University of Birmingham · FACULTY OF PSYCHOLOGY

13344 Chap.8 Randomised Controlled Trials: Design Principles

- one subject, every graph, every model, every mark
6 Chapters4-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 8 of 10 · 13344

Randomised Controlled Trials: Design Principles

Why randomise

The chapter explains how random allocation evens out measured and unmeasured characteristics across arms, why the teaching calls it the most powerful element of a trial, and how allocation concealment differs from blinding.

Running a fair trial

It compares simple, block, stratified and minimisation methods, diagnoses a flawed trial, explains blinding levels and intention to treat, and covers feasibility, acceptability, equipoise and eligibility.

Using it in the proposal

The research proposal options include a randomised trial, and the seminar headings ask for randomisation, randomisation concealment, blinding and outcome measures.

The chapter's counter-example shows how to diagnose each of these in a flawed design and fix it.

In this chapter

What this chapter covers

  • 01

    Rationale for randomisation

  • 02

    Allocation concealment and blinding

  • 03

    Simple, block, stratified and minimisation

  • 04

    Intention to treat and per protocol

  • 05

    Follow-up and differential dropout

  • 06

    Feasibility, acceptability and equipoise

Worked example · free

Concealment or blinding: which protection failed

Q [3 marks]. AskSia-authored practice. Classify each problem: (a) a recruiter can see the next allocation on a printed list; (b) outcome raters know which arm a participant is in; (c) the allocation sequence is generated by a computer but posted on the clinic wall. The marks shown are an AskSia study allocation, not a University marking scheme.
Two protections at two different moments.Two protections at two different moments. The boxes run from screening to outcome measurement; allocation concealment guards everything up to the moment of allocation, shown by the left band, while blinding guards delivery and measurement afterwards, shown by the right band. Screen Consent Allocate Treat Measure concealment protects blinding protects random sequence revealed only here
Two protections at two different moments. The boxes run from screening to outcome measurement; allocation concealment guards everything up to the moment of allocation, shown by the left band, while blinding guards delivery and measurement afterwards, shown by the right band.
  • 1(a) Allocation concealment has failed, because the recruiter knows the next allocation before enrolment.
  • 1(b) Blinding has failed, because outcome assessment happens after allocation with knowledge of the arm.
  • 1(c) Concealment has failed: the sequence is random but visible before allocation.
Problems before allocation are failures of concealment, which invite selection bias; problems after allocation are failures of blinding, which invite performance and ascertainment bias. A random sequence that anyone can see is still unconcealed. In a proposal, name who generates the sequence, who holds it and who is blinded, so a reader can see that each protection is in place.
Sia tip — Ask whether the leak happens before or after allocation.
Glossary

Key terms

Block Randomisation
Block Randomisation is allocation that keeps the arms balanced within each block of participants.
Equipoise
Equipoise is genuine uncertainty about which trial arm is better, without which a trial is unethical.
Per Protocol Analysis
Per Protocol Analysis includes only participants who adhered to their allocated treatment.
Pragmatic Trial
A Pragmatic Trial tests whether an intervention works in real-world settings, with less selection and more flexible delivery.
Stratified Randomisation
Stratified Randomisation is randomising separately within groups defined by a prognostic factor, such as sex or severity.
FAQ

Randomised Controlled Trials: Design Principles FAQ

Why are randomised trials called the gold standard?

Randomisation evens out characteristics between the groups, including ones nobody measured, so a difference in outcomes can be attributed to the intervention. The teaching says this is not possible with any other design, which is why well-conducted trials sit at the top of the evidence hierarchy.

What is the difference between allocation concealment and blinding?

Concealment keeps the next allocation hidden until it is assigned and prevents selection bias; it can always be achieved. Blinding hides the allocation afterwards from participants, staff or assessors and prevents performance and ascertainment bias, but it cannot always be achieved, especially for psychotherapies.

Why analyse by intention to treat?

Analysing everyone in the arm they were randomised to preserves the balance randomisation created and gives an unbiased assessment of the treatment effect. Leaving out people who stopped treatment usually removes those doing worst and exaggerates the benefit.

What is a feasibility study for?

Trials need large samples and are expensive, so a smaller feasibility study first asks whether the design can be delivered, whether enough people can be recruited and retained, and whether the primary outcome is appropriate. Acceptability asks whether the design suits participants. Progression criteria then decide whether to go ahead.

What is the difference between efficacy and effectiveness?

Efficacy is whether an intervention works in the ideal conditions of an explanatory trial. Effectiveness is whether it works in real-world settings, which pragmatic trials test with less selection and more flexible delivery; they tend to be larger and find smaller effects.

How does minimisation allocate trial participants?

It is an allocation method in which the first participant is allocated at random and each later one is allocated to whichever arm reduces imbalance on chosen factors, usually with a random element. CONSORT describes it as the only acceptable alternative to randomisation, and it suits small trials with several important prognostic factors.

Study strategy

Assessment move

Take one published trial and write a single line on how it randomised, concealed, blinded and analysed participants. Then design a flawed version and fix each flaw, as in the chapter's counter-example. When you plan a trial for the proposal, write the randomisation, concealment and blinding sentences first, because each one corresponds to a heading the seminar suggests.

Working through Randomised Controlled Trials: Design Principles in 13344? Sia is AskSia’s AI Psychology tutor — ask any 13344 Randomised Controlled Trials: Design Principles question and get a clear, step-by-step explanation grounded in how 13344 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.

A+Everything unlocked
Unlocks this Bible + all 17 of your subjects at University of Birmingham, plus 1,000+ more Bibles across the AskSia library.
Sia - your 13344 tutor, unlimited, worked the way the exam marks it
The full 4-page Bible + practice bank with worked solutions
Chrome extension - sync your LMS so Sia knows your deadlines
Bilingual EN / Chinese on every Bible and every Sia answer
$0.99 Trial
30-day money-back · cancel in one tap · how it works