Skip to main content

Hypothesis Testing — Previous Year Questions

📊 Topic Overview​

MetricValue
Papers Appeared20 of 27 (74.1%)
Priority🔴 VERY HIGH
Section DistributionSec A: ~10
Most Repeated"Explain hypothesis testing. Describe errors in hypothesis testing."
ProbabilityVery High — prepare full steps + errors

🎯 Must-Know Concepts​

  1. Definition of hypothesis and types (null, alternative, directional, non-directional)
  2. Steps in hypothesis testing (7–8 steps)
  3. Null hypothesis (H₀) and Alternative hypothesis (H₁)
  4. One-tailed vs two-tailed tests
  5. Level of significance (α = 0.05 or 0.01)
  6. Critical region and critical values
  7. Type I and Type II errors

🔥 Most Repeated Question: Hypothesis Testing​

Common phrasings:

  • "What is inferential statistics? Discuss in detail hypothesis testing."
  • "Explain hypothesis testing. Describe errors in hypothesis testing."
  • "Discuss the steps involved in setting up the level of significance."
  • "Define hypothesis testing. Discuss general procedure for testing a hypothesis."
  • "Elucidate hypothesis testing."

Answer (Section A — ~450 words):​

Hypothesis Testing is a statistical procedure for making rational decisions about whether a sample outcome provides sufficient evidence to reject a stated null hypothesis about the population. It forms the cornerstone of inferential statistics.

What is a Hypothesis? A hypothesis is a tentative statement about the relationship between variables or the characteristics of a population. In statistical testing, two complementary hypotheses are stated:

Null Hypothesis (H₀): States that there is no significant effect, no difference, or no relationship. It is the hypothesis that the researcher tries to reject. Example: "There is no significant difference in anxiety scores between group A and group B."

Alternative Hypothesis (H₁): States that a significant effect, difference, or relationship exists. It is what the researcher hopes to support. Example: "Group A has significantly higher anxiety scores than group B."

Steps in Hypothesis Testing:

  1. State the null and alternative hypotheses: Clearly define H₀ and H₁ in testable terms. Decide if the test is one-tailed (directional) or two-tailed (non-directional).

  2. Choose the appropriate statistical test: Based on scale of measurement, sample size, and research design (e.g., t-test for comparing two means, ANOVA for three or more means).

  3. Set the level of significance (α): Typically α = 0.05 (5%) or α = 0.01 (1%). This is the probability of making a Type I error (rejecting a true null hypothesis).

  4. Determine the critical region: Using the chosen significance level and degrees of freedom, find the critical value from the appropriate statistical table. The critical region is the area of the sampling distribution that leads to rejection of H₀.

  5. Collect data and compute the test statistic: Compute the appropriate statistic (t, F, χ², z) from sample data.

  6. Make a decision: Compare the obtained statistic with the critical value.

    • If obtained value ≥ critical value: Reject H₀ (result is statistically significant)
    • If obtained value < critical value: Retain H₀ (result is not significant)

  7. State the conclusion: Interpret the statistical decision in terms of the original research question.

One-tailed vs Two-tailed Tests: A one-tailed test is used when the direction of the effect is predicted (e.g., Group A scores will be higher than Group B). The entire critical region falls in one tail. A two-tailed test is used when no direction is predicted; the critical region is split equally between both tails, making it more conservative.

Errors in Hypothesis Testing:

  • Type I Error (α error): Rejecting H₀ when it is actually true. The probability of this error equals the significance level (α).
  • Type II Error (β error): Retaining H₀ when it is actually false. The probability is denoted β; the power of the test = 1 − β.

The two errors are inversely related: reducing Type I error (using a stricter α) increases the risk of Type II error.

📝 Other Key Questions​

One-tailed and two-tailed tests and their applications (Section B — ~250 words)​

One-tailed test: Used when the researcher has a specific directional prediction based on prior theory or evidence. The critical region is entirely in one tail (left or right). Advantage: higher statistical power for detecting the predicted direction. Risk: If the effect is in the opposite direction, it will be missed entirely.

Two-tailed test: Used when the research question is non-directional ("Is there a difference?"). The critical region is divided equally between both tails (α/2 in each). More conservative but safer for exploratory research.

Decision rule: One-tailed test at α = 0.05 uses the same critical value as a two-tailed test at α = 0.10, making the one-tailed test more powerful (easier to reach significance) when the prediction is correct.

🧠 Memory Aid​

Steps in Hypothesis Testing (SCSCDM-C):

  • State H₀ and H₁
  • Choose statistical test
  • Set level of significance (α)
  • Critical region identification
  • Data collection & compute statistic
  • Make a decision
  • Conclude

📚 Related Content​

✅ Self-Assessment​

  1. What is the null hypothesis? Give an example.
  2. List the 7 steps in hypothesis testing.
  3. What does the level of significance (α = 0.05) mean?
  4. What is the difference between a one-tailed and a two-tailed test?
  5. Explain the inverse relationship between Type I and Type II errors.

💡 Exam Tips​

  • Always state H₀ and H₁ before describing the steps — this earns marks immediately.
  • The 7 steps must be listed and briefly explained, not just listed.
  • Type I and Type II errors are frequently asked within this topic — prepare the decision matrix.
  • Connect hypothesis testing to inferential statistics in your introduction.