Question 1
By what paths do households say "Organisation 1"'s activities led to changes in household income and production and in food and diet, and how many households, out of the 15 whose interviews name "Organisation 1", have such a path that passes through better household health? The answer builds on the earlier answer about what influences brought about changes in household health (8 October 2026, latest revision 10 October 2026), taking its classification of which programme organisation each interview names and its coded health passages and links as they stand. Descriptive, with no rubric; read in the manner of QuIP causal coding. Cases are households (one interview document each), counted out of the 15 that name "Organisation 1", with all 18 also shown; a household can count under several paths and sources. A path starts where a passage itself names "Organisation 1" as the source. Paths are joined from one household's own links across its passages, never across households, and are reported beside the chains a respondent tells whole in one passage. A path passes through better health only where the same household has both a link into household health and a link from household health to income, production, food or diet. Weak passages are counted apart from firm ones and never added to them. Answers to closed questions are context, never evidence of a cause. Numbered organisations are anonymised names and are written in quotation marks.

Question 2
What do households say changed their overall wellbeing, through which paths, and what part do relationships within the family (between spouses, and between generations) play along those paths: as a cause, as something that changed, or both? The answer builds on the answer about paths from "Organisation 1" to household income and food (13 October 2026), taking its coding of what changed income, food, family relationships and wellbeing, and on the earlier answer about influences on household health (8 October 2026, latest revision 13 October 2026), taking its coded health links and its classification of which programme organisation each interview names, all as they stand. Descriptive, with no rubric; read in the manner of QuIP causal coding. Cases are households (one interview document each), counted out of all 18, with the 15 whose interviews name "Organisation 1" shown; a household can count under several paths. A change in wellbeing is any link into wellbeing, for better or worse, with changes for the worse reported apart. Paths are joined from one household's own links across its passages, never across households, and are reported beside the chains a respondent tells whole. Family relationships count as both a cause and something that changed only where the same household has a link into family relationships and a link from them on a path to wellbeing. Each passage about family relationships is also coded for whether it concerns spouses or co-wives, generations, both, or the household without saying. Weak passages are counted apart from firm ones and never added to them. Answers to closed questions are context, never evidence of a cause. Numbered organisations are anonymised names and are written in quotation marks.

Question 3
For each kind of change households report (health, food, income and production, family relationships, overall wellbeing), whom or what do they credit: "Organisation 1", other organisations and services ("Organisation 2", health services and other named organisations), government, their own effort, family members, the community, or weather and markets? Where a household credits "Organisation 1" for a kind of change, does it also credit anything else for the same change? The answer builds on the answer about paths from "Organisation 1" to household income and food (13 October 2026, revised the same day), taking its coding of what changed income, food, family relationships and wellbeing, and on the earlier answer about influences on household health (8 October 2026, latest revision 13 October 2026), taking its coded health links and its classification of which programme organisation each interview names, all as they stand. Descriptive, with no rubric; read in the manner of QuIP causal coding. Cases are households (one interview document each), counted out of all 18; a household can count under several sources. A household credits a source for a kind of change where its own links make a path from that source to the change, directly or through a practice such as farming or hygiene, joined within the household only; practices are the routes credit travels along, not sources. Credit for changes for the better and for the worse is counted apart. A household credits something else beside "Organisation 1" for a change where, among the households whose own firm links make a path from "Organisation 1" to that change, another source also has a path to it. Weak passages are counted apart from firm ones and never added to them. Answers to closed questions are context, never evidence of a cause. Numbered organisations are anonymised names and are written in quotation marks.

Question 4
How do the accounts of women and men respondents differ, and those of province A and province B, in which kinds of change they report (health, food, income and production, family relationships, overall wellbeing) and in how often they credit "Organisation 1"? And how far along the chain does each household carry its credit to "Organisation 1": not at all, to a practice only, to health, to income or food, or to wellbeing? The answer builds on the earlier answers about influences on household health (8 October 2026, latest revision 13 October 2026), about paths from "Organisation 1" to income and food (13 October 2026), and about who households credit for each kind of change (13 October 2026), taking their coded links and their household-level record of which changes each household's links carry from "Organisation 1", all as they stand. Descriptive, with no rubric; read in the manner of QuIP causal coding. Cases are households (one interview document each); the respondent's sex and the province are as the list of documents records them. Every count is given beside the size of its group, and no group is compared by share alone. A household reports a kind of change where it has an explicit link into it, for better or worse, counted apart. A household's reach is the highest rung its explicit links reach from "Organisation 1", in the order given; income and food form one rung. Because the three households naming only "Organisation 2" are all women's interviews in province B, credit to "Organisation 1" is also compared among the 15 households naming "Organisation 1". Implicit passages are counted apart from explicit ones and never added to them. Answers to closed questions are context, never evidence of a cause. Numbered organisations are anonymised names and are written in quotation marks.

Question 5
How strongly do the accounts support the claim that "Organisation 1"'s work contributed to better household wellbeing, against this rubric, which the evaluator fixed before the chapter's documents were read?
- Breadth (counted): strong if at least 10 of the households whose interviews name "Organisation 1" credit it by name for some change; moderate 5 to 9; weak under 5.
- Reach along the chain (counted): strong if at least 5 households carry the credit past a practice to income, food or wellbeing; moderate 2 to 4; weak 0 or 1.
- Attribution not crowded out (read from quoted passages): strong where, when "Organisation 1" is credited, other causes named for the same change are minor or complementary; moderate mixed; weak other causes usually named as the main one.
- Harms (counted): strong if no household credits "Organisation 1" with a change for the worse; moderate 1 household; weak 2 or more.
Overall: strong if at least three criteria are strong and none weak; weak if two or more are weak; moderate otherwise.
The answer builds on the earlier answers about influences on household health (8 October 2026, latest revision 13 October 2026), about paths from "Organisation 1" to income and food, about who households credit for each kind of change, and about differences by sex and province (all 13 October 2026), taking their coded links and their household-level records of how far each household's links carry credit from "Organisation 1", all as they stand. Cases are households (one interview document each). Breadth is counted out of the 15 households whose interviews name "Organisation 1"; a household credits it by name for some change where an explicit link runs from "Organisation 1" to a practice or a change. Reach counts households whose explicit links run from "Organisation 1" to better income, food or wellbeing. Harms counts households whose explicit links run from "Organisation 1" to a change for the worse. Attribution is read from the passages naming another cause for health, food, income, family relationships or wellbeing in the households that credit "Organisation 1" for the same change. Implicit passages are counted apart from explicit ones and never added to them. Answers to closed questions are context, never evidence of a cause. Numbered organisations are anonymised names and are written in quotation marks.

Write your answer to answer.md.