How BchiroMat works

1. Selecting the statements

The questionnaire contains 30 statements in 7 display topics. Items are selected for policy relevance in Israel, party differentiation, clarity, and public evidence. A weak or duplicative item is archived or returned for further research.

Display topics, scoring groups, priority topics, and questionnaire order are now stored as separate versioned maps. Their memberships remain identical to v37 in the current release, so this architectural split does not change results.

2. How a party position is coded

For each statement, the research asks whether the party agrees, disagrees, is neutral, is internally divided, or has no clear public position. A position is never filled in merely from a left–right label.

1. 1A current official party platform or policy programme
2. 2A recorded vote, legislation, official transcript, or documented governing action
3. 3A direct statement by party leadership or an attributable interview
4. 4Reliable reporting of a clear party commitment

Evidence strength describes how strong and current the support is, not how intensely the party holds the view. It does not pull a party toward the midpoint in the primary score.

3. Exact numerical mapping

User answers are mapped to a −1 to 1 scale. Party positions remain three-level so the data does not imply unsupported precision.

Important: “Neither agree nor disagree” is the midpoint (0) and is scored. Choose “Don’t know / skip” if you do not know or do not want the statement to affect the result at all.

4. Topic scoring and base weights

1. Within each scoring group, points are averaged only across answered statements for which the party has a scoreable unified position.
2. A group’s effective weight is its base weight, multiplied by 2 if selected as a priority, and multiplied by that party’s comparable-position coverage in the group.
3. The overall match is the weighted mean of the available scoring groups and is rounded to a whole percentage.

Therefore x2 means twice that topic’s disclosed base weight, not twice every other topic. Base weights are a versioned editorial choice; changing them requires a sensitivity test and a new method version.

5. Missing data and the percentage denominator

A user skip, no documented party position, and documented internal disagreement receive no points and are not treated as agreement, disagreement, or neutrality. Missing positions are neither ideologically imputed nor assigned an arbitrary penalty.

This means each party may have a different denominator: the percentage is agreement between your answers and the documented positions that could be compared for that party. Topic coverage reduces that topic’s weight; it does not make percentages based on different evidence sets statistically identical.

A result with overall comparable coverage below 65% is marked as having less available information. This is a disclosure threshold, not a confidence interval.

6. Ties and display order

Parties are first ordered by the rounded match percentage, highest first. A tie is broken by rounded overall evidence coverage, highest first. If that is also equal, the Hebrew display name is sorted alphabetically. The final rule exists only for stable display and does not imply a substantive difference in fit.