Find My Match · Methodology

How the quiz works

This page explains how we build the quiz questions, how your answers are matched to parties and candidates, and what the results can — and cannot — tell you. We publish this so you can scrutinise and, if you disagree, challenge our approach.

The quiz is a voter-education tool, not a voting recommendation. Your results show which parties and candidates share your stated views on the issues you care about. How you vote remains entirely your decision.

1 What is the Find My Match quiz?

The Find My Match quiz is designed to help voters — especially those voting for the first time — discover which parties and candidates most closely match their views on the issues they personally care about most.

Two principles guide every design choice we've made:

Principle 1
Keep it short

Most people won't complete a 30-question quiz. Ours targets 8–15 questions, finishable in under 4 minutes — covering only the topics you selected.

Principle 2
Be honest about uncertainty

A confident-looking result based on weak data is worse than an honest partial answer. We show what we know, label what we don't, and don't make claims our sources can't support.

2 How the quiz is structured

Step-by-step flow

Quiz flow
Step 1
Pick your topics
Choose 4–5 policy areas
Step 2
Answer questions
2–3 per topic (8–15 total)
Step 3
Weight topics
Optional: flag priorities
Step 4
See your matches
Party & candidate scores

Question types

Every question is directly linked to a real manifesto pledge or public statement. There are two question formats:

Type A — Agreement scale
"The government should introduce a cap on annual rent increases"
Strongly support
Support
Neutral
Oppose
Strongly oppose

Used when one policy pledge clearly defines a topic.

Type B — Trade-off choice
"On energy bills, which comes closer to your view?"
Keep current subsidies — prioritise stability for households
Phase out subsidies — invest savings in structural reform
Neither / not sure

Used when the campaign is being fought on a genuine trade-off.

You can skip any question Skipped questions are excluded from your match score entirely — they don't hurt you, and they're noted in your results so you can see exactly what fed into your match.

Importance weighting (optional)

After answering all questions, you can flag any of your chosen topics as "particularly important to me". Questions in flagged topics carry double weight in your match score. This step is optional — evidence from comparable tools suggests fewer than 30% of users use it, and making it compulsory reduces completion rates.

3 How your match is calculated

Your answers and each party's stated positions are converted to numbers on the same scale. The closer those numbers are, the higher the match.

The match formula — step by step

Step 1 — Convert answers to numbers

Strongly support +2
Support +1
Neutral / not sure 0
Oppose −1
Strongly oppose −2

Step 2 — Measure the gap on each topic

Per-topic alignment = how close your number is to the party's number. The maximum gap is 4 (opposite ends of the scale).

0% Opposite
25% 2 steps apart
50% 1 step apart
75% Near match
100% Identical

Step 3 — Average across all topics

Match score  =  Sum of alignment scores (×2 if topic flagged as important) Sum of topic weights

Expressed as a percentage. Topics where the party has no stated position are excluded from both sides of the fraction.

Candidate scores

Candidates who have completed our structured survey are matched at the question level — your quiz question is compared directly to the candidate's answer to that same question, giving the most precise match possible. Candidates who have not submitted personal positions inherit their party's score, clearly labelled as "Based on party". Independents with no stated positions appear in the candidate list but are not given a match score.

What the score doesn't mean: A match percentage does not predict how a candidate will vote once elected, account for coalition dynamics, or evaluate whether a policy is deliverable or well-costed. It measures stated position alignment only.

4 Where the data comes from

Tier 1
AI-assisted research
Party manifestos, press releases, election speeches, and reporting in mainstream Maltese media (MaltaToday, The Malta Independent, LovinMalta). Each position is logged with its source and a confidence level.
Tier 2
Candidate-submitted
A structured survey sent directly to every candidate. Responses are labelled "Submitted by candidate" throughout the site. These are the positions we can present with the highest confidence.

Confidence labels

Every Tier 1 position carries one of two confidence labels:

Stated
A direct quote or unambiguous policy statement from a primary source. We can point to exactly where it was said.
Inferred
Consistent with the party's broader narrative across multiple sources, but not a single direct statement. These are marked with a ? icon in the results.
🔒
Your quiz answers never leave your device.

Everything is stored only in your browser's local storage. We have no server-side record of your quiz responses. You can verify this by opening your browser's network tab while taking the quiz — no quiz data is sent anywhere.

5 Known limitations

We document these openly rather than wait to be asked about them.

  • 👤
    Most candidates don't have their own positions on record Detailed per-candidate positions exist mainly for party leaders. Until candidates submit their own responses to our survey, most candidate results are based on their party's positions.
  • 📋
    The topic list is editorially curated We chose the topics based on what's being debated in the 2026 election. This judgement may not match yours. The topic picker lets you decide which topics matter most to you.
  • 📊
    Stances are simplified Real policy positions are complex. A 5-point scale can't capture every nuance — two parties may both support rent relief but through completely different mechanisms. We provide a drill-down view so you can read the detail behind each score.
  • 🏝️
    Independents are weakly represented Independent candidates typically have no manifesto and no party to inherit from. They appear in the candidate list but cannot be matched without their own survey submissions.
  • 🧭
    A high match on narrow topics doesn't reflect a party's full ideology The quiz measures alignment on the topics you chose. If you selected a narrow set — for example, just migration or just identity — you may score a high match with a party whose broader ideology, on topics you didn't select, sits very differently from your own values. To address this, every party's results card shows its ideological tradition (sourced from the party's own materials and European political family affiliation), and every party profile includes a full ideological context section. We recommend reading a party's full profile before drawing broader conclusions from a match score.

6 Common questions

Question Our answer
Is the quiz biased — does it favour a particular party? The score formula and all source citations are public. You can run the formula yourself against our published dataset. We welcome corrections.
Why are candidates scored on their party's positions rather than their own? Because most candidates don't have per-candidate positions on record. We label every such case clearly. The alternative — excluding 95% of candidates from matching — would leave you with much less useful results.
Don't match percentages overstate precision? We show percentages because users expect them. But we show the topic-by-topic breakdown on the same screen — we don't present a single number without context.
Do my quiz answers go to a server? No. Your answers are stored only in your browser. No quiz session exists on our servers. You can verify this in your browser's network tab.
Can I challenge or correct a position you've attributed to a party? Yes. Every position has a source citation. If you believe we have mischaracterised a position, contact us and we'll review it.

7 Keeping it up to date

Questions
Locked at launch (~25 May 2026). Changed only for material errors. All changes are logged publicly.
Party positions
Refreshed weekly during the campaign, and immediately after major manifesto launches or policy announcements.
Candidate surveys
Accepted and incorporated continuously through to two days before election day.

This methodology document is versioned. If we make a material change to the formula, quiz structure, or scoring approach, we publish a new version number and log the change.

Are you a candidate, journalist, or researcher? If you want to review the underlying data, challenge a position, or propose a question for consideration, get in touch. All source citations are public and the full dataset will be published after the election.