Evidencing Collaboration

Claim–Evidence Alignment in Academic Library Impact Documents

Authors

Affiliation

Janny Surmieda

School of Library and Information Studies

Dan Anthony Dorado

School of Library and Information Studies

Abstract

Academic libraries increasingly present collaboration as evidence of institutional value, linking partnerships with faculty, research offices, student services, cultural institutions, community organizations, and professional bodies to claims about student success, research support, equity, open scholarship, cultural preservation, and public engagement. Yet public library documents often blur the distinction between collaborative activities, countable outputs, user or institutional outcomes, and longer-term impact. This study examines how academic library public documents construct, evidence, and legitimate claims of collaborative impact. Using qualitative comparative document analysis, it analyzes public annual reports, strategic plans, impact and assessment documents, and professional frameworks at the level of the impact claim rather than the whole document. The article develops a claim–evidence–legitimation framework that codes claims by impact level, evidence type, claim strength, impact domain, partnership domain, beneficiary, evidentiary adequacy, and legitimation mode. The analysis shows how collaborative impact is often framed through student success, access, research support, equity, public value, and institutional alignment, while the evidence used to substantiate these claims varies substantially in strength and fit. The article argues that the central assessment problem is not simply the absence of metrics, but the mismatch between claim strength and evidentiary support. It contributes a framework for distinguishing symbolic, output-based, outcome-oriented, and transformative collaborative impact claims and offers practical guidance for aligning public reporting with the level of evidence required by different kinds of impact claims.

1 Introduction

1.1 Collaborative Impact as a Strategic Problem for Academic Libraries

Academic libraries increasingly occupy an interstitial role in higher education: they are service providers, research infrastructure, learning partners, cultural-memory institutions, open-scholarship advocates, and visible participants in institutional strategy. This expanded role has made collaboration central to how libraries describe their value. Partnerships with faculty, research offices, information technology units, teaching and learning centers, student-services divisions, cultural institutions, community organizations, and consortia are now routinely invoked as mechanisms through which libraries contribute to student success, research visibility, equity, open science, cultural preservation, and public engagement. Yet the more expansive the claimed contribution becomes, the more difficult it is to establish what counts as impact.

The academic library assessment literature has long recognized this problem. Impact evaluation requires more than enumerating activities or reporting service outputs; it requires a defensible account of how library work changes users, institutions, scholarly practices, or publics (Jager, 2017). Urquhart (2018) similarly frames impact assessment as a matter of principles, methods, and value frameworks rather than a simple exercise in reporting performance data. Studies linking library use to student success have demonstrated the value of moving beyond simple usage counts toward outcome-oriented evidence, including retention, degree completion, learning, and academic performance (Haddow & Joseph, 2010; Nurse et al., 2018; Soria & Fransen, 2017; A. Thorpe et al., 2016). At the same time, work on information literacy, research-data management, open science, and scholarly communication shows that many of the most strategically important library contributions are collaborative and distributed across institutional systems rather than contained within library-controlled transactions (ACRL Research Planning and Review Committee, 2020; Marín-Arraiza, 2019; Miller, 2018; Rice, 2019).

This creates a persistent evidentiary tension. Libraries are under pressure to demonstrate institutional and societal value, but many collaborative contributions are difficult to attribute, observe, or quantify. Attendance counts, repository deposits, satisfaction scores, and narratives of engagement may be useful evidence, but they do not automatically demonstrate durable learning, research-culture change, equity, or systemic transformation. The central problem is that documents often compress activities, outputs, outcomes, and long-term impacts into the same strategic vocabulary.

1.2 From Measurement to Representation

This study begins from the premise that public library documents are not transparent mirrors of actual organizational performance. Annual reports, strategic plans, impact reports, assessment webpages, and professional guidelines are institutional artifacts: they communicate priorities, make claims of value, select evidence, and position the library in relation to the expectations of university leaders, funders, accrediting bodies, disciplinary communities, and publics. Following Bowen (2009), document analysis is therefore appropriate not because documents reveal the full reality of collaborative impact, but because they reveal how public documentary claims of collaborative impact are constructed, evidenced, and legitimated.

The distinction matters. A document that states that a library “transformed student learning” may provide evidence of a program, a participation count, a satisfaction survey, a student quotation, a learning assessment, or no evidence at all. These are not equivalent forms of substantiation. Directed qualitative content analysis permits theoretically informed categories, such as outputs, outcomes, and impacts, while allowing new categories to emerge (Hsieh & Shannon, 2005). Framework analysis supports comparison across document genre and partnership domain while preserving transparent cross-case interpretation (Gale et al., 2013).

The present article therefore treats public documentary claims of collaborative impact as the primary empirical object. A collaborative initiative is defined as a documented activity, service, program, infrastructure, or project involving the academic library and at least one identifiable internal or external partner. An impact claim is defined as a statement attributing value, benefit, change, contribution, transformation, or public/institutional significance to such an initiative. This move shifts the analysis away from asking whether collaboration actually “worked” and toward a more defensible question: how do academic libraries construct documentary evidence of collaborative value?

1.3 Impact Evidence, Public Value, and Institutional Legitimacy

The study is positioned at the intersection of three literatures. First, library assessment scholarship has emphasized the need to demonstrate value in ways aligned with institutional missions, user outcomes, and public accountability (Clunie & Parrish, 2018; Oakleaf, 2010; Urquhart, 2018). Second, public value scholarship is useful because academic libraries frequently claim benefits that exceed individual service use, including democratic access to knowledge, open scholarship, community engagement, and cultural stewardship (Brewer, 2013; Moore, 1995). Third, institutional theory helps explain why public documents may adopt broadly shared vocabularies of innovation, transformation, equity, collaboration, and measurable value. This article uses institutional theory specifically as a theory of strategic legitimacy management: documents are read as attempts to align library work with valued field-level expectations and institutional priorities (DiMaggio & Powell, 1983; Scott, 2014; Suchman, 1995). Academic library documents are therefore not neutral containers of facts; they are strategic texts through which libraries make themselves intelligible and valuable within higher education.

Figure 1 summarizes the claim-evidence-legitimation framework guiding the manuscript. Collaborative initiatives are first bounded as library work involving at least one identifiable partner. Those initiatives generate public documentary impact claims, which are coded by impact level, evidence type, claim strength, impact domain, partnership domain, and beneficiary. Evidentiary adequacy is then assessed as the degree of fit among claim strength, impact level, and evidence type. The final interpretive layer identifies the legitimation modes through which documents make claims credible.

Figure 1: Claim-evidence-legitimation framework for collaborative impact in academic library documents.

1.4 Research Gap and Contribution

Existing research has advanced the assessment of academic library value, particularly in relation to learning outcomes, student success, library use data, assessment practice, and research-support services (Blummer & Kenton, 2018; Oakleaf, 2010, 2011; Rice, 2019; Salisbury & Peasley, 2018; Soria & Fransen, 2017; A. Thorpe et al., 2016). However, less is known about how academic libraries publicly assemble and substantiate collaborative impact claims across document genres. The gap is not simply the absence of indicators. Rather, it is the lack of comparative evidence about how public documents distinguish, blur, or hierarchize collaborative activities, outputs, outcomes, long-term impacts, and legitimation claims when collaboration is presented as a source of institutional and societal value.

This article addresses that gap through a qualitative comparative document analysis of academic library documents and professional frameworks. The study is not designed to estimate the prevalence of claim-evidence practices across the academic library sector; rather, it uses a purposive, genre-varied corpus to identify and theorize recurring patterns in how collaborative impact claims are publicly constructed and substantiated. It contributes to academic library assessment by shifting attention from indicators alone to the alignment between public impact claims and the evidence used to substantiate them. It demonstrates how claim-level document analysis can reveal differences among symbolic, output-based, outcome-oriented, and transformative claims, and how institutional legitimacy vocabularies shape the evidentiary burden of collaborative impact reporting.

Table 1 presents the revised research questions that organize the study. The sequence is deliberate. RQ1 establishes how collaborative impact is explicitly defined or implicitly framed before evidence is assessed. RQ2 examines claim-evidence alignment across activity, output, outcome, and long-term impact levels. RQ3 links domains to evidentiary form rather than treating domain visibility as a simple frequency exercise. RQ4 examines document genre within the analyzed corpus rather than making broad sector-level genre claims. RQ5 links evidentiary mismatch to institutional legitimation.

Table 1: Research questions and analytic purpose.

RQ	Question	Purpose
RQ1	How do academic library public documents explicitly define or implicitly frame collaborative impact?	Establish the construct before judging evidence.
RQ2	What types of evidence are used to support claims of collaborative impact, and how do these evidence types align with activity, output, outcome, and long-term impact levels?	Operationalize claim-evidence alignment.
RQ3	Which domains of collaborative impact are most visible across the corpus, and how do domains differ in the evidence used to substantiate them?	Link domain visibility to evidentiary form.
RQ4	How are claim strength and evidentiary adequacy shaped by document genre within the analyzed corpus?	Make genre the main comparative axis while limiting generalization.
RQ5	Where do documents make claims whose rhetorical strength exceeds the evidence provided, and what forms of institutional legitimation are associated with these mismatches?	Connect evidentiary mismatch to institutional theory.

The argument is intentionally modest in causal terms. The study does not claim to measure the real-world effectiveness of collaborative library initiatives. Instead, it examines the documentary evidence regimes through which academic libraries define, evidence, and legitimate the impact of collaboration.

2 Review of Related Literature

2.1 Library Impact Assessment and the Problem of Evidentiary Scale

Academic library assessment has shifted from a narrow concern with activity counts toward a broader concern with value, outcomes, and institutional contribution. This shift is visible in the foundational value literature and in work on impact evaluation, which distinguishes between the availability or use of services and the changes those services may produce for users, institutions, or publics (Jager, 2017; Oakleaf, 2010; Urquhart, 2018). Głowacka (2019) similarly treats impact as a multidimensional assessment problem involving areas, methods, and indicators rather than a single metric. The literature therefore gives this study its first analytical premise: the evidence needed to support an impact claim depends on the scale and strength of the claim.

The student-success literature illustrates this evidentiary shift. Studies have connected library use with retention, persistence, GPA, degree completion, engagement, and academic achievement, often by linking library data with institutional records (Haddow & Joseph, 2010; Nurse et al., 2018; Soria & Fransen, 2017; A. Thorpe et al., 2016). These studies move beyond service use toward educational outcomes, but they also show that outcome evidence requires integrated data, comparison, and attention to attribution. When public documents present attendance, downloads, consultations, or satisfaction as evidence of transformation, they may shift from output evidence to outcome language without doing the evidentiary work that outcome claims require.

This distinction is especially important for collaborative initiatives. Collaboration multiplies the value potential of library work, but it also complicates attribution. A library embedded in a first-year experience program, open science initiative, doctoral training pathway, community archive, or research-impact service may contribute meaningfully to outcomes that are co-produced with other units. The problem is not that library impact is absent. The problem is that impact is distributed, mediated, and often only partially visible in documents.

2.2 Collaboration as a Mode of Library Value Creation

The literature on academic library collaboration shows that contemporary library value is increasingly produced through relationships rather than standalone services. In research support, libraries have expanded into bibliometrics, research data management, scholarly communication, open access, researcher identity management, and research-impact support (Haddow & Mamtora, 2017; Kennan et al., 2014; Si et al., 2019). In open science and research data management, libraries are positioned as infrastructure builders, educators, advocates, and mediators between researchers, institutional policy, publishers, funders, and technical systems (Marín-Arraiza, 2019; Rice, 2019).

Collaboration is also visible in learning and student-success work. Information literacy, reference consultations, and embedded instructional support are increasingly discussed as contributions to learning outcomes and academic achievement rather than merely as services delivered (Blummer & Kenton, 2018; Miller, 2018). This has two implications for the present study. First, collaborative library work is coded not only by the service area involved, but also by the partnership domain through which the work is enacted. Second, evidence is assessed relative to the implied theory of change. A library-faculty instructional partnership that reports numbers of workshops and students reached is evidencing activity or output; a claim about improved student capability requires evidence of learning or behavioral change.

The social mission literature broadens this view further. Academic libraries are described as contributors to community engagement, employability, public knowledge, cultural heritage, and social transformation (Corrall, 2022). The Sustainable Development Goals literature similarly positions academic libraries as actors that can map, assess, and communicate contributions to broader societal goals (Missingham, 2024; C. Thorpe & Gunton, 2021). These literatures support the inclusion of access, learning, research support, equity, cultural preservation, community engagement, and policy influence as impact domains.

2.3 From Outputs to Outcomes to Impact Claims

One recurring problem in library assessment is the conceptual slippage among outputs, outcomes, and impacts. Outputs are countable products of activity, such as visits, consultations, downloads, deposits, events, or participants. Outcomes refer to changes for users or institutions, such as improved information literacy, research visibility, data practices, open access compliance, or student engagement. Longer-term impacts refer to broader and more durable effects, such as strengthened research capacity, equitable knowledge access, cultural preservation, or policy influence.

The literature supports this distinction but also shows why it is hard to sustain in practice. Student-success studies can connect library use to educational outcomes when they have access to institutional data (Soria & Fransen, 2017; A. Thorpe et al., 2016). Research-support studies can document emerging service models, staff capabilities, service maturity, and researcher satisfaction (Kennan et al., 2014; Sawe et al., 2024; Si et al., 2019). SDG-related studies can map library activities to societal goals, but they also show that mapping activities to broad goals is not the same as demonstrating long-term societal impact (Missingham, 2024; C. Thorpe & Gunton, 2021). Thus, the evidentiary burden rises as claims move from activity to output, outcome, and impact.

Table 2 summarizes the sensitizing distinctions that guide the coding framework. The table is not a findings table; it is a conceptual bridge between the literature review and the empirical design. It clarifies why the analysis must code both the level of the claim and the evidence used to support it.

Table 2: Sensitizing distinctions guiding the analysis of collaborative impact claims.

Distinction	Definition	Evidence	Literature
Activity	A reported library action or collaborative effort.	Program descriptions, lists of partnerships, event announcements, service narratives.	Bowen (2009); Kennan et al. (2014)
Output	A countable product of activity.	Attendance, consultations, downloads, deposits, workshops, visits, processed collections.	Jager (2017); Głowacka (2019)
Outcome	A change for users, researchers, students, or institutional processes.	Learning assessment, retention/GPA association, satisfaction linked to service change, compliance or behavior change.	A. Thorpe et al. (2016); Soria & Fransen (2017); Nurse et al. (2018)
Long-term impact	A durable institutional, scholarly, public, or societal contribution.	Policy uptake, sustained capability development, cultural preservation evidence, equity indicators, longitudinal institutional change.	Urquhart (2018); C. Thorpe & Gunton (2021); Missingham (2024)
Legitimation claim	A statement that frames library work as strategically valuable or transformational.	Strategic rhetoric, alignment with university goals, public value language, transformation narratives.	Suchman (1995); DiMaggio & Powell (1983); Collins (2007)

As shown in Table 2, the same collaborative initiative may be represented at multiple evidentiary levels. A repository program may be an activity when described as a service, an output when deposits or downloads are counted, an outcome when researcher behavior changes, and a long-term impact when linked to open scholarship, public access, or policy. The value of claim-level analysis is that it prevents the strongest language in a document from determining the evidentiary classification. Instead, the classification depends on what the claim says and what evidence is provided.

2.4 Impact Domains in Academic Library Collaboration

The domains used in this study are derived from recurring areas in the literature and from the proposed corpus. Access remains foundational: libraries provide access to collections, spaces, technologies, expertise, and scholarly resources. Learning appears through information literacy, instruction, reference consultations, and embedded pedagogical work (Blummer & Kenton, 2018; Miller, 2018). Research support includes research data management, bibliometrics, open access, researcher identifiers, scholarly communication, and impact services (Haddow & Mamtora, 2017; Kennan et al., 2014; Rice, 2019; Si et al., 2019). Equity appears through access, inclusion, social justice, accessibility, open scholarship, and SDG-related reporting (Missingham, 2024; C. Thorpe & Gunton, 2021). Cultural preservation and community engagement appear in the social mission of academic libraries, especially where libraries steward special collections, archives, institutional repositories, oral histories, community knowledge, and public-facing heritage materials (Corrall, 2022). Policy influence is less likely to appear as a direct effect but may emerge through open science mandates, research assessment, scholarly communication policy, national library strategies, and institutional governance.

These domains are analytically useful because they differ in evidentiary difficulty. Access and service outputs are comparatively easy to count. Learning and research outcomes require stronger assessment designs. Equity, cultural preservation, community engagement, and policy influence often require longitudinal, qualitative, or external evidence. This variation is central to RQ3 and RQ5: the study asks not only which domains are visible, but whether the evidentiary basis matches the strength and time horizon of the claim.

2.5 Documents as Institutional Evidence and Genre

Document analysis provides the methodological bridge between library assessment and institutional theory. Documents are not passive containers of organizational facts; they are produced for audiences and purposes. Bowen (2009) defines document analysis as a systematic procedure for reviewing and evaluating documentary material, and this study uses that logic to examine the representational work performed by annual reports, strategic plans, impact reports, assessment webpages, and professional guidelines.

Genre matters because different public documents perform different communicative work. Strategic plans articulate future orientation, priorities, and desired identities; annual reports narrate completed activities and accomplishments; standards and frameworks express field-level norms; impact reports select evidence to make value claims visible. Research on strategic plans as organizational genres shows that such texts are not merely containers of strategy but official textual expressions of strategy practice (Cornut et al., 2012). Critical genre analysis similarly treats organizational documents as sites where broader institutional discourses are assembled and made persuasive (Fage-Butler, 2015). These insights support the article’s decision to treat document genre as the central comparative axis.

Institutional theory sharpens the interpretation of these documents. Organizations seek legitimacy by aligning themselves with valued norms, rules, and expectations in their fields (DiMaggio & Powell, 1983; Scott, 2014; Suchman, 1995). Academic libraries operate in a higher education environment where the valued vocabulary includes student success, research excellence, innovation, equity, transformation, public engagement, and measurable outcomes. Public value scholarship adds that organizations serving public or civic purposes often justify themselves not only through efficiency or use, but through broader claims about citizen, societal, cultural, and democratic value (Brewer, 2013; Moore, 1995).

This theoretical perspective does not imply that library documents are deceptive. Rather, it suggests that they are selective, strategic, and genre-bound. In this corpus, strategic plans provide aspirational and future-oriented claims, annual reports provide retrospective outputs and accomplishments, impact reports foreground selected evidence, and professional guidelines articulate field-level ideals. For this reason, RQ4 compares claim strength and evidentiary adequacy by document genre within the analyzed corpus rather than treating all documents as equivalent evidence.

2.6 Claim-Evidence-Legitimation Framework

The study uses a claim-evidence-legitimation framework to convert the literature into a coding and interpretation model. The framework begins with collaborative initiatives: library activities, services, programs, infrastructures, or projects involving at least one identifiable partner. These initiatives become visible in public documents through impact claims. Claims are then analyzed by impact level, evidence type, claim strength, impact domain, partnership domain, beneficiary, and evidentiary adequacy. The final interpretive layer identifies legitimation modes: strategic alignment, student success, research excellence, equity/social justice, public value, innovation/transformation, and compliance/accountability.

The framework rests on one boundary condition: the study analyzes public representation, not actual institutional impact or causal effectiveness. A document may make a strong claim about transformation, but the analysis evaluates only whether the document provides evidence adequate to that claim. This distinction is central to the contribution of the article. It makes possible a critical analysis of documentary evidence practices without making unsupported judgments about whether the underlying initiatives succeeded.

2.7 Sensitizing Expectations

The reviewed literature leads to four expectations that orient the analysis without functioning as formal hypotheses. First, output evidence is likely to be more common than outcome or long-term impact evidence because outputs are easier to count and communicate (Głowacka, 2019; Jager, 2017). Second, transformative claims are likely to be most common in strategic plans and professional frameworks, where aspirational language is genre-appropriate (Collins, 2007; Suchman, 1995). Third, research support and open scholarship claims are likely to involve multiple partnership domains because these services connect libraries to research offices, faculties, publishers, funders, and technical infrastructures (Haddow & Mamtora, 2017; Kennan et al., 2014; Rice, 2019). Fourth, equity, public value, cultural preservation, and transformation claims are likely to carry higher evidentiary risk because they require broader, longer-term, or externally corroborated evidence (Corrall, 2022; Missingham, 2024; C. Thorpe & Gunton, 2021).

Together, these expectations define the study’s contribution. The article does not ask whether academic libraries create impact in general. The literature already shows that libraries can contribute to learning, research, access, open scholarship, and public value. The more precise question is how public documents define, evidence, compare, and sometimes overextend those claims. That question is where the literature on library assessment, collaboration, public value, and institutional legitimacy converges.

3 Methodology

3.1 Research Design

This study uses qualitative comparative document analysis to examine how academic libraries define, evidence, and legitimate the impact of collaborative initiatives in public documents. The design combines directed qualitative content analysis with framework analysis. Directed content analysis is appropriate because the study begins with theoretically informed categories from library impact assessment, public value, and institutional theory, including outputs, outcomes, impacts, evidence type, and claim strength (Hsieh & Shannon, 2005). Framework analysis is appropriate because the study compares coded material across document genres and partnership domains while preserving a transparent matrix of analytic categories (Gale et al., 2013).

Document analysis is used here as a method for studying representation, not as a proxy for measuring actual institutional performance. Following Bowen (2009), public documents are treated as materials that can be systematically reviewed, categorized, and interpreted. However, they are also genre-bound institutional texts. Annual reports, strategic plans, impact frameworks, assessment webpages, and professional guidelines are produced for particular audiences and purposes. They can show what libraries claim, emphasize, count, narrate, and omit; they cannot, by themselves, verify whether collaborative initiatives caused real-world transformation.

3.2 Corpus Construction and Sampling Controls

The corpus is constructed as a purposive, comparative sample of public documents published between 2019 and 2026, with preference for 2019-2025 documents to capture post-pandemic reporting, open science, digital transformation, and emerging AI-related language. The corpus includes four document categories: annual reports, strategic plans, impact reports or assessment frameworks, and professional or sector guidelines. Professional guidelines are included as a comparison group because they articulate normative expectations against which institutional documents can be interpreted.

The sampling logic is summarized in Table 3. The design is purposive rather than statistically representative. The aim is not to estimate the prevalence of impact practices across all academic libraries globally, but to build a structured corpus capable of supporting comparative analysis of documentary impact representation. Document genre is the primary comparative axis. Region is retained as descriptive metadata and is used only when the corpus has sufficient balance to support interpretation.

Table 3: Corpus structure and comparison rules.

Dimension	Rule	Rationale
Genre	Include annual reports, strategic plans, impact/assessment reports, and professional frameworks; compare by genre before any broader inference.	Separates genre effects from substantive differences in impact representation.
Collaboration	Include only documents containing at least one identifiable collaborative initiative or partnership claim.	Keeps the empirical object focused on collaborative impact rather than general library value.
Region	Record region descriptively; use regional comparison only if each region has adequate institutional-document coverage.	Prevents underpowered regional interpretation.
Timeframe	Prioritize 2019-2025; include 2026 only when retrieval is stable and complete.	Captures post-pandemic digital, open science, equity, and transformation language.
Language	English-language public documents unless translation capacity is available.	Maintains coding consistency while acknowledging language bias.
Institutional type	Include research-intensive and teaching-focused institutions where public documents are available.	Reduces prestige bias and avoids over-reliance on elite research universities.

As shown in Table 3, the corpus is structured to make comparison possible without overstating generalizability. Document genre is treated as the central sampling dimension because strategic plans and annual reports are likely to differ in genre conventions. Strategic plans may emphasize aspiration, alignment, and future-oriented transformation, whereas annual reports are more likely to contain retrospective achievements and countable outputs. Professional guidelines are analyzed separately before being used as interpretive context for institutional documents.

Preprocessing identified possible impact statements using a search dictionary containing impact terms, collaboration terms, and domain terms. The extracted unit was the sentence, with adjacent context consulted during manual screening when a sentence depended on a heading, list item, or preceding sentence. The preprocessing pass retained sentences that contained impact-oriented language and either collaboration language or domain-specific language related to access, learning, research support, equity, cultural preservation, community engagement, or policy influence. Duplicate headers, table artifacts, OCR fragments, and boilerplate text were removed during manual screening.

3.3 Inclusion and Exclusion Criteria

Documents are eligible for inclusion if they are publicly accessible, attributable to an academic library, university library system, consortium, or relevant professional body, and contain explicit or implied discussion of impact, value, outcomes, strategic contribution, assessment, or evidence. For inclusion in the core analysis, a document must also contain at least one identifiable collaborative initiative or partnership claim. Documents may be PDF documents, HTML reports, strategic webpages, or institutional repository records, provided the text is stable and publicly available.

Documents are excluded if they require login access, cannot be attributed to a clear institution or body, duplicate another document in the same reporting period, contain only operational statistics without strategic, impact, value, or collaboration language, or contain no identifiable collaborative initiative. Purely promotional webpages are excluded unless they contain substantive strategy or assessment content. Documents with minimal impact language may be retained as contrast cases only outside the core collaborative-claim analysis.

3.4 Unit of Analysis: Collaborative Impact Claim

The unit of analysis is the collaborative impact claim rather than the whole document. An impact claim is any statement that attributes value, change, benefit, contribution, outcome, or transformation to a library service, partnership, program, infrastructure, or initiative. A collaborative impact claim is the subset of impact claims in which the claim names, implies, or is situated within a library initiative involving at least one identifiable internal or external partner. The analysis distinguishes all impact claims, collaborative impact claims, and claims with unspecified collaboration; only collaborative impact claims form the core analysis.

In this article, “validated” means manually screened and retained for the core claim-level dataset. It does not mean externally verified institutional impact. A claim was manually validated when it met the definition of a collaborative impact claim and contained enough contextual information to code impact level, evidence type, claim strength, partnership domain, and evidentiary adequacy.

The decision to code impact claims responds directly to the conceptual problem identified in the literature: library documents may use the same vocabulary of “impact” to describe activities, outputs, outcomes, and long-term impacts (Głowacka, 2019; Jager, 2017; Urquhart, 2018). Treating the claim as the unit of analysis prevents the document as a whole from being classified too broadly and allows the analysis to examine whether each claim is adequately evidenced.

3.5 Coding Framework

The study uses a hybrid deductive-inductive coding framework. Deductive codes are derived from the literature on library impact assessment, public value, institutional legitimacy, and qualitative document analysis (Bowen, 2009; Hsieh & Shannon, 2005; Saldaña, 2021; Schreier, 2012; Suchman, 1995). Inductive codes are added during pilot coding when documents contain recurring concepts not captured by the initial framework.

The main coding dimensions are presented in Table 4. These dimensions are designed to answer the research questions sequentially: definitions and implied meanings of impact for RQ1, evidence and impact levels for RQ2, domains for RQ3, comparative variables for RQ4, and evidentiary adequacy for RQ5.

Table 4: Claim-level coding framework.

Dimension	Definition	Categories
Impact definition	How impact is explicitly defined or implicitly framed.	Usage/value; learning; research contribution; equity; transformation; public value; unspecified.
Level	The level of change or contribution asserted by the claim.	Activity; output; outcome; long-term impact.
Evidence type	The form of support offered for the claim.	Metric; assessment result; case example; testimonial; policy citation; benchmark; external recognition; narrative assertion; no evidence.
Impact domain	The area in which value or change is claimed.	Access; learning; research support; equity; cultural preservation; community engagement; policy influence.
Partnership domain	The internal or external partner context associated with the claim.	Faculty; research office; IT; student services; cultural institution; community partner; government/policy body; consortium; unspecified.
Strength	The rhetorical force of the claim.	Descriptive; contributory; causal; transformative.
Adequacy	Alignment between claim strength and evidence provided.	Strong; moderate; weak; absent.
Mode	Institutional logic used to make the claim credible.	Strategic alignment; student success; research excellence; equity/social justice; public value; innovation/transformation; compliance/accountability.
Beneficiary	The actor or group said to benefit.	Students; faculty; researchers; institution; community; public/society; library staff.
Time horizon	The implied or stated timeframe of the benefit.	Immediate; short-term; long-term; unspecified.

As shown in Table 4, the coding framework separates what a claim is about from how strongly it is made and how well it is evidenced. This separation is crucial for RQ5. A claim that a library “supported student success” with attendance data is not coded in the same way as a claim supported by learning assessment or institutional student-outcome data. Likewise, a claim about “transforming research culture” is coded differently depending on whether it is supported by longitudinal evidence, policy change, researcher behavior change, or only strategic rhetoric.

Table 5: Evidentiary adequacy rubric for claim-level coding.

Claim level/type	Minimum evidence	Weak/absent when
Activity	Description of the action, partner, program, or collaborative mechanism.	The document names a value or priority but does not identify an action or partner.
Output	Count, completed deliverable, product, participation figure, deposited item, event, or service volume.	The document describes activity but provides no product, count, or completed deliverable.
Outcome	Evidence of change in behavior, learning, research practice, satisfaction linked to action, policy/process change, or institutional outcome.	The document uses outcome language but provides only activity or output evidence.
Long-term impact	Longitudinal evidence, external validation, sustained adoption, policy uptake, cultural/institutional change, or public-value evidence.	The document asserts broad or durable change without longitudinal, external, or systemic evidence.
Transformative claim	Systemic, longitudinal, external, or independently corroborated evidence proportionate to the claimed transformation.	The document uses transformation language as aspiration, alignment, or branding without commensurate evidence.

Table 5 provides the decision rules used to reduce subjectivity in adequacy coding. The rubric does not assume that every claim must provide long-term evidence. Instead, it applies proportionality: stronger and more durable claims require stronger evidence.

3.6 Coding Procedure and Reliability

Coding proceeded in six stages. First, all documents were logged in a sampling matrix with document ID, region, country, institution, document type, title, year, language, URL, retrieval status, local file path, inclusion rationale, and notes. Second, documents were screened against the inclusion and exclusion criteria. Third, a pilot subset of documents was coded to test the clarity and sufficiency of the codebook. The pilot included more than one document type and more than one region to expose genre and context variation early.

Fourth, the codebook was revised based on the pilot. Revisions included clarifying decision rules for evidentiary adequacy, separating unspecified partnership claims from validated collaborative claims, and adding legitimation mode as an interpretive code. Fifth, claims were coded at the claim level. Sixth, coded claims were exported into a structured analysis matrix for descriptive mapping, cross-tabulation, and qualitative interpretation.

Possible claims identified during preprocessing were manually screened against the definition of a collaborative impact claim. Claims were retained only when they attributed value, benefit, contribution, change, or transformation to a library initiative and when an internal or external partner, partnership domain, or collaborative mechanism was identifiable. Claims were excluded from the core analytic set when they made general impact statements but did not specify a collaborative mechanism. A stability check was conducted by re-reviewing a subset of claims after codebook revision, with attention to impact level, evidence type, partnership domain, claim strength, and evidentiary adequacy. Ambiguous cases were resolved through analytic memoing and application of the adequacy rubric in Table 5.

The analysis used a single-coder design with structured validation safeguards rather than formal intercoder reliability statistics. To strengthen reliability, the coder re-reviewed 25% of the validated collaborative impact claims across all document genres after codebook revision. Stable categories reviewed in this check were impact level, evidence type, impact domain, partnership domain, and claim strength. Evidentiary adequacy was treated as an interpretive judgment and rechecked against the adequacy rubric. Disagreements between initial and review coding were recorded in analytic memos and used to refine decision rules. This procedure supports coding transparency, but it remains a limitation compared with independent second-coder validation.

3.7 Analytical Strategy

The analysis combines qualitative interpretation with descriptive mapping. RQ1 is addressed through thematic analysis of explicit definitions and implied meanings of collaborative impact. RQ2 is addressed through coding evidence type and impact level, followed by cross-tabulation of evidence types across activity, output, outcome, and long-term impact claims. RQ3 is addressed by mapping impact domains and comparing the evidence used to substantiate each domain. RQ4 is addressed through comparison by document genre. RQ5 is addressed through evidentiary adequacy analysis, comparing claim strength against the type and level of evidence provided and identifying associated legitimation modes.

Table 6 links each research question to its analytic procedure and expected output. This table functions as a methodological audit trail, showing how each research question is operationalized rather than leaving the connection between questions and methods implicit.

Table 6: Alignment between research questions, data, and analysis.

RQ	Codes	Procedure	Output
RQ1	Impact definition; claim excerpt; document genre	Thematic analysis of explicit and implied meanings of collaborative impact.	Typology of impact meanings.
RQ2	Evidence type; impact level; claim strength	Cross-tabulation and qualitative interpretation of claim-evidence alignment.	Distribution of evidence across activities, outputs, outcomes, and impacts.
RQ3	Impact domain; evidence type; beneficiary	Domain mapping linked to evidence form and evidence depth.	Taxonomy of collaborative impact domains and evidence profiles.
RQ4	Document genre; claim strength; evidentiary adequacy	Comparative matrix analysis within the analyzed corpus by annual report, strategic plan, impact report, and professional framework.	Corpus-bounded genre patterns of claim strength and evidence adequacy.
RQ5	Claim strength; evidence type; impact level; evidentiary adequacy; legitimation mode	Gap analysis comparing rhetorical force with evidentiary basis and institutional logic.	Identification of claim-evidence mismatch and legitimation patterns.

As Table 6 shows, the study is designed so that no research question requires evidence beyond what public documents can reasonably provide. This is especially important for RQ5. The study does not judge whether a library actually failed to create transformation. It judges whether the document provides evidence commensurate with a transformative claim.

3.8 Use of R in the Analytical Strategy

R was used to support transparent data management, descriptive mapping, and reproducible tables and figures. The document-level sampling log and claim-level coding matrix were stored as structured CSV files. R scripts were used to inspect missing metadata, summarize the corpus by region and document type, tabulate claim-level codes, and generate tables for the manuscript. The purpose of R was not to convert the study into a statistical inference project; rather, it supported auditability and reduced the risk of manual counting errors.

R produced frequency tables, cross-tabulations, and visual summaries such as distributions of claim level by evidence type, impact domain by document type, legitimation mode by evidentiary adequacy, and evidentiary adequacy by claim strength. These descriptive outputs were interpreted alongside qualitative excerpts from the documents. Excerpts were used selectively to illustrate patterns, deviant cases, and conceptual distinctions.

3.9 Trustworthiness, Validity, and Bias Safeguards

The study addresses trustworthiness through transparent sampling, a structured codebook, pilot coding, an audit trail, and negative case analysis. Construct validity is strengthened by defining impact, evidence, collaboration, claim strength, and evidentiary adequacy before full coding. Credibility is strengthened by preserving document metadata, using claim-level excerpts, and reporting discrepant or weakly evidenced cases rather than only strong examples.

Several biases are anticipated. Availability bias may privilege institutions that publish polished reports. English-language bias may exclude relevant documents from non-English-speaking contexts. Prestige bias may overrepresent research-intensive universities with well-developed communications infrastructure. Genre bias may cause annual reports, strategic plans, and professional guidelines to appear substantively different when some differences reflect document purpose. Region is therefore treated as descriptive context unless the final sample contains adequate institutional-document coverage to support comparison.

Robustness checks were built into the analysis. First, claims with unspecified partnership domains were excluded from the core collaborative-impact analysis and reported separately. Second, professional frameworks were analyzed both with and without institutional documents to test whether sector-level norms distorted institutional patterns. Third, source-page texts were analyzed separately from full reports because some pages are substantive documents while others are retrieval portals. Fourth, a subset of claims was re-reviewed after codebook revision to assess coding stability. Fifth, automated claim extraction was treated only as preprocessing; the reported results are based on the screened collaborative-claim set.

3.10 Ethical Considerations

The study uses publicly available institutional and professional documents, so human-subject risk is low. Nevertheless, ethical interpretation remains important. Documents are cited accurately and interpreted in context. The study avoids deficit framing of institutions from under-resourced settings and does not treat limited public documentation as evidence of poor practice. Where claims are weakly evidenced, the analysis attributes that weakness to documentary representation rather than to the actual absence of institutional impact.

3.11 Methodological Limitations

The principal limitation is that document analysis cannot verify actual collaborative impact. It can identify how impact is defined, evidenced, emphasized, and legitimated in public documents. It cannot determine whether library initiatives caused the reported outcomes, nor can it capture internal assessment practices that are not publicly documented. A second limitation is sample availability: institutions differ substantially in how much they publish, how long documents remain accessible, and whether documents are available in English. A third limitation is interpretive subjectivity in coding evidentiary adequacy. The study mitigates this through codebook definitions, pilot coding, negotiated agreement, and transparent reporting of decision rules.

These limitations are consistent with the purpose of the study. The aim is not to measure the effectiveness of academic library collaboration directly. The aim is to analyze the documentary evidence regimes through which collaborative impact is constructed, communicated, and sometimes overstated.

4 Results

4.1 Corpus and Validated Claim Set

The results are based on the screened collaborative-claim set drawn from the document-level corpus. The retrieval process began with 43 public records. Twelve records were successfully retrieved as valid PDFs and converted to machine-readable text. Reachable source pages were used as a retrieval supplement but were not combined with the document-level claim analysis because some pages were substantive strategy texts while others were landing pages pointing to reports, plans, or repository files.

The 12 PDF documents contained 43,801 extracted words. Preprocessing identified 319 possible impact statements, which were then screened against the study definition of a collaborative impact claim. The final analytic set contains 145 validated collaborative impact claims. Claims were excluded from the core set when they made general impact statements but did not identify a partner, partnership domain, or collaborative mechanism. Table 7 summarizes this screening process.

Table 7: Corpus and validated collaborative-claim set.

Category	n
Public records in retrieval set	43
Valid PDF documents retrieved and converted	12
Possible impact statements screened after preprocessing	319
Validated collaborative impact claims retained	145
Claims excluded because partnership domain was unspecified	174
Reachable source pages retained as retrieval supplement	20

Table 7 establishes the denominator for the Results section. The tables and figures that follow report the 145 validated collaborative impact claims rather than the broader preprocessing universe.

Table 8: Document-level corpus included in the validated collaborative-claim analysis.

ID	Region	Genre	Title	Words	Screened
AF109	Africa	Annual report	DUT Library annual report	17493	84
AU102	Australia	Strategic plan	UWA Library strategic plan PDF	1984	10
AU108	Australia	Strategic plan	University of Sydney Library strategy	2166	20
EU102	Europe	Strategic plan	Library strategic plan PDF	160	0
NA111	North America	Strategic plan	University Library strategic plan	742	21
NA112	North America	Strategic plan	UC Riverside University Library strategic plan	915	1
NA113	North America	Strategic plan	University Libraries strategic plan	561	3
NA114	North America	Strategic plan	Library strategic plan	676	3
NA115	North America	Strategic plan	University Libraries strategic plan	358	9
NA116	North America	Strategic plan	UNG Libraries strategic plan	282	2
GL102	Professional	Standard	Standards for Libraries in Higher Education PDF	7353	56
GL105	Professional	Impact report	Academic Library Contributions to Student Success	11111	110

Table 8 shows the document-level corpus from which validated claims were drawn. The corpus includes annual report, strategic plan, impact report, and standard/framework genres. Region is reported descriptively; document genre is the main comparative axis.

Because a claim-level study can be distorted when one or two documents contribute many claims, Table 9 reports claim concentration by document. The largest contributor was the student-success impact report, followed by the annual report and professional standard. Together, these three documents account for a substantial majority of the validated claim set, which is why the Results are interpreted as patterns within a purposive corpus rather than sector-wide prevalence estimates.

Table 9: Document-level concentration of validated collaborative impact claims.

ID	Genre	Title	Claims	%
GL105	Impact report	Academic Library Contributions to Student Success	43	29.7
AF109	Annual report	DUT Library annual report	36	24.8
GL102	Standard	Standards for Libraries in Higher Education PDF	28	19.3
NA111	Strategic plan	University Library strategic plan	14	9.7
AU108	Strategic plan	University of Sydney Library strategy	10	6.9
AU102	Strategic plan	UWA Library strategic plan PDF	6	4.1
NA115	Strategic plan	University Libraries strategic plan	5	3.4
NA113	Strategic plan	University Libraries strategic plan	2	1.4
NA112	Strategic plan	UC Riverside University Library strategic plan	1	0.7

4.2 How Collaborative Impact Is Framed

Across the validated claims, impact was more often implied than explicitly defined. Documents tended to frame collaborative impact through domains of contribution: student learning, access to resources and spaces, research support, community engagement, equity, cultural preservation, and policy alignment. The most visible implied definition was educational: impact was associated with student success, learning, capability development, teaching support, and curriculum alignment. A second prominent meaning was access-oriented: impact appeared through language about collections, resources, spaces, open access, discovery, and availability. Research support was also visible, especially in documents that discussed scholarly communication, repositories, data, publishing, and research capacity.

This pattern indicates that impact operates less as a stable construct than as a strategic umbrella term. Documents rarely paused to distinguish activities, outputs, outcomes, and long-term impacts. Instead, they used impact language to connect collaborative library work with institutional priorities. Table 10 provides curated examples of the range of claim forms used in the corpus.

Table 10: Clean illustrative excerpts from validated collaborative impact claims.

Form	ID	Excerpt	Coding rationale
Output-based	AF109	Information literacy integration into the Cornerstone module was negotiated by the library and teaching partners.	The claim identifies a collaborative instructional output and links it to a concrete program integration.
Outcome-oriented	AF109	Workshops were associated with improvement in the submission of long-overdue theses and dissertations.	The sentence implies institutional/user change, but the evidence remains report-based rather than independently verified.
Transformative	AF109	A Wiley transformative agreement enabled DUT researchers to publish without article processing charges.	The claim uses transformative publishing language and asserts research benefit, but does not show longer-term research-culture change.
Symbolic	AU108	Library people, collections, and spaces empower the university community to pursue academic excellence.	The claim aligns library work with institutional mission and public value without operational evidence.
Equity-oriented	AU102	Services and spaces are framed as inclusive, welcoming, and responsive to community diversity.	The claim invokes equity and inclusion, but no user-level barrier-reduction evidence is provided.
Policy/accountability	AF109	The library awaited approval of a research data management policy and formalized RDM processes.	The claim is tied to research governance and policy infrastructure rather than direct user outcome evidence.

Table 10 shows why claim-level analysis is necessary. The examples include output-based, outcome-oriented, transformative, symbolic, equity-oriented, and policy/accountability claims. They also show how documentary language can imply stronger impact than the evidence provided.

4.3 Claim Level and Evidence Type

Table 11 summarizes how evidence types were distributed across impact levels in the validated collaborative-claim set. The most common evidence category was no explicit evidence, followed by metrics. Metrics were especially concentrated in output and some outcome-level claims. By contrast, activity-level claims were often descriptive and unsupported by specific evidence. Long-term impact claims were relatively uncommon, but when they appeared, they often relied on either metrics or narrative language rather than robust longitudinal evidence.

Table 11: Evidence type by impact level among validated collaborative impact claims.

Level	Case study	External recognition	Metric	Narrative	No evidence	Policy citation	Testimonial
Activity	5	0	0	3	28	2	1
Long-term impact	0	0	5	0	4	0	0
Outcome	3	1	30	3	28	2	0
Output	0	0	28	0	2	0	0

The distribution in Table 11 reveals a central evidentiary asymmetry. Output claims were strongly associated with metrics, which is expected because outputs are countable. Outcome claims were more mixed: many were supported by metrics, but many also lacked explicit evidence. This matters because outcome claims imply change, not merely activity. Long-term impact claims were the most evidentially demanding but were not consistently supported by correspondingly strong evidence.

Figure 2: Distribution of validated collaborative impact claims by impact level.

Figure 2 shows that outcome claims were the largest category, followed by activity and output claims. This is not evidence that the documents conclusively demonstrated outcomes. Rather, it shows that the language of outcomes, improvement, success, and development appeared frequently in validated collaborative claims. The analytical issue is whether those outcome-oriented statements were backed by evidence strong enough to support them.

4.4 Domain-Specific Evidence Profiles

The most visible impact domains were learning, access, and research support. Learning accounted for the largest share of validated collaborative claims. Access and research support were also prominent, while community engagement, equity, policy influence, and cultural preservation were less visible.

Figure 3: Primary impact domains identified in validated collaborative impact claims.

Figure 3 illustrates a clear hierarchy of documentary visibility. Learning and access are the dominant vocabularies of collaborative impact in the validated set. Research support is visible but secondary, despite the strategic importance of open scholarship, research data management, and scholarly communication in the literature. Equity, cultural preservation, and policy influence appear least often. This does not mean these domains are absent from academic library work. Rather, they are less frequently articulated as collaborative impact claims in the analyzed documents.

4.5 Genre Differences

Within this corpus, variation by document genre was more interpretable than variation by region because the verified corpus was regionally imbalanced. Table 12 shows that the impact report was dominated by learning claims, while the annual report showed a broader distribution across learning, access, research support, and community-oriented claims. Strategic plans contained access, learning, research support, equity, and cultural preservation claims, but many of these were brief and aspirational. The professional standard contained access, learning, community engagement, policy influence, and research-support language, reflecting its broader normative purpose.

Table 12: Impact domain by document genre among validated collaborative impact claims.

Genre	Access	Community engagement	Cultural preservation	Equity	Learning	Policy influence	Research support	Unspecified
Annual report	6	3	0	0	13	0	8	6
Impact report	1	4	0	1	32	0	3	2
Standard	4	7	0	0	9	3	3	2
Strategic plan	18	2	2	1	8	0	6	1

As Table 12 indicates, document genre affects what becomes visible as impact within the analyzed corpus. The strategic plans in this corpus describe desired directions and institutional alignment; the annual report provides retrospective accounts and more concrete achievements; the professional documents in this corpus articulate standards, ideals, and sector-level expectations. Genre is not noise in the data; it is part of how impact is constructed.

Table 13: Primary partnership domains identified in validated collaborative impact claims.

Partner domain	Claims	Percent
Community partners	40	27.6
Consortium/professional body	6	4.1
Cultural institutions	5	3.4
Faculty/academic schools	41	28.3
Government/policy body	5	3.4
IT/digital infrastructure	26	17.9
Research office/researchers	17	11.7
Student services	5	3.4

Table 13 shows that faculty and academic schools, community partners, IT/digital infrastructure, and research-related partners were the most common partnership domains. The pattern matters because collaboration is not only a topic in the study; it is the mechanism through which impact is claimed.

4.6 Evidentiary Adequacy and Claim-Evidence Mismatch

The most important critical finding concerns evidentiary adequacy. Table 14 compares claim strength with evidentiary adequacy. Most claims were descriptive or contributory. Transformative and causal claims were comparatively rare, but they were also the most vulnerable to evidentiary mismatch. Several transformative claims were coded as absent or weak in evidentiary support, meaning the claim used strong change-oriented language without commensurate documentary evidence.

Table 14: Evidentiary adequacy by claim strength among validated collaborative impact claims.

Strength	Strong	Moderate	Weak	Absent
Causal	0	0	1	2
Contributory	0	24	3	24
Descriptive	0	50	3	34
Transformative	0	0	2	2

Table 14 shows that the main problem is not rampant causal overclaiming. Rather, the more common issue is evidentiary thinness in descriptive and contributory claims, combined with a smaller but important set of transformative claims that exceed the evidence provided. In practical terms, many documents imply that library work supports learning, access, research, or institutional value, but they do not always specify whether the evidence is a count, a narrative, an assessment result, a benchmark, or a demonstrated change over time.

Figure 4: Distribution of evidentiary adequacy ratings among validated collaborative impact claims.

Figure 4 reinforces this point visually. No claims in the validated set were coded as strongly adequate. Moderate and absent evidence dominate the validated collaborative-claim set. The finding is consistent with the article’s core argument: academic library documents often make collaboration and impact strategically visible, but the evidentiary infrastructure for substantiating stronger claims is uneven.

4.7 Legitimation Modes

RQ5 also asks what forms of institutional legitimation accompany claim-evidence mismatch. Table 15 summarizes the main modes identified in the validated claims. Student success was the most frequent legitimation mode, followed by research excellence and public value. Equity/social justice and innovation/transformation appeared less often, but they carried high evidentiary risk because they tended to involve broader claims about institutional or societal change.

Table 15: Legitimation modes associated with validated collaborative impact claims.

Mode	Claims	Form	Evidence	Risk	Excerpt
Student success	87	Collaboration supports learning, success, or academic achievement.	Metric	Outcome language may be supported by participation or service evidence.	Information literacy was integrated into a core module through library-teaching collaboration.
Research excellence	20	Library infrastructure enables research visibility, open scholarship, or data practice.	No evidence	Reach, access, or service outputs may stand in for research-culture change.	A transformative publishing agreement was framed as enabling researchers to publish without APCs.
Public value	19	Library collaboration benefits communities, publics, or cultural memory.	No evidence	Community or cultural value may be asserted through narrative rather than outcome evidence.	Libraries engage campus and broader communities to advocate educate and promote value.
Compliance/accountability	6	Library work meets standards, policies, or accountability expectations.	Metric	Policy or standard alignment may be treated as impact.	Sample outcomes and metrics were linked to standards and accountability expectations.
Strategic alignment	5	Library work is aligned with institutional mission or priorities.	Metric	Strategic alignment may substitute for evidence of change.	Subject librarians collaborated with academic staff across faculties and sites.
Innovation/transformation	4	Library work changes service models, infrastructure, or institutional capacity.	No evidence	Transformative language may lack longitudinal or systemic evidence.	A digital library program was linked to improved discovery and user experience.
Equity/social justice	4	Library work expands inclusive access or reduces barriers.	No evidence	Values language may lack user-level or barrier-reduction evidence.	Services and spaces were framed as inclusive welcoming and responsive to diversity.

As Table 15 shows, legitimation modes do not simply decorate claims; they shape the evidentiary burden. Student-success claims require evidence of learning, persistence, engagement, or achievement if they move beyond participation. Research-excellence claims require evidence that distinguishes reach or visibility from changes in research practice. Equity, public value, and transformation claims require evidence that extends beyond values language or strategic alignment.

Table 16: Legitimation mode by evidentiary adequacy.

Mode	Absent	Moderate	Weak
Compliance/accountability	2	4	0
Equity/social justice	4	0	0
Innovation/transformation	4	0	0
Public value	9	8	2
Research excellence	11	7	2
Strategic alignment	2	3	0
Student success	30	52	5

Table 16 makes the mismatch visible by showing how adequacy varies across legitimation modes. The pattern supports the article’s theoretical claim: when documents use legitimacy vocabularies such as transformation, public value, equity, or research excellence, the evidence required to substantiate the claim becomes broader than the evidence usually available in routine reporting.

4.7.1 How Legitimation Is Textually Produced

The legitimation modes operate through recognizable textual moves, not only through topic labels. Table 17 shows recurrent moves: connecting library collaboration to student success, linking library infrastructure to research excellence, extending library value to publics or communities, and framing library work as innovation, transformation, or equity.

Table 17: Textual production of legitimation modes in collaborative impact claims.

Mode	Move	Excerpt	Note
Student success	Positions library collaboration as a contributor to learning, academic success, or capability development.	Information literacy integration into a core module was negotiated with teaching partners.	The rhetorical move links library-faculty collaboration to student learning infrastructure.
Research excellence	Positions library infrastructure, agreements, or services as enabling research visibility and publishing capacity.	A transformative agreement enabled researchers to publish without article processing charges.	The claim ties library-mediated infrastructure to research productivity, but long-term research impact remains unevidenced.
Public value	Extends library value beyond internal service use toward community, cultural, or civic benefit.	Libraries engage campus and broader communities to advocate, educate, and promote value.	The language makes public-facing engagement credible through civic and community orientation.
Innovation/transformation	Frames library work as modernizing infrastructure, service models, or institutional practice.	An expanded digital library program was linked to improved discovery and user experience.	Innovation language raises the evidentiary burden because it implies change beyond ordinary service provision.
Equity/social justice	Frames collections, spaces, or services as inclusive, accessible, and culturally responsive.	Collections were framed as providing diverse voices and Aboriginal and Torres Strait Islander perspectives.	The claim legitimates library value through inclusion, but documentary evidence of changed access is limited.

Table 17 demonstrates how legitimacy is built linguistically. Student-success language makes collaboration credible by tying library work to learning infrastructure. Research-excellence language makes library infrastructure visible as part of research capacity. Public-value language moves the beneficiary beyond the campus. Innovation and equity language raise the evidentiary burden because they imply changed systems, practices, or access conditions rather than simply completed activity.

4.8 Taxonomy of Collaborative Impact Claims

The results support a four-part taxonomy of collaborative impact claims: symbolic, output-based, outcome-oriented, and transformative. Symbolic claims align library work with valued institutional priorities but provide little or no evidence. Output-based claims provide counts or products of activity. Outcome-oriented claims assert change for users, researchers, students, or institutional processes. Transformative claims assert durable systemic, cultural, policy, or societal change.

Table 18: Proposed taxonomy of collaborative academic library impact claims.

Type	Feature	Evidence	Risk
Symbolic	Aligns library work with valued institutional or societal priorities.	Strategic language, mission alignment, values statements.	Legitimation without operational evidence.
Output-based	Reports countable products of collaborative activity.	Events, attendance, consultations, downloads, deposits, collections processed.	Treating outputs as if they demonstrate change.
Outcome-oriented	Claims change for users, researchers, students, or institutional processes.	Learning assessment, behavior change, satisfaction linked to service change, institutional data linkage.	Insufficient attribution or weak evidence of change.
Transformative	Claims long-term, systemic, cultural, policy, or public-value impact.	Longitudinal evidence, policy uptake, sustained capability development, cultural preservation outcomes, external corroboration.	Aspirational language exceeding available evidence.

Table 18 synthesizes the main conceptual contribution of the Results. The taxonomy does not rank claim types from bad to good. Symbolic claims are sometimes appropriate in strategic documents. Output-based claims are necessary for accountability. Outcome-oriented claims are essential for demonstrating user and institutional benefit. Transformative claims matter because libraries increasingly position themselves in relation to open scholarship, equity, cultural memory, and public value. The problem arises when documents blur these categories or use evidence suitable for one type of claim to support another.

4.9 Summary of Results

The results answer the research questions in five ways. First, impact is more often implied through domains of institutional contribution than explicitly defined. Second, evidence is unevenly distributed: metrics support many output and outcome claims, but a large share of claims provide no explicit evidence. Third, learning, access, and research support are the most visible domains, while equity, cultural preservation, and policy influence are comparatively underdeveloped in the validated set. Fourth, document genre shapes impact representation more clearly than region in the analyzed corpus. Fifth, the most serious evidentiary risk is not simply the absence of measurement, but mismatch: documents sometimes use outcome-oriented or transformative language without evidence adequate to the strength of the claim.

These findings support the study’s central methodological decision: the collaborative impact claim must be the unit of analysis. Whole-document summaries would obscure the coexistence of output metrics, outcome language, strategic rhetoric, and evidentiary gaps within the same document.

5 Discussion

5.1 Impact as Genre-Bound Documentary Construction

The results show that academic library impact is less often defined as a stable assessment construct than assembled through recurring domains of institutional contribution. Documents tended to imply impact through learning, access, research support, community engagement, equity, cultural preservation, and policy alignment. This finding answers RQ1 by showing that “impact” functions as an integrative strategic vocabulary. It allows libraries to connect diverse activities to institutional value, but it also permits conceptual slippage among activities, outputs, outcomes, and long-term impacts.

This pattern aligns with the literature on library impact assessment, which warns that usage, value, outcomes, and impact are related but not interchangeable (Głowacka, 2019; Jager, 2017; Urquhart, 2018). The present analysis extends that literature by showing how the slippage appears in public documents. The documents do not merely report evidence; they construct a public account of why library work matters. In this sense, annual reports, strategic plans, standards, and impact reports are organizational texts that translate library work into claims legible to university leaders, funders, professional bodies, and publics.

This interpretation also supports the institutional-theory framing of the study. The language of transformation, student success, access, equity, open scholarship, and strategic alignment reflects the normative environment in which academic libraries operate (DiMaggio & Powell, 1983; Suchman, 1995). Libraries use these vocabularies because they are meaningful within contemporary higher education. The analytical problem is not that such language is illegitimate, but that the evidentiary basis for the language varies substantially.

5.2 Collaboration Is Often Under-Specified as a Mechanism

The study defined collaboration as an inclusion condition, yet the screening process showed that many impact statements did not specify a partnership domain. This matters because collaborative impact is not demonstrated simply by invoking collaborative values. A claim is stronger when it identifies the partner, the shared objective, the library’s contribution, and the evidence linking the collaborative mechanism to the stated benefit.

The validated claims suggest that collaboration is most visible when documents name faculty, academic schools, community partners, IT/digital infrastructure, research offices, consortia, or professional bodies. Even then, many claims describe benefits to students, researchers, the university, or the public without fully explaining how the partnership operated. This weakens the capacity of public documents to distinguish collaborative impact from general library service activity.

5.3 Evidence Gaps Are Claim-Evidence Mismatches

The findings for RQ2 indicate that evidence is unevenly distributed across claim levels. Metrics are common for outputs and some outcome-oriented claims, while many activity and outcome claims contain no explicit evidence. Long-term impact claims are comparatively rare but evidentially demanding. This suggests that the central problem is not simply that libraries need more metrics. The deeper problem is evidentiary fit.

Different claims require different kinds of evidence. Output claims can be supported with counts. Outcome claims require evidence of change. Transformative claims require evidence of sustained change, institutional uptake, cultural shift, capability development, policy influence, or public value. A count of workshops delivered may be useful, but it cannot by itself substantiate a claim about transformed student capability. A repository download count may signal reach, but it does not necessarily demonstrate a change in research culture or knowledge equity.

The student-success literature illustrates what stronger outcome evidence can look like: linking library interactions to institutional outcomes, using multimethod designs, or interpreting library use in relation to retention, academic performance, engagement, and degree completion (Haddow & Joseph, 2010; Nurse et al., 2018; Soria & Fransen, 2017; A. Thorpe et al., 2016). However, many collaborative impact domains are harder to evaluate with such methods. Community engagement, cultural preservation, equity, and open scholarship may require longitudinal, qualitative, policy, or network evidence. The results therefore suggest that libraries need differentiated evidence strategies rather than a single impact-measurement template.

5.4 Transformative Language Carries the Highest Evidentiary Burden

The findings for RQ5 show that overclaiming is best understood as evidentiary mismatch rather than simple exaggeration. Most claims in the validated set were descriptive or contributory. Strong causal and transformative claims were relatively rare. However, when transformative language appeared, it was often weakly supported or unsupported. The more widespread issue was not excessive causal language, but the routine use of outcome-oriented language without evidence adequate to the level of change implied.

This distinction is important. A document can be rhetorically restrained and still evidentially thin. For example, a claim that a library “supports student success” is contributory rather than causal, but it still requires some indication of what support means, who benefited, and what evidence links the activity to student success. Likewise, a claim that library collaboration “advances open scholarship” may be plausible, but it is stronger if supported by deposits, policy compliance, researcher adoption, publishing outcomes, or documented changes in practice.

The study therefore reframes the problem from “libraries overstate impact” to “documents often under-specify the evidentiary chain between collaboration and impact.” This is a more useful diagnosis because it points to improvement. Libraries do not need to avoid ambitious claims. They need to align claim strength with evidence type and make the theory of change visible.

5.5 Claim-Evidence Alignment Can Improve Library Reporting Practice

The practical implication is that libraries can design impact reporting around claim-evidence alignment. A more rigorous report would not simply add more indicators; it would specify the level of each claim and match it with appropriate evidence. If the claim is about activity, descriptive evidence may be sufficient. If the claim is about output, counts and service statistics are appropriate. If the claim is about outcomes, the report needs evidence of change. If the claim is transformative, the report needs longitudinal, corroborated, or external evidence, or else frames the claim as an aspiration rather than a demonstrated impact.

The findings also suggest that collaborative initiatives benefit from greater relational specificity. Documents can identify partners, shared goals, the library’s contribution, the evidence collected, and the beneficiary. This would make collaboration more than a rhetorical marker. It would turn collaboration into an analyzable pathway through which impact is produced and evidenced.

Finally, genre clarity matters. Strategic plans can articulate ambitions and theories of change without presenting aspirations as achieved impacts. Annual reports can document activities and outputs without implying that counts automatically demonstrate outcomes. Impact reports can bridge the two by explaining how activities and outputs connect to outcomes and longer-term impacts.

Table 19: Reporting recommendations by claim type.

Type	Recommendation
Activity	Identify the partner, purpose, collaborative mechanism, and institutional alignment.
Output	Provide counts or deliverables and explain why they matter for the stated collaborative goal.
Outcome	Provide evidence of change, not only participation or service volume.
Long-term impact	Use longitudinal, policy, behavioral, cultural, institutional, or external evidence.
Transformative	Avoid transformation language unless evidence demonstrates systemic or durable change.

Table 19 translates the article’s analytic framework into reporting practice. The central recommendation is proportionality: the stronger the claim, the more explicit the theory of change and the stronger the evidence required.

5.6 Theoretical Contribution

The article contributes to library assessment scholarship by shifting attention from indicators alone to evidence regimes. The question is not simply which indicators libraries use, but how documents construct acceptable evidence of value. It also contributes to institutional theory by showing how academic library documents use field-level legitimacy vocabularies–student success, innovation, equity, transformation, open scholarship, and public value–to position the library within higher education. Finally, it contributes methodologically by demonstrating the value of claim-level document analysis. The claim-level approach makes visible the coexistence of symbolic, output-based, outcome-oriented, and transformative claims within the same document.

5.7 Limitations

The study is limited to public documents and therefore analyzes representation rather than actual institutional effectiveness. The findings cannot determine whether a collaborative initiative succeeded, caused user change, or produced long-term public value. They can only assess whether documents provided evidence proportionate to the claims they made.

The English-language and availability criteria also shape the corpus. Institutions with stronger communications infrastructure may be more visible, and documents from under-resourced or non-English-speaking contexts may be underrepresented. Genre imbalance is another limitation: strategic plans, annual reports, impact reports, and standards perform different communicative functions, and comparisons across genres require that difference in mind.

Claim concentration is a further limitation. The student-success impact report, annual report, and professional standard contributed a large share of the validated claims. This concentration may amplify learning, access, standards, and student-success language relative to other domains of collaborative library impact. The Results therefore support analytic pattern identification rather than prevalence claims.

Finally, evidentiary adequacy requires interpretive judgment. The rubric reduces subjectivity by applying proportionality between claim strength and evidence type, but borderline cases remain possible. The study used single-coder analysis with re-review and memoing rather than independent intercoder reliability statistics. The study therefore treats adequacy coding as an analytic judgment about public documentation, not as an evaluation of institutional competence.

6 Conclusion

Academic libraries increasingly use public documents to position collaborative work as evidence of institutional value, student success, research excellence, equity, public engagement, open scholarship, and transformation. This article reconstructs that problem as one of claim-evidence-legitimation: how public documents construct collaborative impact claims, what evidence they provide, and how they legitimate claims whose strength may exceed their evidentiary basis.

The central finding is conceptual as much as empirical. Collaborative impact is not a generic synonym for library value. It is better analyzed through the relationship among collaborative initiative, impact claim, impact level, evidence type, claim strength, evidentiary adequacy, and legitimation mode. This approach clarifies why participation counts, narratives, strategic alignment, learning assessment, policy uptake, and transformation language are not interchangeable forms of evidence.

For academic library reporting, the implication is direct: documents need to align the strength of their claims with the evidence provided. Activity claims identify partners and purposes. Output claims provide counts and explain their relevance. Outcome claims demonstrate change. Long-term and transformative claims require longitudinal, external, policy, behavioral, cultural, or institutional evidence. Where such evidence is not yet available, documents can still articulate aspirations, but they need to avoid presenting aspiration as demonstrated impact.

For research on library value, the contribution is a claim-level method for studying public evidence regimes in academic librarianship. The approach preserves the value of strategic communication while making its evidentiary assumptions visible. By showing how academic libraries publicly evidence and legitimate collaborative value, the study clarifies where claim-evidence alignment can improve both assessment practice and institutional reporting.

References

ACRL Research Planning and Review Committee. (2020). 2020 top trends in academic libraries: A review of the trends and issues affecting academic libraries in higher education. College & Research Libraries News, 81(6), 270. https://doi.org/10.5860/crln.81.6.270

Blummer, B., & Kenton, J. M. (2018). Academic libraries and student learning outcomes. Performance Measurement and Metrics, 19(1), 75–87. https://doi.org/10.1108/PMM-11-2017-0053

Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27–40. https://doi.org/10.3316/QRJ0902027

Brewer, J. D. (2013). The public value of the social sciences. Bloomsbury Publishing. https://doi.org/10.5040/9781472545121

Clunie, S., & Parrish, D. A. (2018). How assessment websites of academic libraries convey information and show value. Performance Measurement and Metrics, 19(3), 203–212. https://doi.org/10.1108/PMM-12-2017-0061

Collins, R. (2007). The BBC and "public value". Medien & Kommunikationswissenschaft, 55(2), 164–184. https://doi.org/10.5771/1615-634X-2007-2-164

Cornut, F., Giroux, H., & Langley, A. (2012). The strategic plan as a genre. Discourse & Communication, 6(1), 21–54. https://doi.org/10.1177/1750481311432521

Corrall, S. (2022). The social mission of academic libraries in higher education. In The social future of academic libraries (pp. 109–148). Facet. https://doi.org/10.29085/9781783304738.007

DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147. https://doi.org/10.2307/2095101

Fage-Butler, A. (2015). Investigating interdiscursivity in hospital strategic plans using foucauldian discourse analysis. HERMES - Journal of Language and Communication in Business, 27(54), 35–52. https://doi.org/10.7146/hjlcb.v27i54.22946

Gale, N. K., Heath, G., Cameron, E., Rashid, S., & Redwood, S. (2013). Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Medical Research Methodology, 13(1), 117. https://doi.org/10.1186/1471-2288-13-117

Głowacka, E. (2019). Research on the impact of academic libraries: Areas, methods, indicators. Library Management, 40(8/9), 583–592. https://doi.org/10.1108/LM-02-2019-0005

Haddow, G., & Joseph, J. (2010). Loans, logins, and lasting the course: Academic library use and student retention. Australian Academic & Research Libraries, 41(4), 233–244. https://doi.org/10.1080/00048623.2010.10721478

Haddow, G., & Mamtora, J. (2017). Research support in australian academic libraries: Services, resources, and relationships. New Review of Academic Librarianship, 23(2-3), 89–109. https://doi.org/10.1080/13614533.2017.1318765

Hsieh, H.-F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277–1288. https://doi.org/10.1177/1049732305276687

Jager, K. de. (2017). Approaches to impact evaluation in academic libraries. IFLA Journal, 43(3), 282–287. https://doi.org/10.1177/0340035217696321

Kennan, M. A., Corrall, S., & Afzal, W. (2014). "Making space" in practice and education: Research support services in academic libraries. Library Management, 35(8/9), 666–683. https://doi.org/10.1108/LM-03-2014-0037

Marín-Arraiza, P. (2019). ORCID in the open science scenario: Opportunities for academic libraries. Mitteilungen Der Vereinigung Österreichischer Bibliothekarinnen Und Bibliothekare, 72(2), 478–493. https://doi.org/10.31263/voebm.v72i2.2811

Miller, R. E. (2018). Information literacy and instruction: Reference consultations and student success outcomes. Reference & User Services Quarterly, 58(1), 16–21. https://doi.org/10.5860/rusq.58.1.6836

Missingham, R. (2024). Australian academic libraries and the united nations sustainable development goals. IFLA Journal, 51(1), 17–31. https://doi.org/10.1177/03400352241252973

Moore, M. H. (1995). Creating public value: Strategic management in government. Harvard University Press.

Nurse, R., Baker, K., & Gambles, A. (2018). Library resources, student success and the distance-learning university. Information and Learning Sciences, 119(1/2), 77–86. https://doi.org/10.1108/ILS-03-2017-0022

Oakleaf, M. (2010). The value of academic libraries: A comprehensive research review and report. Association of College; Research Libraries.

Oakleaf, M. (2011). Are they learning? Are we? Learning outcomes and the academic library. The Library Quarterly, 81(1), 61–82. https://doi.org/10.1086/657444

Rice, R. (2019). Supporting research data management and open science in academic libraries: A data librarian’s view. Mitteilungen Der Vereinigung Österreichischer Bibliothekarinnen Und Bibliothekare, 72(2), 263–273. https://doi.org/10.31263/voebm.v72i2.3303

Saldaña, J. (2021). The coding manual for qualitative researchers (4th ed.). SAGE Publications.

Salisbury, F., & Peasley, J. (2018). Measuring the academic library. Information and Learning Sciences, 119(1/2), 109–120. https://doi.org/10.1108/ILS-07-2017-0068

Sawe, E., Mwai, N., & Oyieke, L. I. (2024). Doctoral students’ satisfaction with research support services at public universities in kenya. South African Journal of Libraries and Information Science, 90(1). https://doi.org/10.7553/90-1-2291

Schreier, M. (2012). Qualitative content analysis in practice. SAGE Publications.

Scott, W. R. (2014). Institutions and organizations: Ideas, interests, and identities (4th ed.). SAGE Publications.

Si, L., Zeng, Y., & Guo, S. (2019). Investigation and analysis of research support services in academic libraries. The Electronic Library, 37(2), 281–301. https://doi.org/10.1108/EL-06-2018-0125

Soria, K. M., & Fransen, J. (2017). The impact of academic library resources on undergraduates’ degree completion. College & Research Libraries, 78(6), 812. https://doi.org/10.5860/crl.78.6.812

Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of Management Review, 20(3), 571–610. https://doi.org/10.5465/amr.1995.9508080331

Thorpe, A., Lukes, R., & Bever, D. J. (2016). The impact of the academic library on student success: Connecting the dots. Portal: Libraries and the Academy, 16(2), 373–392. https://doi.org/10.1353/pla.2016.0027

Thorpe, C., & Gunton, L. (2021). Assessing the united nation’s sustainable development goals in academic libraries. Journal of Librarianship and Information Science, 54(2), 208–215. https://doi.org/10.1177/09610006211005528

Urquhart, C. (2018). Principles and practice in impact assessment for academic libraries. Information and Learning Sciences, 119(1/2), 121–134. https://doi.org/10.1108/ILS-06-2017-0053