**4. Discussion**

The findings from our analysis show differing impacts of SSM, depending on field of study. The distribution of articles related to SSM varied considerably since 1980, with certain periods most highly representing its impact. Some authors had more outsized impact on the use of SSM and systems thinking than others. The following sections explore these outcomes with accompanying data visualizations.

#### *4.1. Impact on Academic Discourses Represented by Publication Trends and Impact Factors*

The first major outcome is the publication trend tied to soft systems methodology from 1980 to 2018 as we approach 40 years of SSM discussion and use. Since this research approach was meant to take a global view of SSM's impact on academic output, we did not discriminate by region, instead choosing to view academic impact as an aggregate, worldwide outcome. Figure 1 presents the number of publications that included some reference to SSM (1980–2018).

**Figure 1.** Number of publications by year (1980–2018).

In this figure, we see small growth in the 1980s, with a robust spike in use during the 1990s, following the publication of Checkland's [20] highly cited (569) piece in the *Human Systems Management* journal as well as Mingers and Taylor's [21] "The use of soft systems methodology in practice". Each piece simplified the process and provided examples that practitioners and theorists could apply. During that time more than 90 publications discussed or applied SSM in practice, showing high interest in the methods that continued through the 2000–2009 period. The period from 1990 to around 2010 included the largest number of SSM-related publications. With the decline in academic publications that employ SSM over the last three years, it is possible that the method may be perceived

as inappropriate to answering today's research questions, or that they require more rigor than time allows. It may also be that some publications are not ye<sup>t</sup> stored in databases accessible to the authors, so are unrepresented in this dataset. However, for those authors that have invested in the method and find it of value, this decline may be of concern if it is an indication that SSM's value is no longer clear to academic researchers. From an academic impact perspective, the outlets where SSM research and theory are published may also be of concern as noted in the next section.

#### *4.2. SSM Most Common and Most Impactful Publication Outlets*

While examining the impact of SSM as a function of the number of publication outlets is valuable to show how widely distributed the method is in academia, it is also important to understand where these pieces have been published to ge<sup>t</sup> a better sense of how accessible they are. Further, it is important to know the perceived impact of those journals that have been evaluated using objective measures such as SJR and h-index scores to provide a better sense of SSM's academic impact in the social sciences more broadly. The publication outlets that are most highly represented are included in Figure 2.

**Figure 2.** Highest frequency SSM publication outlets.

While the figure does not incorporate citation counts, the largest number of publications related to SSM are book chapters. This is problematic in academia, at least in the USA, because book chapters, like books, are less accessible to researchers that may be impacted by them. Finding and using book chapters as sources requires both knowledge that the edited book exists and a financial investment that many academics today may be unable or unwilling to make. This situation negatively impacts how broadly disseminated the research or theory outcomes are in academia and, in turn, limits the reach of SSM. Five of the top six journals that published SSM pieces are located in Europe, likely increasing knowledge of them and dissemination there. However, given the limitations of access to library databases in the USA due to increasing cost at higher education institutions, it is possible that researchers miss important SSM pieces, since the work is often behind a publisher paywall [22].

The top nine most represented journals each had published at least two SSM pieces, though only eight had citations because two of the pieces in Service Science were recently published. The following Figure 3 shows the citations for the top eight journals.

Systems Research and Behavioral Science was both a top destination for SSM pieces and had the strongest impact on the field based on citation counts. Most pieces in that journal were not research-based according to our analysis; rather, they discussed the development of SSM as an approach to research in a particular discipline. Many articles offered significant adaptations and additions to the methodology to make it easier to use or more applicable in different fields, but failed to report research outcomes from testing those changes. This is an issue noted by Holwell [16] with articles through the 1990s that continues to today. While Systemic Practice and Action Research included significant publications, the journal's impact was less evident on the field by citation count, though the number of research studies using SSM was greater than most other journals. However, the most significant publication outlets were not journals, as shown in Table 1.


**Table 1.** Citation counts for SSM-related publication outlets.

As mentioned earlier, limited knowledge about and access to books and book chapters reduces dissemination of ideas and research due to cost or marketing of materials, especially from one continent to another. This may negatively impact the reach of SSM in some regions, such as the USA where systems thinking topics tend to be associated with Senge & Sterman [23], Banathy and Jenlink [14], Reigeluth [24], or other authors that more commonly publish conceptual or "thought pieces" rather than research outcomes. With the heavy focus on positivist, numbers-driven research methods in the USA, it is possible that American academics' exposure to systems thinking from these sources has provided a limited picture of SSM and related methods. They may therefore view such approaches as less rigorous and, therefore, less valuable. This situation may account for the difference in where SSM pieces have been published as well, with European journals significantly more represented than in USA journals. This condition could indicate either an implicit or explicit bias among journal editors and researchers against soft research methods, and qualitative approaches more broadly, in different fields that must be overcome with better teaching and training [25].

For SSM as a topic of discourse and use as a research method, Figure 4 visualizes the high impact of four books and thirty-five book chapters had versus journals. 11,144 total book citations came from Checkland's and Scholes' 1990 book [26], which gave that publication the largest impact on other authors. Removing that text as an outlier, SSM-related books still had almost twice the citation impact of book chapters and nearly a treble impact over the top journal's pieces, limiting the impact of academic research pieces that employed SSM by comparison.

**Figure 4.** Citation impact of SSM dissemination outlets as a percentage of all citation.

Based on citation counts, books had the largest percentage impact by a significant margin, with book chapters following substantially lower. Chapters and books combined accounted for 80.35% of all measured SSM citation impact, the top journals accounting for only 19.08%. This means all other journals accounted for only 0.57% of all citation impact, which limits perception that the SSM work contained in them reached an interested audience, which may be partly responsible for the decline of published academic pieces that employ SSM in recent years. This differs by field, with some disciplines such as engineering and business showing historically stronger affinity with SSM methods than others. Findings regarding the disciplines impacted by publications containing information about or research using SSM are included in the next section.

#### *4.3. Disciplines Most Impacted by SSM*

To capture which disciplines are most impacted by SSM, each publication was reviewed and coded according to the Scimago journal subject area that most closely aligned with the content. While white papers, dissertations, books, and chapters do not have subject areas, those were coded in accordance with similarity to journal articles containing the same subject matter. Figure 5 presents the distribution of articles according to coding for Scimago subject area.

**Figure 5.** Top 10 SJR coded subjects of publications using SSM.

"Systems thinking and systems theory" constituted the largest subject, with "Management science and operations management" trailing considerably. This was, in part, because our content analysis revealed that most pieces coded as systems thinking and systems theory discussed the development of SSM from the perspective of a discipline, but often contained no research application. Most also only described proposed alterations of the methodology for a particular purpose [27] (e.g., software development by adding UML) or a description of SSM to a new audience [28,29] (e.g., marketing, medical settings, etc.) and failed to provide evidence of their effectiveness. The Scimago codes above were classified into broad subject categories based on the topics of the articles in the database and are presented in Figure 6.

**Figure 6.** Disciplinary category representation of SSM-related articles.

Business was the largest discipline impacted by SSM research in publications from 1980–2018 with ten subjects represented. Engineering followed with eight highly coded Scimago subjects. Physical science category pieces tended to focus on large scale, messy problems like water allocation in countries with poor access to clean drinking water, making SSM an appropriate tool for research. Education and health sciences also had ill-defined problems that made SSM useful for studying complex systems and, while SSM was less impactful than with business or engineering, there was some influence. The authors that contributed to these pieces had differential impacts on SSM research and practice, as shown in the following section.

#### *4.4. Major Contributors to SSM Theory and Research*

As measured by the number of publications tied to SSM, twelve had the highest impact. These are presented in Figure 7, showing the percentage of the 286 SSM-related publications analyzed here that they are responsible for as an author since 1980. Each had three or more publications related to SSM.

**Figure 7.** Top author representation among all SSM-related publications.

The top six authors combine to represent 23.78%, or nearly a quarter of all SSM-related publications since 1981. The remaining 75% of pieces written by other authors indicates a broad distribution of the ideas and application of SSM, with those above serving as what Lave and Wenger [30] called core participants in what may be considered Community of Practice centered on the development and use of soft systems methodology for academic research and theory development. While the distribution of the work is fairly broad, the following Figure 8 shows that the citation impact is substantially different.

**Figure 8.** Author impact on field by citation representation percentage.

Checkland's impact, as expected, is clearly massive, regardless of the field of influence. His work on SSM garnered 23,780 citations, with 11,114 alone for his 1981 book. Mingers contributed substantial work as well. Boardman, with his variant of SSM applied in engineering and business settings, and his work with Sauser, showed strong impacts among the remaining authors. However, with less than 23% of all citations coming from authors besides Checkland, the academic impact distribution of SSM as a research approach did not have the reach that its proponents may have hoped.
