How to Write Chapter 2 (RRL) for IT Capstone — With Examples

Chapter 2 (RRL) review of related literature capstone is where most capstones get sent back for revision. The pattern is the same every time.

The student spent two weeks finding sources, copied summaries from each one into a Word document, listed them under “Related Literature” and “Related Studies,” and submitted to their adviser. The adviser writes “needs synthesis” in red ink and hands it back.

The problem isn’t the sources. The problem is what the student did with them.

Chapter 2 isn’t a book report. It’s an argument. It says, “here’s what people know about this topic, here’s what others have tried, here’s where the gap is, and here’s how my project fills it.” That’s the version that gets approved.

This guide walks through how to write that version. The structure. The sources. The paragraph format. The synthesis section most students skip. APA citations done right. A full sample excerpt at the end so you can see what excellence looks like.

By the time you finish reading, you should be able to write your first three RRL paragraphs tonight.

What Chapter 2 is actually for

Three things every Chapter 2 must demonstrate:

• You read the literature. You can name authors, years, and findings. You’re not making it up.
• You understood it. You can explain what each source contributes, not just summarize their abstracts.
• You identified the gap. You can point at a specific problem or opportunity that existing work hasn’t fully solved — and that’s where your project sits.

If your Chapter 2 hits those three marks, the panel will respect it. If it hits only the first one (“you read it”), they’ll send it back.

The difference between book report mode and researcher mode is one sentence per paragraph. Book report says “Smith (2024) studied X and found Y.” Researcher says “Smith (2024) studied X and found Y, which directly supports the present study’s choice to…”

That second sentence is what panels look for.

The 5 subsections of Chapter 2

Standard structure in most PH and Indian BSIT programs:

• 2.1 Related Literature — books, expert articles, government publications, established theory
• 2.2 Related Studies — other researchers’ work similar to yours (theses, journal articles, conference papers)
• 2.3 Synthesis — your analysis tying everything together (the section students skip)
• 2.4 Conceptual Framework — your model of how your project will work (IPO is standard)
• 2.5 Theoretical Framework — established theories that ground your work (sometimes optional)

Some schools merge 2.1 and 2.2 into one section called “Related Literature and Studies.” Some split each into Foreign and Local. Check your school’s manual. The substance below works for all variants.

Where to find good sources (free and legitimate)

Most students go straight to Google and grab the first 10 results. That’s how you end up citing blog posts and Wikipedia. Use real research databases.

Primary sources to use:

• Google Scholar (scholar.google.com) — free, broadest coverage, good citation tracking
• ResearchGate — free with academic account, full PDFs from authors
• ScienceDirect — your school likely has a subscription; ask the librarian
• IEEE Xplore — best for computing-related theses
• DOAJ (Directory of Open Access Journals) — entirely free, peer-reviewed
• ACM Digital Library — some articles free, more with school subscription
• Past capstones in your school library — your school’s own thesis archive is gold
• Government publications and statistics — PSA, DOST, DICT for PH; equivalents for India

Sources to AVOID citing:

• Wikipedia — useful for learning, never cite directly
• Random blog posts — not peer-reviewed
• Sources older than 5 years — unless they’re foundational (Davis 1989 for TAM is fine; a random 2008 paper isn’t)
• Sources you haven’t actually read — panels can tell when you’re citing something you only saw the abstract for

When using Google Scholar, click the “Cited by” link below each result to find newer papers that built on the original. This is the fastest way to build a chain of related work.

How many sources do you need

Most BSIT schools in the Philippines require 15 to 25 sources minimum. Some require more for thesis-track students. Quality matters more than count.

Recommended distribution:

• 60-70% recent (published in the last 5 years)
• 30-40% foundational (classic works in your area, can be older)
• 60% foreign sources (international researchers and contexts)
• 40% local sources (Philippine or Asian context — important for relevance)

For an AI-themed capstone in 2026, you should have at least 5 sources published in 2024 or later. If your most recent source is 2020, the panel will assume you didn’t search recently.

How to write a single RRL paragraph correctly

This is the most important section in this guide. Get the paragraph structure right and your entire Chapter 2 improves immediately.

Every RRL paragraph has three parts:

Part 1: Source introduction. Name the author, year, and what they studied or wrote about. One sentence.

Part 2: Key findings. What did they actually discover or argue? Two to three sentences.

Part 3: Connection to your study. How does this relate to, support, or differ from your project? One to two sentences.

Here’s a fully annotated example using a hypothetical AI tutoring capstone:

PART 1: Cruz and Reyes (2024) investigated the effectiveness of AI tutoring chatbots among Filipino senior high school students in their study titled Adoption Patterns of AI-Powered Learning Assistants in Metro Manila Public Schools.

PART 2: They surveyed 312 students across 8 schools and found that AI tutors significantly improved math performance for students in the lowest quartile, with an average score increase of 14% over a 6-week trial. However, they also identified that engagement dropped sharply after the third week when the chatbot did not adapt to student progress.

PART 3: This finding directly informs the present study, which addresses the engagement-drop issue by incorporating an adaptive difficulty mechanism that adjusts question complexity based on the student’s running accuracy.

Notice the second sentence in part 3 — “the present study addresses the engagement-drop issue.” That’s the sentence that turns a summary into a contribution. Without it, you’re just describing what Cruz and Reyes did. With it, you’re showing why you cited them.

Every RRL paragraph in your Chapter 2 should follow this 3-part structure. Without exception.

APA 7th citation format

Most BSIT schools in the Philippines and India require APA 7th edition. Here’s the short version.

In-text citations

When you mention a source in your paragraph, cite the author and year in parentheses:

• One author: (Smith, 2024)
• Two authors: (Smith & Cruz, 2024)
• Three or more authors: (Smith et al., 2024)
• With a direct quote: (Smith, 2024, p. 45)
• Multiple sources together: (Smith, 2024; Cruz, 2023; Reyes, 2022)

If you mention the author’s name in the sentence itself, only the year goes in parentheses:

• “Smith (2024) found that…”
• “According to Cruz and Reyes (2024)…”

Reference list entries

The reference list at the end of your documentation is alphabetical by author surname. Hanging indent. Here are the formats for the source types you’ll use most:

Book:

Smith, J. A. (2024). Modern capstone methodologies. Academic Press.

Journal article:

Cruz, M. R., & Reyes, A. B. (2024). Adoption patterns of AI-powered learning assistants. Journal of Philippine Computing Studies, 12(3), 45-58. https://doi.org/10.xxxx/jpcs.2024.12.3.45

Online article or report:

Department of Information and Communications Technology. (2024). Annual digital readiness report. DICT Philippines. https://dict.gov.ph/reports/2024

Thesis or capstone:

Aquino, M. T. (2023). AI-assisted capstone topic selection for BSIT students [Unpublished capstone project]. University of the East.

Conference paper:

Lim, K., & Tan, S. (2024). Adaptive learning systems in Southeast Asian schools. In Proceedings of the 2024 IEEE Conference on Educational Technology (pp. 112-119). IEEE.

Use a reference manager (Zotero is free) to handle the formatting. Manual APA almost always has errors panels will catch.

2.1 Related Literature — template and example

The Related Literature section covers what experts and authorities have written about your topic. Books, government publications, established review articles. Usually 5 to 10 entries.

Each entry follows the 3-part paragraph structure above. Group by theme, not by source type. Don’t list books first then articles — group by what they say.

Full example paragraph for the AI tutoring capstone:

According to the Department of Information and Communications Technology (2024) in its annual digital readiness report, only 38% of Filipino public school students have consistent access to digital learning resources outside of school hours.

The report highlights the disparity between urban and rural learners, noting that students in Metro Manila have nearly three times the access to supplementary educational tools compared to students in Region V.

This finding establishes the contextual need for low-resource AI tutoring solutions that can operate offline or on minimal connectivity, which the present study addresses through a TFLite-based mobile deployment strategy.

Notice three things in this example:

• The source is a government report, treated with the same paragraph structure as any other RRL entry
• Specific statistics (38%, three times the access) make the claim verifiable
• The last sentence connects to the present study’s “offline” feature

2.2 Related Studies — template and example

The Related Studies section covers other researchers’ work that’s similar to yours. Theses from your school library, journal articles, conference papers. Usually 8 to 15 entries.

Same 3-part structure. Different source types.

Full example paragraph:

Lim and Tan (2024), in their paper Adaptive Learning Systems in Southeast Asian Schools presented at the IEEE Conference on Educational Technology, developed an adaptive AI tutoring platform tested across schools in the Philippines, Vietnam, and Indonesia.

Their system used a Bayesian Knowledge Tracing model to estimate student mastery and adjust question difficulty accordingly. They reported a 22% improvement in long-term retention compared to a control group using static practice.

The present study extends Lim and Tan’s work by adapting their Bayesian Knowledge Tracing approach to a single-school deployment context with offline capability, addressing the network-dependency limitation they acknowledged in their future work section.

Again, notice the structure. Source introduction. Key findings (specific numbers). Connection to your study (extends, adapts, addresses a limitation).

2.3 The synthesis — the section students skip

This is where Chapter 2 either passes or fails review. After listing 15-25 sources, students often submit Chapter 2 without ever connecting the dots. The synthesis section is one to two paragraphs that show you understood everything you read.

What synthesis does:

• Identifies common themes across multiple sources
• Notes where sources agree and where they disagree
• States the gap in existing research
• Establishes the position of your study in that gap

Full example synthesis paragraph:

The reviewed literature and studies converge on three findings about AI tutoring systems. First, multiple authors agree that adaptive difficulty is the strongest factor in sustained student engagement (Cruz & Reyes, 2024; Lim & Tan, 2024; Aquino, 2023).

Second, while AI tutoring shows consistent positive outcomes in urban high-resource contexts, evidence from rural or low-connectivity contexts is significantly thinner (Department of Information and Communications Technology, 2024; Reyes, 2023).

Third, none of the reviewed studies addressed the combined challenge of adaptive difficulty AND offline capability in a single system. This gap motivates the present study, which integrates Bayesian Knowledge Tracing adaptation with TFLite mobile deployment to support both engagement and rural accessibility.

This single paragraph does what a panel wants to see — themes, agreements, gap, contribution. If your Chapter 2 has a synthesis paragraph this strong, the chapter is likely approved.

2.4 Conceptual Framework

Most BSIT capstones use the Input-Process-Output (IPO) model. It’s a diagram plus a paragraph explaining it.

Standard IPO structure:

Example for the AI tutoring capstone:

Explanation paragraph:

Figure 2.1 illustrates the conceptual framework of the study using the Input-Process-Output (IPO) model. The Input layer captures the student’s profile, current topic of study, and historical performance data from previous sessions. The Process layer applies Bayesian Knowledge Tracing to estimate the student’s mastery of each subtopic, then selects the next question at an optimal difficulty level using the adaptive algorithm proposed by Lim and Tan (2024). The Output layer delivers a personalized question to the student, provides real-time feedback on their response, and updates the progress report visible to both the student and their teacher.

Create the actual diagram in draw.io or Lucidchart, export as PNG, embed it as Figure 2.1.

Some studies use flowcharts or system diagrams instead of IPO. Check what your school prefers.

2.5 Theoretical Framework (when required)

If your school requires a Theoretical Framework section, pick an established theory that aligns with your project. Common choices in BSIT capstones:

• Technology Acceptance Model (TAM) — Davis (1989). Most popular for software adoption studies. Useful when your system needs user acceptance.
• Diffusion of Innovations — Rogers (1962). Useful when your project introduces something new to a community.
• ISO/IEC 25010 Software Quality Standards — for studies focused on software quality evaluation.
• ADDIE Model — for instructional design and educational technology projects.
• Theory of Reasoned Action (TRA) — Fishbein & Ajzen (1975). For studies focused on user behavior and intent.
• Unified Theory of Acceptance and Use of Technology (UTAUT) — Venkatesh et al. (2003). Modern extension of TAM.

Don’t pick three theories. Pick one. Apply it to your study explicitly.

Example application of TAM to the AI tutoring capstone:

This study is grounded in the Technology Acceptance Model (TAM) developed by Davis (1989). TAM posits that user adoption of a technology depends on two primary constructs: Perceived Usefulness (the degree to which a user believes the system will improve their performance) and Perceived Ease of Use (the degree to which the user believes the system requires minimal effort).

The present study operationalizes these constructs through user acceptance testing measuring both dimensions on a 5-point Likert scale, allowing direct comparison of AI tutoring acceptance against Davis’s original findings.

If your school doesn’t require a theoretical framework, you can skip this section. If they do require it, pick a theory you can actually apply, not just name-drop.

Sample 3-paragraph Chapter 2 excerpt

Here’s what excellence looks like — three paragraphs that would together pass any BSIT panel review. All three reference the same hypothetical AI tutoring capstone.

2.1 Related Literature (excerpt)

According to the Department of Information and Communications Technology (2024) in its annual digital readiness report, only 38% of Filipino public school students have consistent access to digital learning resources outside of school hours. The report highlights the disparity between urban and rural learners, noting that students in Metro Manila have nearly three times the access to supplementary educational tools compared to students in Region V.

This finding establishes the contextual need for low-resource AI tutoring solutions that can operate offline or on minimal connectivity, which the present study addresses through a TFLite-based mobile deployment strategy.

2.2 Related Studies (excerpt)

Lim and Tan (2024), in their paper Adaptive Learning Systems in Southeast Asian Schools presented at the IEEE Conference on Educational Technology, developed an adaptive AI tutoring platform tested across schools in the Philippines, Vietnam, and Indonesia.

Their system used a Bayesian Knowledge Tracing model to estimate student mastery and adjust question difficulty accordingly. They reported a 22% improvement in long-term retention compared to a control group using static practice. The present study extends Lim and Tan’s work by adapting their Bayesian Knowledge Tracing approach to a single-school deployment context with offline capability, addressing the network-dependency limitation they acknowledged in their future work section.

2.3 Synthesis (excerpt)

The reviewed literature and studies converge on three findings about AI tutoring systems. First, multiple authors agree that adaptive difficulty is the strongest factor in sustained student engagement (Cruz & Reyes, 2024; Lim & Tan, 2024; Aquino, 2023). Second, while AI tutoring shows consistent positive outcomes in urban high-resource contexts, evidence from rural or low-connectivity contexts is significantly thinner (Department of Information and Communications Technology, 2024; Reyes, 2023).

Third, none of the reviewed studies addressed the combined challenge of adaptive difficulty AND offline capability in a single system. This gap motivates the present study, which integrates Bayesian Knowledge Tracing adaptation with TFLite mobile deployment to support both engagement and rural accessibility.

Adapt this structure to your own project. Replace the AI tutoring topic with yours. Replace the citations with your actual sources. The format works for any capstone topic — IT, business, education, healthcare.

Quick reference — what makes a good RRL paragraph

Three-question checklist for every paragraph you write:

1. Did I name the author and year? If not, fix it before moving on.
2. Did I state what the source actually found, with specifics? Numbers, percentages, sample sizes — not “the author talked about X.”
3. Did I connect the finding to my project? “This relates to…” or “This informs…” or “The present study extends…”

If yes to all three, the paragraph is defensible. If even one is missing, the paragraph is a book report.

Common mistakes that get Chapter 2 sent back

Eight patterns that consistently kill Chapter 2 reviews:

1. Just summarizing without connection. Every paragraph must connect back to your study. No exceptions.
2. Sources that don’t actually relate. If you can’t say in one sentence why a source is relevant, drop it.
3. Citing sources you didn’t read. Panels can tell. They’ll ask follow-up questions, and you won’t have answers.
4. No synthesis section. This is the most common reason Chapter 2 gets sent back. The synthesis is mandatory, not optional.
5. APA format inconsistencies. Mixed comma styles, missing italics, wrong hanging indent. Use Zotero.
6. Missing in-text citations. Every claim from a source needs a parenthetical citation. Even paraphrased ones.
7. References don’t match in-text citations. Every source cited in your chapters must appear in References, and vice versa.
8. Conceptual framework is just a list, not a model. IPO is a relationship between elements, not three bullet points side by side.

If your draft has any of these, fix it before submitting to your adviser. The adviser is the practice run for defense, and you don’t want to waste that pass on formatting issues.

How long should Chapter 2 be

Most BSIT capstones have Chapter 2 between 15 and 25 pages. The page count breakdown roughly:

• 2.1 Related Literature — 5 to 8 pages (5-10 entries)
• 2.2 Related Studies — 7 to 12 pages (8-15 entries)
• 2.3 Synthesis — 1 to 2 pages
• 2.4 Conceptual Framework — 1 to 2 pages with the diagram
• 2.5 Theoretical Framework — 1 to 2 pages (if required)

If your Chapter 2 is under 10 pages, it’s probably underdone. If it’s over 30, you’re probably padding or including irrelevant sources.

Now go to Google Scholar. Open 5 tabs. Start writing.

Chapter 2 doesn’t get written by waiting until you’ve found “enough” sources. It gets written one paragraph at a time, starting from the first source you can connect to your study.

Open Google Scholar. Search your topic. Pick five papers. Read the abstracts. Pick the one most relevant to your project. Write the paragraph. Repeat.

By the time you have 20 paragraphs, you have a Chapter 2.

If you haven’t seen the full Chapter 1-5 structure yet, our Capstone Chapter 1-5 Template covers every chapter with examples.

For UML diagrams that go in Chapter 3, our UML guides cover every diagram type panels expect. If you haven’t picked your capstone topic yet, browse 150 Best Capstone Project Ideas for IT Students 2026.

For free working source code to study and reference in your synthesis, see our Python projects library and free projects library for other stacks.

Start tonight. One paragraph. Author, finding, connection. That’s the only formula you need.