Import the replicate-weight design, not just a collection of weight columns
A plausible point estimate can coexist with a misinterpreted variance-weight package.
Read the lesson →THE CREATIVE FIELDBOOK / SNS MEDIA TECH
Practical lessons
1095 source-led reads, 6 per page. These numbered pages work without JavaScript. Search every study and lesson →
Investigate a one-PSU stratum before silencing its variance warning
A solitary primary sampling unit can reflect different design situations with different consequences.
Read the lesson →Make a weighted survey percentile reproducible across analytical tools
Two valid percentile conventions can produce different values from the same weighted responses.
Read the lesson →Keep population totals distinct from normalized survey weights
A rescaled weight can preserve a weighted mean while changing the meaning of a weighted total.
Read the lesson →Preserve the first-phase sample when selecting detailed follow-up measurements
A detailed follow-up sample needs a selection contract, not just a list of interesting initial responses.
Read the lesson →Reconcile every selected case before naming a survey response rate
A completion percentage among starters is not the response rate for the selected sample.
Read the lesson →