Research methods exam tips

Exam tips

• Exclude ties when calculating the sign test

  Remove any pairs with a difference of zero before counting positives and negatives.

  Ties do not contribute to the direction of the median difference and would otherwise inflate the sample size, leading to an incorrect binomial calculation.

• Use statistical tables efficiently

  Memorise the most common critical values for chi‑square, t, and z tests, and practise quick lookup of degrees of freedom.

  This saves time during exams and reduces the chance of calculation errors.

• Balance significance level and power

  When answering questions about error types, explicitly state how the chosen α level and the study’s power influence the likelihood of Type I and Type II errors.

  Demonstrating this relationship shows a deeper understanding of inferential testing and satisfies higher‑level assessment criteria.

• Check the variable type before choosing a test

  Create a quick table: Variable type → Appropriate test (e.g., nominal → chi‑square; interval/ratio → t‑test/ANOVA).

  This ensures you select the correct inferential test and avoids common pitfalls in exam answers.

• Structure your answer clearly

  Start by stating the design, then the type of data, followed by the assumptions, and finally name the appropriate test.

  Examiners look for a logical flow that demonstrates you understand how design influences statistical choice.

• Choosing the right test quickly

  Make a quick checklist: Are variables continuous? Are groups independent or related? Is data normal? Then pick the appropriate test.

  This systematic approach reduces errors and saves time during the exam.