---

📚 Research Methodology & Article Writing - Easy Notes

---

📌 CHAPTER 1: Introduction to Research

Types of Research

Key Terms

• Research Objective - What the researcher clearly intends to find out. (e.g., "To determine the prevalence of diabetes in adults aged 40-60 in Karachi")
• Rationale - Why this study? What gap does it fill? What was wrong or missing in previous studies?
• Operational Definition - Exactly how you will measure your variables. (e.g., "Hypertension = BP ≥140/90 on two separate readings")
• Literature Search - Systematically searching journals, conference papers, and databases to understand what's already known before starting your research.

---

📌 CHAPTER 2: Types of Study Designs

1. Descriptive Studies (Describe what's happening - no hypothesis)

2. Analytical Studies (Test a hypothesis - always have 2+ groups)

Observational (researcher watches, doesn't intervene)

Memory trick: Cohort = forward in time (cause → effect). Case-Control = backward (effect → cause).

Experimental (researcher assigns exposure)

Blinding Types

Why blind? To prevent bias in results. (e.g., if you know you're getting the real drug, your pain might "feel better" due to psychology - the placebo effect)

---

📌 CHAPTER 3: Sampling

What is a Sample?

Example: Instead of testing all 200 million Pakistanis, you test 400 people representing the whole country.

Sampling Techniques

1. Probability Sampling (everyone has an equal chance - statistically valid)

2. Non-Probability Sampling (not statistically ideal, but practical)

---

📌 CHAPTER 4: Variables, Data & Presentation

Variables

• Dependent Variable = The OUTCOME you're measuring. (e.g., blood sugar level, death, recovery)
• Independent Variable = The CAUSE or EXPOSURE you think influences the outcome. (e.g., diet, smoking, drug type)

Example: Studying if smoking causes lung cancer → Smoking = independent variable; Lung cancer = dependent variable.

Types of Data

DATA
├── QUALITATIVE (categories)
│   ├── Nominal - No order (e.g., blood group: A, B, O, AB; gender: M/F)
│   └── Ordinal - Has order but no fixed distance (e.g., cancer stage I, II, III, IV; pain scale: mild/moderate/severe)
│
└── QUANTITATIVE (numbers)
    ├── Discrete - Whole numbers only (e.g., number of children: 0, 1, 2, 3)
    └── Continuous - Any value, including decimals (e.g., weight: 65.4 kg, height: 170.2 cm)

How to Present Data

Pie chart example: 60% of patients are male, 40% female - shown as slices of a circle. Histogram example: Distribution of patient ages - shows if data is normally distributed.

---

📌 CHAPTER 5: Summarization of Data

Measures of Central Tendency (Where is the "middle"?)

Measures of Variation (How spread out is the data?)

Key insight: Small SD = data points are close together. Large SD = data is spread widely.

Standard Error & Confidence Interval

• Sampling Variation - Every sample you take will give slightly different results - this is natural.
• Standard Error (SE) - Estimates how much your sample mean might differ from the true population mean.
• Confidence Interval (CI) - A range within which the true value likely falls.

Example: "The mean blood pressure is 130 mmHg (95% CI: 125-135)" means you are 95% sure the true population mean is between 125-135 mmHg.

---

📌 CHAPTER 6: Estimation & Hypothesis Testing

Hypothesis

• Null Hypothesis (H₀) = "There is NO difference/relationship." (e.g., "Drug A and Drug B have the same effect")
• Alternate Hypothesis (H₁) = "There IS a difference/relationship." (e.g., "Drug A is better than Drug B")

P-Value

• P-value = Probability that your result occurred by chance if H₀ were true.
• If p ≤ 0.05 → Result is statistically significant → Reject H₀ (accept H₁)
• If p > 0.05 → Do NOT reject H₀

Example: p = 0.02 means there's only a 2% chance your result was due to chance - so your finding is likely real!

Errors in Hypothesis Testing

• Power = Probability of correctly detecting a real difference = 1 - Beta. Higher power = better study.

---

📌 CHAPTER 7: Measures of Disease Frequency

Prevalence vs. Incidence

Memory trick: Prevalence = existing (like a photograph). Incidence = new (like a video recording new events).

---

📌 CHAPTER 8: Measures of Association

Correlation Coefficient (r)

• Measures the relationship between two variables
• Ranges from -1 to +1
• +1 = perfect positive correlation (both go up together)
• -1 = perfect negative correlation (one goes up, other goes down)
• 0 = no relationship

Relative Risk (RR) - Used in Cohort Studies

RR = Incidence in Exposed ÷ Incidence in Unexposed

Example: Smokers develop lung cancer at 15%, non-smokers at 1% → RR = 15/1 = 15 (smokers are 15x more likely to get lung cancer)

Odds Ratio (OR) - Used in Case-Control Studies

OR = (a × d) ÷ (b × c) using a 2×2 table

• Used when you cannot calculate incidence (because you started with diseased people)
• Interpretation same as RR: OR > 1 = increased risk

---

📌 CHAPTER 9: Bias, Confounding & Effect Modification

Bias = Systematic error that distorts your results

Confounding

Classic Example: Studies find coffee drinkers get more lung cancer. But coffee drinkers also tend to smoke more. Smoking is the confounder - not the coffee!

• At design stage: Restriction, Matching, Randomization
• At analysis stage: Stratification, Multivariate Analysis

Effect Modifier

Example: A drug works well in women but not men - gender is an effect modifier.

---

📌 CHAPTER 10: Sample Size Calculation

What you need to calculate sample size:

Key Terms

• Significance Level = Risk of a false positive (usually 5% = 0.05)
• Margin of Error (d) = How much deviation from the true prevalence you can tolerate (e.g., ±5%)
• Power = Ability to detect a real difference (usually 80-95%). More power = larger sample needed.

---

📌 CHAPTER 11: Screening Tests

The 2×2 Table

                 Disease +     Disease -
Test Positive  |    a (TP)   |    b (FP)  |
Test Negative  |    c (FN)   |    d (TN)  |

Example: HIV ELISA has high sensitivity (>99%) - catches almost all HIV+ people. Confirmatory Western Blot has high specificity - rules out false positives.

Memory trick: SN-out (Sensitive test, if Negative → rules OUT disease). SP-in (Specific test, if Positive → rules IN disease).

---

📌 CHAPTER 12: Questionnaire Development

Types of Questions

Likert Scale

Example: "The doctor explained my condition clearly." Strongly Agree | Agree | Neutral | Disagree | Strongly Disagree

---

📌 CHAPTER 13: Bioethics

The 4 Core Principles

Ethical Considerations in Research

• Informed Consent - Participants must know the purpose, risks, and benefits before agreeing
• No Plagiarism - Give credit, cite sources properly
• Conflict of Interest - Disclose any financial or personal ties to the research
• Data Confidentiality - Keep participant data secure
• Animal Welfare - Follow ethical guidelines when using animals

Consequences of Unethical Research

• Retraction of published papers
• Loss of credibility and public trust
• Legal action and professional penalties
• Inability to replicate findings

---

📌 CHAPTER 14: Publishing Original Research

IMRaD Structure (The Standard Format)

Key Qualities of a Good Research Article

• Novelty - New findings, not published before
• Scientific Rigor - Clear hypothesis, valid methods
• Ethical Approval - Mandatory for human/animal studies (ERC/IRB)
• Reproducibility - Methods described in enough detail for others to replicate

Reporting Guidelines by Study Design

Publishing Process (Step-by-Step)

1. Develop a clear research question
2. Get Ethics Committee (ERC/IRB) approval
3. Register the trial (if it's a clinical trial)
4. Follow the relevant guideline (CONSORT, STROBE, etc.)
5. Get synopsis approved by REU (CPSP)
6. Conduct the study
7. Write with clarity and scientific integrity
8. Submit to a peer-reviewed, indexed journal (as per HEC/CPSP criteria)

---

🧠 Quick Summary Table