University of Split

EXECUTIVE SUMMARY AND STRATEGIC VISION

The University of Split demonstrates a commendable overall integrity profile, characterized by a low aggregate risk score of 0.248. This performance is anchored in strong controls in areas such as multiple affiliations and publication channel selection. The institution's thematic strengths are notable, with SCImago Institutions Rankings data placing it among the top national performers in key areas including Engineering, Computer Science, Environmental Science, and Business, Management and Accounting. However, this profile of excellence is contrasted by significant vulnerabilities, most critically in the Rate of Hyper-Authored Output, and medium-level risks related to impact dependency and authorship concentration. These specific risks present a direct challenge to the University's mission to uphold "high standards of excellence, ethics and moral" and to foster "leadership roles." Addressing these integrity vulnerabilities is crucial to ensure that the institution's recognized thematic excellence is built upon a foundation of transparent, ethical, and sustainable research practices, thereby fully realizing its commitment to societal contribution and academic leadership.

ANALYSIS BY INDICATOR

Rate of Multiple Affiliations

With an institutional Z-score of -0.906, significantly lower than the national average of -0.821, the University of Split exhibits an exemplary and very low-risk profile in this area. This result indicates a clear and consistent affiliation policy that aligns with national standards while demonstrating even greater control. The absence of risk signals suggests that the institution's collaborative practices are transparent and well-governed. While multiple affiliations can be a legitimate outcome of researcher mobility or partnerships, the university's data shows no signs of strategic attempts to inflate institutional credit or engage in “affiliation shopping,” reflecting a robust and ethically sound operational framework.

Rate of Retracted Output

The institution's Z-score for retracted output is -0.061, which is statistically normal and closely aligned with the Croatian average of -0.095. This low-risk level is what would be expected for an institution of its context and size, indicating that its pre-publication quality control mechanisms are functioning effectively. Retractions are complex events, and a low, controlled rate can signify responsible supervision and the honest correction of unintentional errors. The data suggests that there is no systemic failure in the institution's integrity culture or methodological rigor that would lead to an abnormal rate of retractions.

Rate of Institutional Self-Citation

The University of Split shows a Z-score of 0.322 in institutional self-citation, a medium-risk value that reflects a systemic pattern also observed at the national level (Z-score: 0.288). This alignment suggests that the institution's practices are in line with a broader trend within the country. A certain level of self-citation is natural, reflecting the continuity of established research lines. However, this medium-level indicator serves as a caution against potential scientific isolation or 'echo chambers' where work is validated without sufficient external scrutiny. It warns of a shared risk of endogamous impact inflation, where academic influence may be shaped more by internal dynamics than by recognition from the global community.

Rate of Output in Discontinued Journals

With a Z-score of -0.319, the institution's rate of publication in discontinued journals is low and statistically normal when compared to the national average of -0.284. This indicates that the university's researchers are, on the whole, exercising appropriate due diligence in selecting reputable dissemination channels for their work. The low-risk profile suggests that there is no significant channeling of scientific production through media that fail to meet international ethical or quality standards, thereby protecting the institution from the reputational risks associated with 'predatory' or low-quality publishing practices.

Rate of Hyper-Authored Output

The institution exhibits a significant risk in this indicator, with a Z-score of 2.738 that sharply accentuates the medium-level vulnerability already present in the national system (Z-score: 0.472). This high value is a critical alert. While extensive author lists are legitimate in 'Big Science' disciplines, this result strongly suggests a systemic tendency towards author list inflation that goes beyond the national norm, diluting individual accountability and transparency. It is imperative for the institution to investigate whether this pattern stems from necessary massive collaboration or from problematic 'honorary' or political authorship practices that compromise research integrity.

Gap between Impact of total output and the impact of output with leadership

The University of Split has a Z-score of 1.673 for its impact gap, a medium-risk value that indicates high exposure to this vulnerability, especially when compared to the national average of 0.807. This wider-than-average positive gap signals a potential sustainability risk, suggesting that the institution's scientific prestige may be more dependent on external partners than that of its national peers. This metric invites a strategic reflection on whether the university's high-impact metrics result from genuine internal capacity and intellectual leadership, or from strategic positioning in collaborations where it does not hold a primary role, a dynamic that could undermine long-term research autonomy.

Rate of Hyperprolific Authors

With a Z-score of 0.906, the institution presents a medium-level risk for hyperprolific authorship, a moderate deviation from the low-risk national standard (Z-score: -0.608). This discrepancy suggests the university is more sensitive to this risk factor than its peers and warrants a review of its causes. While high productivity can reflect leadership, extreme individual publication volumes challenge the limits of meaningful intellectual contribution. This indicator alerts to potential imbalances between quantity and quality, pointing to risks such as coercive authorship or the assignment of authorship without real participation—dynamics that prioritize metrics over the integrity of the scientific record.

Rate of Output in Institutional Journals

The institution's Z-score for output in its own journals is 1.003, a medium-risk value that nonetheless indicates differentiated and more effective management compared to the higher national average of 1.531. While in-house journals can create conflicts of interest by making the institution both judge and party, the university appears to moderate this risk better than its national counterparts. This suggests a more controlled approach that mitigates the potential for academic endogamy and the use of internal channels as 'fast tracks' to inflate publication counts without undergoing rigorous, independent external peer review.

Rate of Redundant Output

The University of Split has a Z-score of -0.125 for redundant output, which, while categorized as low risk, signals an incipient vulnerability as it is slightly higher than the national average of -0.247. This suggests that while the issue is currently contained, the institution's research output shows a slightly greater tendency towards data fragmentation or 'salami slicing' than its peers. This practice, which involves dividing a study into minimal publishable units to inflate productivity, distorts scientific evidence. The signal warrants proactive monitoring to ensure this trend does not escalate and to reinforce the principle of publishing significant, new knowledge over sheer volume.