By Cristo Leon, Ph.D.

Last reviewed August 21, 2026.

Teaching APA Citation: Why Students Struggle and What Faculty Can Do

Faculty regularly encounter student papers with missing references, incorrect in-text citations, incomplete source information, or citations added almost as an afterthought. These problems are often interpreted as evidence that students simply “do not know APA.”

Generative artificial intelligence (AI) makes this problem even more visible. A student may use an AI tool to brainstorm ideas, develop an outline, revise paragraphs, generate a table, or proofread an assignment. The student then searches online for how to cite the AI tool, adds a reference at the end of the paper, and assumes the requirement has been satisfied.

From the student’s perspective, the task may appear complete: I used something, and I cited it.

From the faculty perspective, however, important information is still missing.

The problem may therefore be larger than students making APA formatting mistakes. Students may not have been taught to distinguish among citation, disclosure, and documentation or to understand why these practices matter to scholarly communication (León & Kudelka, 2026; Lipuma & León, 2024).

Why Do College Students Struggle With APA Citation?

American Psychological Association (APA) Style provides a standardized approach to scholarly communication. Uniformity and consistency allow readers to focus on ideas rather than formatting and make it easier to locate key information, findings, and sources. APA guidelines also encourage authors to disclose essential information and appropriately credit the sources that contributed to their work (American Psychological Association, 2013).

Yet students may encounter APA primarily as a formatting system.

They learn where to place parentheses, when to italicize a title, how to construct a reference list, and where commas and periods belong. Those conventions are necessary, but mastering them does not necessarily mean that students understand why scholarly work requires citation.

This creates a pedagogical problem.

Faculty may evaluate students according to conventions we assume they already understand. Students, meanwhile, may perceive citation as a technical requirement to complete before submitting an assignment.

The predictable cycle is familiar: students submit incorrect citations, faculty identify the errors, students correct the formatting, and the underlying conceptual problem remains largely unchanged.

Teaching Why We Cite Before Teaching How We Cite

One way to address this problem is to begin citation instruction one step earlier.

Before asking students to construct an APA reference, we can ask:

Why do scholars cite sources in the first place?

Citation allows writers to identify where ideas and evidence originated and enables readers to locate and evaluate those sources. In this sense, citation is part of participating in a scholarly conversation rather than simply a mechanism for avoiding plagiarism.

APA Style supports this process by providing consistency. Its purpose is not to make every paper look identical for its own sake. Consistent conventions reduce unnecessary distractions and allow readers to concentrate on the argument, evidence, and sources.

Once students understand this purpose, the mechanics of APA have a clearer function.

The instructional sequence therefore becomes:

Why do we cite? → How does APA communicate source information? → How do we apply those principles to the tools we use?

The final question has become increasingly important with the widespread adoption of generative AI.

Generative AI Exposes a Gap in Traditional Citation Instruction

Consider a student who uses an AI system throughout an assignment.

The student asks the AI to suggest possible topics, develops an outline from the response, asks it to revise several paragraphs for clarity, generates a table, and finally uses the tool to proofread the paper.

The student then includes the AI system in the reference list.

The student has addressed one question:

What did you use?

But that citation does not necessarily answer:

How did you use it?

Nor does it answer:

What evidence can you provide of that use?

These questions represent three related but distinct responsibilities that become particularly useful when teaching students about AI-supported academic work:

Citation = What did you use?Student Disclosure of Work = How did you use it?Documentation = What evidence can you provide?

The distinction provides faculty with a simple framework for moving students beyond treating AI citation as something they can find on the internet and attach to the end of an assignment.

1. Citation: What Did You Use?

Citation identifies the source or tool.

For an AI tool, students need to understand how the required citation system identifies information such as the developer, date or version, name of the tool, and source information.

For example, the instructional handout developed for our students presents the general AI-tool reference format as:

Company Name. (Year of version). AI Tool Name (Version information) [Large language model]. URL

Students must also distinguish between parenthetical and narrative in-text citations.

These are teachable—and assessable—skills.

But correctly citing the tool establishes only what was used.

It does not establish what the tool actually did.

2. Student Disclosure of Work: How Did You Use It?

The second responsibility is therefore Student Disclosure of Work.

Disclosure concerns the process through which the student created the academic work, including the contributions of AI or other software tools.

For undergraduate writing, this can be organized around three phases:

Phase 1: Research and formulation
Did the student use the tool to explore a topic, formulate ideas, conduct preliminary research, develop questions, or organize an argument?

Phase 2: Creating the written work
Did the tool contribute to text, images, tables, figures, or other content?

Phase 3: Editing and proofreading
Was the tool used only for spelling and grammar, or did it contribute more substantially by changing tone, suggesting citations, providing content, or revising the writing?

This distinction becomes especially important for sophisticated writing assistants. The instructional guidance, for example, distinguishes basic proofreading from substantive interventions involving tone, content, citations, or other significant changes.

Disclosure therefore answers a question that a citation alone cannot:

What role did the technology play in creating this work?

3. Documentation: What Evidence Can You Provide?

The third responsibility is documentation.

Documentation provides evidence supporting the student’s disclosure of how a tool was used.

Depending on the assignment and the nature of the AI contribution, documentation may include prompts, AI-generated responses, transcripts, or other records of the interaction. The current instructional guidance, for example, requires the prompts and responses to be included when a full transcript is provided for direct quotations or extended paraphrasing of AI-generated material.

Documentation is therefore distinct from both citation and disclosure.

A student could cite an AI tool without adequately disclosing how it contributed to the work.

A student could also disclose that an AI tool contributed to the work without providing sufficient documentation of what occurred.

Teaching these as separate concepts gives students a clearer framework for understanding their responsibilities.

The Faculty Problem Is Also an Instructional-Design Problem

Once these distinctions are established, a more uncomfortable question emerges:

If we expect students to know these things, have we explicitly taught them?

And a second question follows:

If we taught them, how do we know they understood?

Simply distributing a handout does not establish that students read or understood it. Conversely, repeatedly correcting APA errors in submitted assignments is an inefficient way to determine whether students possess foundational citation knowledge.

A relatively simple intervention is to pair explicit instructional materials with a short, automatically graded knowledge check.

The objective is not to assess sophisticated scholarly writing. It is simply to establish whether students understand the information they have been given.

Students can identify the correct APA reference from several options. They can distinguish parenthetical from narrative citation. They can match parts of a reference with their appropriate positions. They can determine whether a particular use of AI constitutes proofreading or substantive editing. They can identify when an appendix or disclosure is required.

These questions can be assessed through multiple-choice, matching, ordering, and true/false items without creating an additional open-ended grading burden for faculty.

From “Students Should Know APA” to Evidence of Learning

This creates an important shift in how faculty interpret student errors.

Without explicit instruction and assessment, an incorrect citation could reflect many different problems: lack of prior preparation, misunderstanding, carelessness, confusion about disciplinary expectations, or an inability to apply a known rule.

A handout paired with a knowledge check establishes a clearer instructional sequence:

Instruction → Comprehension → Application

The information was provided.

The student demonstrated that they understood it.

The student was subsequently asked to apply that knowledge in authentic academic work.

This does not guarantee correct application—and that is precisely where the approach becomes interesting.

If students perform well on an APA knowledge check but continue to make significant citation, disclosure, or documentation errors in their academic work, the problem can no longer be explained simply as “students do not know APA.”

Instead, we need to ask:

Why are students failing to transfer what they know about scholarly attribution into what they actually produce?

That is a different educational problem.

Citation Knowledge Is Not the Same as Citation Practice

The increasing use of generative AI provides faculty with an opportunity to reconsider citation instruction more broadly.

AI did not create the underlying problem. It made the problem easier to see.

Students have always needed to understand where ideas originate, how sources contribute to their work, and how readers can evaluate those contributions. AI complicates those relationships because a technological tool may participate at multiple points in the intellectual and communicative process.

Teaching students merely how to construct an AI reference therefore addresses only part of the challenge.

Students need to understand three separate responsibilities:

Citation: What did you use?

Student Disclosure of Work: How did you use it?

Documentation: What evidence can you provide?

Before evaluating whether students fulfill those responsibilities, however, faculty should establish that students have actually been taught what they mean.

A short instructional resource and knowledge check will not solve every problem associated with APA Style, academic integrity, or generative AI. What they can do is remove an important ambiguity.

Instead of beginning with “Why didn’t the student follow the rule?”, we can begin with three more useful questions:

Did we explicitly teach the rule? Did the student demonstrate understanding? And can the student transfer that understanding into academic practice?

For faculty struggling with recurring APA errors, that may be the more productive place to start.

Sources

American Psychological Association. (2013). Publication Manual of the American Psychological Association (6th ed.). American Psychological Association. http://www.apastyle.org/products/4200066.aspx

León, C., & Kudelka, M. (2026). Auditing GenAI Literature Search Workflows: A Replicable Protocol for Traceable, Accountable Retrieval in Student-Facing Inquiry. AI in Education, 2(2), 42. /Research/Education (CLDM_Ds_Peer-reviewed). https://doi.org/https://doi.org/10.3390/aieduc2020008

Lipuma, J., & León, C. (2024). Disclosure of Support Statement: Increasing Student Transparency About Support from Software Like ChatGPT. Atena Editora Journal of Engineering Research, 4(7), 11. /Research/Education (CLDM_Ds_Peer-reviewed). https://doi.org/https://doi.org/10.22533/at.ed.317472426025