that TIW is often described or associated with keywords related to information behavior and information-seeking behavior. However, other more specific types of information behaviors were used as keywords, such as information practices, information literacy, information sharing, and information overload.

The majority of articles coded with TA used mixed methods or qualitative methods, and both of the two articles that specified multiple methods approaches combined only qualitative methods. Only one of the studies was specifically focused on quantitative methods. Burnet and Jaeger (2010) note that TIW is intended to be amethodological and thus usable with whichever methods researchers deem best for their research questions. However, given the heavy focus on qualitative methods in the research articles so far, more quantitative studies could be conducted for further theory testing and development. Relatedly, propositions or hypotheses for TIW have yet to be officially published by Burnett and Jaeger (2010). Additional quantitative studies could help in their development.

Qualitative interviews and content analysis appear to be the most used methods for data collection and analysis with TIW in the research articles collected. Survey and field research were also common but to a lesser degree. Cross-referencing these findings with those from Park (2022), which exampled 14 TIW-using dissertations, sees similar trends: 7 used qualitative methods and 7 used mixed methods, with 10 overall using interviews. It should also be noted that 6 of the dissertations analyzed by Park (2022) overlap with 7 of the articles in the TA sample, meaning that 6 of the scholars successfully published a research article using TIW from their dissertations, and one published 2. This finding is also encouraging because publishing articles out of one’s dissertation is highly encouraged and helpful for emerging scholars, especially those on the job market.

The TA research articles also focused on a variety of groups and communities in both online and offline contexts. Eight (44.4%) of the articles focused on marginalized or protected communities, such as immigrants, children or young adults, women, and seniors. This was also observed by Park (2022), who found that 7 or 50% of the dissertations focused on marginalized communities.

Given scope and space limitations, additional implications and applications of the findings from TA articles, as well as TC and TG articles, will be explored through content analysis in a follow-up study. This follow-up study will address how TIW has been used alongside other theories, frameworks, or models and how TIW has been developed or revised based on the findings.

5.4. Limitations

As mentioned, there are several limitations to this study. As stated in the Methods Section 3, the literature review screening and sampling process were limited due to the databases and full-text articles available through their respective institutional libraries, and the respective scope and coverage limitations of said databases and library collections. However, the researchers attempted to counteract this by using an effective search strategy as well as external sources and services like Google Scholar, email, or the academic social network ResearchGate,⁸ to identify and secure additional access to full-text versions. The data collection period also ended in March 2022. Given the trends identified here, it is likely that relevant articles have been published between then and prior to the publication of this article.

This study also focused on the trends associated with a single theory, so the findings may not be representative or similar to general theory citations or use trends in LIS or other specific theories in LIS. Relatedly, while based on an existing framework, this study is the first to use the revised single theory talk model. To ensure its trustworthiness and accuracy, the authors revised and evaluated the codebook through multiple rounds of inter-coder reliability testing.