How to Write How-To Guide from Your Reading Notes (With Citations)

The Three Types of Reading Notes and What They're Good For

Not all reading notes are equally useful when writing a how-to guide from your reading notes. Before diving into the process, it helps to identify what type of notes you've actually taken — because each type requires a different approach when converting to instructional content.

Type 1: Summary notes What they are: Your paraphrase of what the author said. "In Chapter 3, Field argues that the choice of statistical test depends on the distribution of the data and the research hypothesis." How-to guide use: These can provide the background for why a step exists, but they require verification against the source before being cited as established facts. Summary notes can drift from what the author actually said.

Type 2: Reaction/interpretation notes What they are: Your analysis or interpretation of what you read. "This means that running a t-test on non-normally distributed data isn't just inaccurate — it's the wrong question entirely. You're testing whether means differ when the means may not be the meaningful measure." How-to guide use: These are the most useful for how-to guides because they represent your synthesis of the principle — what it actually means for practice. They can't be attributed to the author (they're your interpretation), but they can become the "what this means" layer of a step.

Type 3: Application notes What they are: Your observations from actually trying the technique. "Tried the normal distribution assumption check with Q-Q plots on the response time data. The plots looked wrong — clear S-curve suggesting non-normality. Switched to Mann-Whitney U test. This worked." How-to guide use: These are gold. Application notes record your direct experience with a procedure, which is the most grounded form of how-to knowledge — and one no other note type provides.

---

How Reading Notes Differ From Highlights as How-To Source Material

Article 813 in this series covered how-to guides from highlights. This article addresses reading notes — your own synthesis and interpretation, not verbatim extracted text.

Highlights (813)	Reading notes (814)
Exact verbatim text from the source	Your words, based on your reading
Can be quoted directly with page number	Cannot use quotation marks
Attribution is clear: author said this	Attribution requires care: did author say this, or is this your interpretation?
Closer to the source; less interpretive	Closer to practice; more interpretive
Show what the author said	Show what you understood/applied
Verification step optional for direct quotes	Verification required before citing as author's claim

The reading notes advantage: notes that represent your synthesis are already one step closer to how-to guide prose than highlights, because you've already done the interpretation work. The risk is attribution drift — when notes you think record what an author said actually record your interpretation of what an author said.

---

The Verification Requirement

Before using any reading note as a citation-backed claim in a how-to guide, verify it against the source.

The verification rule:

• If you're writing "According to [Author], the correct procedure is X" — verify that the author actually says this in the source.
• If you're writing "Based on my reading of [Source], the key principle is X" — no need to verify; you're attributing interpretation to yourself, not to the author.
• If you're writing "X works because [reason]" with no attribution — this is either common knowledge or an uncited claim. Either it doesn't need a citation, or it needs one you find.

A simple attribution checklist for each note you use:

FOR EACH NOTE USED IN THE HOW-TO GUIDE:

Note text: [What my note says]
Note type: [Summary / Reaction-interpretation / Application]

For summary notes:
□ Have I verified this against the source? [Y/N]
□ If yes: page number where this claim appears: [p. N]
□ Does my note accurately represent what the author says? [Y/N]
□ If not: correct the note before using it

For reaction/interpretation notes:
□ I am not attributing this to the author — I am attributing this to myself
□ Attribution form: "In my reading of [Source], this means..." or "The practical implication is..."

For application notes:
□ I am attributing this to my experience — "In practice, I found..."
□ If I want to link to supporting evidence: [Source that corroborates this observation]

---

The Three-Layer How-To Guide Structure From Reading Notes

A how-to guide built from reading notes has three layers that each type of note supplies:

STEP [N]: [Verb phrase — the specific action]

Action: [What the reader does — second person, imperative voice]
[May come from application notes or from understanding the procedure]

Why this works (from summary/interpretation notes):
[The principle behind this step — sourced from your reading]
Attribution: Either verified: "According to [Author] (Source, Year, p. N)..."
Or interpreted: "My reading of [Source] suggests that this step works because..."

What to watch for (from application notes):
[What you observed in practice when you tried this step]
Attribution: "In applying this, I found that..." or "Practitioners consistently report..."
[If citing another source for this observation: attribution]

Common mistake:
[What goes wrong — from your interpretation or application notes]

---

Step 1: Sort Notes by Type and Assess Verifiability

Go through your reading notes on the topic and sort them:

NOTE INVENTORY

Summary notes (paraphrase of what author said):
1. "[Note text]" — [Source] — Verified against source? [Y/N]
2. "[Note text]" — [Source] — Verified? [Y/N]

Interpretation notes (my understanding/synthesis):
1. "[Note text]" — my interpretation of [Source]
2. "[Note text]" — my interpretation

Application notes (what I observed in practice):
1. "[Note text]" — from my own practice on [Date/Context]
2. "[Note text]"

Notes that need verification before use:
[List summary notes that haven't been verified against the source]

Notes I can use without verification:
[Interpretation and application notes, clearly labeled as such]

---

Step 2: Build the Procedural Skeleton

As with highlights-based guides (813), write the steps from memory before working with your notes:

1. What do you do first?
2. What do you do next?
3. What comes after that?

The notes will provide the rationale and evidence for each step, but the step order should come from your procedural understanding. If you can't produce the steps from memory, your notes have given you conceptual knowledge that hasn't yet become procedural — which is useful information about where you need more practice.

---

Step 3: Map Notes to Steps and Identify Gaps

With your steps listed, assign each note to the step it supports:

STEP [N] — [Title]:
Action: [The step — from my procedural understanding]

Summary note (verified): "[Verified paraphrase]" — [Author, Source, p. N]
Interpretation note: "[My synthesis of what this means practically]"
Application note: "[What I observed when doing this step]"

If gap (no note for this step):
Option A: I'll note this is from practice: "In practice, [what I've found]..."
Option B: I need to find a supporting source
Option C: This step is obvious enough not to require citation

---

Before/After Worked Example

Context: A graduate student in quantitative psychology has read extensively in Andy Field's Discovering Statistics Using IBM SPSS Statistics (5th edition, 2018) and has 80 reading notes from her study sessions. She wants to write a how-to guide for her lab group on "How to choose the right statistical test for your research question."

Her note types (sample):

Summary note (unverified): "Field says that the choice between parametric and non-parametric tests depends on whether your data is normally distributed." [Needs verification against exact page]

Interpretation note: "What this really means in practice: if your DV is a Likert scale (1-5 or 1-7), it probably isn't normally distributed, which means most psychology studies using Likert scales should be using non-parametric or ordinal tests — but most researchers I've read use t-tests and ANOVA anyway. This is a widespread practice issue."

Application note: "When I ran the Shapiro-Wilk test on my response time data, it came back significant (p = .03), which means I should reject the assumption of normality. Field recommends looking at the histogram and Q-Q plot alongside the test. Histogram showed positive skew. Switched to Mann-Whitney U. Note: Field warns (I think Chapter 5) that Shapiro-Wilk is oversensitive with large samples — so with N > 100, the test might show non-normality even when the data is basically normal."

Her steps from memory:

1. Identify the type of research question (difference, relationship, or prediction)
2. Identify the number of groups or variables
3. Check assumptions for parametric tests
4. Select the parametric or non-parametric test

Verification check: The summary note ("Field says parametric vs. non-parametric depends on normality") needs to be verified against the source. She checks Field (2018) Chapter 5: confirmed — Field discusses assumption testing in Chapter 5, specifically p. 168-169 on the Shapiro-Wilk test. Application note about Shapiro-Wilk being oversensitive with large samples — she verifies: "With larger sample sizes, Shapiro-Wilk becomes very sensitive and can detect trivial deviations from normality." (Field, 2018, p. 169) — confirmed.

Before (notes as outline, no differentiation of note types):

How to Choose a Statistical Test:

• Type of question (difference/relationship)
• Number of groups
• Check normality
• Field says parametric depends on normality
• I found Shapiro-Wilk misleading for large samples

No structure, mixes verified and unverified notes, no attribution clarity.

After (how-to guide with note-type differentiation):

---

How to Choose the Right Statistical Test for Your Research Question

What you'll be able to do after this guide: identify the correct statistical test for a research question based on the type of question, the number of groups or variables, and the distribution of your dependent variable.

Prerequisites: a defined research question; collected data; access to SPSS or R.

This guide does NOT cover: running the tests in software; interpreting the output; effect size calculations.

---

Step 1: Identify the type of research question.

Ask yourself: is my question about a difference between groups (e.g., does Group A score higher than Group B?), a relationship between variables (e.g., does X correlate with Y?), or a prediction (e.g., does X predict Y?)?

The test family depends on this answer: difference questions use t-tests, ANOVAs, or their non-parametric equivalents; relationship questions use correlation or regression; prediction questions use regression or discriminant analysis.

Why this comes first: This distinction is foundational for test selection — the test family (difference vs. relationship vs. prediction) is determined by the question, not by the data. Choosing a test before clarifying the question type is the most common initial error in statistical decision-making. (This reflects my synthesis of Field, 2018, Chapters 7-9, where each test family is organized by question type.)

Step 2: Check assumptions before selecting the specific test.

For difference questions, the central assumption check is whether your dependent variable is normally distributed. Run the Shapiro-Wilk test. If the result is non-significant (p > .05), the normality assumption holds and you can use a parametric test (t-test, ANOVA). If significant (p ≤ .05), consider a non-parametric equivalent.

According to Field (2018, p. 169): "With larger sample sizes, Shapiro-Wilk becomes very sensitive and can detect trivial deviations from normality" — meaning that with N > 100, a significant result doesn't automatically mean your data is meaningfully non-normal. Look at the histogram and Q-Q plot alongside the test result.

From my own application of this: I've found that the Q-Q plot gives better intuitive information than the test alone — an S-curve shape on the Q-Q plot is a strong signal of non-normality regardless of sample size. The Shapiro-Wilk p-value plus the visual inspection together produce a better decision than either alone.

[Continues for Steps 3-4]

---

Sources: Field, A. (2018). Discovering Statistics Using IBM SPSS Statistics (5th ed.). SAGE Publications.

---

The three note types are clearly differentiated: summary notes (verified and cited), interpretation notes (labeled as synthesis), and application notes (labeled as practice observations). Together they produce a guide that's both grounded in the literature and honest about what came from practice.

---

Prompts to Reuse

How-To Guide From Reading Notes

I'm writing a how-to guide on: [Specific procedure]
Sources I've read: [Book(s)/Paper(s) with reading notes]
My steps (from memory, before checking notes): [List N steps]
Target reader: [Their background level]

My notes, sorted by type:
Summary notes (paraphrases of source — need verification):
  1. "[Note]" — Source: [Author, Title, Year] — Verified? [Y/N] — Page: [N if verified]
  
Interpretation notes (my synthesis — clearly labeled as mine):
  1. "[Note]" — My interpretation of [Source]
  
Application notes (from my practice):
  1. "[Note]" — From my practice on [Context]

Step-to-note mapping:
Step [N] — [Title]:
  Action: [What reader does]
  Supporting summary note (verified): "[Note]" — [Attribution]
  My interpretation: "[What this means practically]"
  Practice observation: "[What I found when doing this]"
  Gap: [Step with no note support — how I'll handle it]

Draft a how-to guide:
- Title: "How to [Task]"
- Intro: outcome + prerequisites + scope boundary
- Numbered steps in execution order
- Attribution clearly differentiated:
  Verified summaries: "According to [Author] ([Source, Year, p. N])..."
  My interpretation: "My reading of [Source] suggests..." / "In practice, this means..."
  Application notes: "In applying this, I found..."
- Sources section at the end with full citations

Accuracy constraint: Do not attribute interpretation notes to the author. 
Flag any summary note not yet verified against the source.

---

Key Takeaways

1. Know your note type before using it: summary notes can be attributed to the author (after verification); interpretation notes can only be attributed to yourself; application notes record your practice.
2. Verify summary notes against the source before citing them: reading notes drift from what the author actually said — never cite a summary note as the author's position until you've confirmed the source.
3. Interpretation and application notes are the most useful for how-to guides: they represent your synthesis and practice, which are what instructional content needs.
4. Build the step list from memory before reading your notes: this diagnostic tells you whether your understanding is procedural (you can produce the steps) or conceptual (you understand the principles but not the procedure).
5. Label attribution type clearly in the guide: readers need to know whether each claim comes from a verified source, your synthesis of a source, or your practice — conflating these undermines the guide's credibility.

---

Conclusion

Reading notes from your study sessions are more interpretive and practice-grounded than highlights — and those properties make them well-suited for how-to guide writing, where instructional application matters more than exact quotation. The challenge is attribution discipline: summary notes require verification before being cited; interpretation notes must be attributed to yourself, not the author; application notes document what you actually found in practice. A how-to guide that clearly differentiates these three layers is both more credible and more useful than one that presents all note types as if they were citations from the source.

Try WebSnips free — save your study session notes organized by source and note type, so your next how-to guide can pull verified summary notes, your interpretive synthesis, and your practice observations from a single searchable collection rather than from scattered notebooks.