Skip to content

Rank plus stars page

Measure preferences with a ranking question and a star rating together on one survey page.

Use this when you want two complementary preference measures in one place: a ranking forces the options apart, and a star rating captures overall liking that a ranking cannot express. Asking only for ratings is the naive alternative, and it routinely fails through ceiling effects, with every option “liked” and nothing separating them; the ranking removes that escape while the stars keep the absolute signal. Pick a likert battery instead when items should be judged independently on a shared scale rather than traded off against each other.

typecontentheadresponseTyperesponseOptionsrequiredpageBreakbutton1
instructionsYou will order some snacks by preference and then rate how much you enjoy snacking overall.NEXT
formAnswer both questions and press FINISH.
formRank these snacks from most preferred at the top to least preferred at the bottom.rankCrisps;Chocolate;Fruit;Nuts;Salted popcorn1
formOverall how much do you enjoy snacking?stars11

The three type=form rows are consecutive, so they render as one survey page. The first of them deliberately has a blank responseType: on a form row any unrecognised value, including a blank cell, renders as display only text with no input, which makes it a free on page intro line above the two questions.

The ranking row sets responseType to rank and lists the items to be ordered in responseOptions, separated by semicolons; the initial order shown is the written order, because responseOptionsRandom is not used. The stars row sets responseType to stars, which renders a five star rating; stars is one of the widgets that ignores responseOptions entirely, so that cell stays genuinely empty.

required set to exactly 1 on both question rows stops the page being submitted unanswered; on form rows only the literal 1 has this effect. pageBreak set to 1 on the stars row ends the page, and since this is the last page of the file its button automatically reads FINISH rather than NEXT.

The usual form trap applies: responseType values are case sensitive, so Rank or Stars silently renders more display only text instead of a question.

  • Counterbalance the starting order of the ranked items by adding responseOptionsRandom with 1 on the rank row; responseCode in the results still counts positions in the order written in responseOptions.
  • Rank a different item set or a longer one by editing the semicolon separated list in responseOptions.
  • Give each question its own screen by moving the pageBreak value 1 onto the rank row as well.
  • Ask for star ratings of several aspects by duplicating the stars row with a new head for each aspect.