Skip to content

GUIDES

The craft of surveys that produce trustworthy data.

Short and practical, written from the research literature and from building the tool. No tool required to use any of it, though every technique here is one toggle in Plumeform.

8 MIN READ

Choosing a survey tool for Prolific studies

The seven capabilities that decide your fielding week, and an honest comparison of Qualtrics, Typeform, Google Forms, and Plumeform against them.

6 MIN READ

Prolific completion codes, set up right

The three URLs of a panel study, why you redirect AND show the code, what screened-out participants need, and the two-minute pre-launch test.

7 MIN READ

How to stop duplicate survey participants

Where duplicates really come from, why cookies aren't a control, and the layered defense that keeps a dataset one-person-one-row.

8 MIN READ

Screening participants without burning money

Unguessable eligibility questions, quotas that close themselves, and screen-out etiquette that keeps your panel reputation intact.

7 MIN READ

Random assignment in survey experiments

Running a real A/B experiment inside a survey: server-side assignment, condition-driven branching, and the audit trail your analysis needs.

7 MIN READ

Informed consent for online surveys

What the consent screen must communicate, a copy-paste template, honest withdrawal language for anonymous data, and a gate that records nothing on decline.

7 MIN READ

How many survey responses do you need?

The margin-of-error math without the jargon, the sample sizes worth memorizing, and why your response rate decides more than your audience size.

9 MIN READ

How to catch low-quality survey responses

Speeders, straight-liners, bots, and duplicates each leave a signature. How to detect them, design them out, and clean without fooling yourself.

8 MIN READ

How to write attention check questions

The four types of checks, the writing rules that catch careless responders without punishing honest ones, and what to do with failures.

7 MIN READ

Survey randomization, explained

What order effects do to your data, what to shuffle and what to pin, and the one method sentence that satisfies reviewers.

9 MIN READ

Likert scale best practices

Five or seven points, midpoint or not, labels, item wording, matrix batteries people finish, and what the numbers honestly support.