Methodology v0.2 · Research draft
A compass you can check.
The questions
14 statements cover housing permission, housing delivery, streets, policing and drug policy. The two-dimensional graphic describes housing only. The other issues stay visible in the answer comparison and do not move that dot.
Five choices, plus uncertainty
Strongly disagree = −2, disagree = −1, neither = 0, agree = +1 and strongly agree = +2. Not sure, skips and missing sources have no numeric value. We use a modest agree/disagree placement for a clear direction; the extreme categories require explicit categorical support or a direct response at that strength. Silence never means neutrality.
What a party placement means
Every placement is an editorial interpretation proposed from the linked source, not a response received from the party. Dates, qualifications and reasoning appear next to it. The editor has authorised documented inference, which does not mean each code has been individually reviewed or confirmed by a party. Estimated rows show low or medium editorial confidence. These labels are judgments about evidence, not statistical probabilities. Named candidate statements may inform a party estimate when explicitly labelled as a proxy; they do not claim agreement by every candidate. Turn estimates off on the results page to see how much the comparison depends on them. Historical platforms remain background unless reaffirmed; current campaign proposals issued before a mayoral withdrawal carry that qualification. Where a broader proposal reasonably suggests a direction, we may estimate agree or disagree and explain the missing conditions. We do not infer a budget growth rate from calls for adequate funding, fill silence with neutrality, or assign extreme answers to estimates. A candidate’s own statement does not automatically become a party position.
Your housing point
Each axis averages the four relevant answers, accounting for reverse wording, and multiplies by 50. Three substantive answers per axis are required. A central dot can reflect mixed views. Equal spacing between answers is a modelling assumption; this is not a scientifically validated scale.
Party dots and match scores
Select a party on the results page to compare housing dots when there are at least three shared sourced answers on each axis. Both dots use that same question set; your comparison dot can differ from your full result. Missing evidence never becomes a central coordinate. Agreement percentages average individual shared answers across all five topics, with at least two shared answers required. Identical answers score 100%; each step apart subtracts 25 percentage points. Each question has equal weight, including estimates. Turning estimates off recomputes scores and dots. Low confidence estimates can materially change a result; the confidence count shows that dependence. Coverage and the included questions appear beside each percentage. These are limited-evidence draft comparisons, not overall platform matches. Different parties can cover different questions, so percentages should not be used as a ranking. Dots describe housing only; percentages also include other answered topics.
Privacy
Your answers stay in this tab’s session storage so a refresh does not erase them. Use Clear & start again to remove them. The site sends no answers to its server or newsletter provider. Optional browser assistants can read the answers when you use those tools. Browser session restoration may restore tab data.
Corrections and updates
Write to hello@sidestreetvancouver.ca with the question, party, source and proposed correction. Publication requires a source check; a party dispute should be recorded. Sources are checked on the displayed date, not continuously monitored. Side Street is independent of CBC and its Vote Compass.