University of Birmingham · FACULTY OF PSYCHOLOGY

13344 Chap.4 Cross-Sectional Studies and Surveys

- one subject, every graph, every model, every mark
6 Chapters3-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 4 of 10 · 13344

Cross-Sectional Studies and Surveys

Prevalence at one moment

The chapter explains what a cross-sectional survey measures, how a sampling frame and eligibility criteria are set, and the four main sampling strategies, from simple random to cluster sampling.

Errors and analysis

It sets out the warnings the teaching gives about direction, response rates below about 80% and rare disorders, works through weighting a stratified sample back to its population, and summarises how surveys are analysed and judged.

Using surveys in the essay

A cross-sectional survey is one of the designs most often chosen for a title about associations.

The chapter ends by setting out its strengths, from speed and low cost to usefulness for planning services, and its limits, from snapshot timing to non-response, so you can weigh it fairly.

In this chapter

What this chapter covers

  • 01

    Prevalence and its uses

  • 02

    Sampling frames and eligibility

  • 03

    Simple, stratified and cluster sampling

  • 04

    Response rates and selection bias

  • 05

    Two-phase surveys

  • 06

    Weighting and confidence intervals

Worked example · free

Sampling fractions and weights for an over-sampled borough

Q [4 marks]. AskSia-authored practice. A borough has 6,000 adults aged 65 to 74, 3,000 aged 75 to 84 and 1,000 aged 85 and over. A survey samples 300, 300 and 250 from these groups. Find each sampling fraction and the weight each respondent carries. The marks shown are an AskSia study allocation, not a University marking scheme.
Over-sampling a small group on purpose.Over-sampling a small group on purpose. The top bar shows the age make-up of an invented borough of 10,000 older adults; the bottom bar shows a stratified sample of 850 that deliberately includes more of the oldest group, which is why its results must be weighted back. Borough Sample 65 to 74: 60% 75 to 84: 30% 85+: 10% 35% 35% 85+: 29% oldest group over-sampled, then weighted back
Over-sampling a small group on purpose. The top bar shows the age make-up of an invented borough of 10,000 older adults; the bottom bar shows a stratified sample of 850 that deliberately includes more of the oldest group, which is why its results must be weighted back.
  • 1Aged 65 to 74: 300 ÷ 6,000 = 1 in 20, so each respondent carries a weight of 20.
  • 1Aged 75 to 84: 300 ÷ 3,000 = 1 in 10, a weight of 10.
  • 1Aged 85 and over: 250 ÷ 1,000 = 1 in 4, a weight of 4.
  • 1Check: 300 × 20 + 300 × 10 + 250 × 4 = 6,000 + 3,000 + 1,000 = 10,000.
The oldest group is sampled five times as intensively as the youngest, so its respondents carry the smallest weight. Applying the weights restores the borough's 60%, 30% and 10% age make-up before prevalence is reported. Reporting the sampling fractions alongside the weights lets a reader check that the oldest group was over-sampled deliberately rather than by accident.
Sia tip — Check that weighted counts add back to the population total before using them.
Glossary

Key terms

Sampling Frame
A Sampling Frame is the list or definition from which a survey sample is drawn, and it must represent the base population.
Cluster Sampling
Cluster Sampling is randomly selecting groups such as schools, then units within them.
Period Prevalence
Period Prevalence is the proportion with a condition at any time during a stated period, such as twelve months.
Two-Phase Survey
A Two-Phase Survey screens everyone briefly and then gives likely cases a full clinical assessment.
Response Rate
A Response Rate is the share of sampled people who take part, and below about 80% it reduces a survey's validity.
FAQ

Cross-Sectional Studies and Surveys FAQ

Can a cross-sectional survey show that one thing causes another?

No. Exposure and outcome are measured at the same time, so the design can show an association but not which came first. The teaching's example is that poor sleep and depression measured together cannot tell you which caused the other, or whether something else caused both.

Why does a low response rate matter if the sample is large?

The teaching warns that a response rate below about 80% reduces validity because responders may differ systematically from non-responders, possibly in their exposure or outcome. A larger sample does not fix that difference.

Why are survey results weighted?

When strata are sampled at different fractions, for example to include enough of a small group, the raw sample no longer matches the population. Weighting each respondent by the inverse of their sampling fraction restores the population's make-up before prevalence is reported.

What is the difference between point, period and lifetime prevalence?

Point prevalence is the proportion with a condition when data are collected. Because conditions relapse and remit, surveys sometimes report period prevalence over a stated time such as twelve months, and some assessments ask about lifetime prevalence. Say which one a study reports before comparing it with another.

Why are cross-sectional surveys inefficient for rare disorders?

Most respondents will not have the condition, so very large samples are needed to find enough cases, and a screen will produce many false positives. The teaching recommends a two-phase design for rarer conditions, or a different design altogether.

What is a sampling frame and why does it matter?

It is the list or definition from which a sample is drawn, bounded by eligibility criteria such as age, residence and time period. If it does not represent the base population, the survey risks selection bias however carefully the sample is then drawn, so define it before recruiting anyone.

Study strategy

Assessment move

For each seminar sampling scenario, write the sampling frame, the strata you would use and the most likely source of selection bias, then compare your answers with a classmate's. Practise one weighting calculation with your own strata and sampling fractions, checking that the weighted counts add back to the population.

Working through Cross-Sectional Studies and Surveys in 13344? Sia is AskSia’s AI Psychology tutor — ask any 13344 Cross-Sectional Studies and Surveys 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 3-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