Panelist experience
This page is for the study team. It briefs you on the path a panelist walks and on the feedback the live page renders. The pages written to the panelist are under What a Delphi panel is. The signed-in article What the panelist sees is the staff reference. Sign in, then open Help & docs in the account menu.
The path
A study email opens the panelist's pages. The person does not use the study-team screens.
| Step | What the panelist does | What it does not do |
|---|---|---|
| Invitation | Says whether they are interested. | Saying yes is not consent and does not enroll them. |
| Questions, if the study asks | Answers them before the consent form opens. | The questions appear only when the study asks them. |
| Consent | Reads the published form. Agrees, or says they do not agree. | The page shows the published form and does not add text to it. |
| Each open round | Rates the statements. Answers save as they go. They submit while the round is open. | A late answer is refused at the cutoff. |
| A later round | Reads what the group said last time, then rates again. | No statement is given a decision label. |
Leaving, and what a removal does to earlier answers, is Leaving a panel. What can appear in a publication is What is published.
What the live feedback shows
On a later round the heading is What the group said in Round and the earlier round's number. The introduction says to read it, then rate again. The person can keep the rating or change it.
Under a statement, when the study chose to show them and the numbers are not withheld, the block can include:
- How many people answered.
- Median rating.
- Middle half of ratings, from the first quartile to the third.
- The share in the agree zone, In between, and the disagree zone. The words at the ends of the scale are the study's.
- You rated this, with the person's own earlier rating, or the study's unable-to-rate label.
- What other panelists said, for comments the team chose to share.
- Themes in the comments, when the team published themes. The introduction says each theme was raised by several different panelists. The rule that admits a theme is the study's smallest reportable group, which starts at 5 and can be set to 1. The word "several" is the screen's wording, not that floor.
The introduction defines the word Status: "Status is the study team's reading of where the group stands on that statement." That sentence is not a row. The figures under a statement do not include a Status row, a classification, or a decision word.
If too few people answered, the page says the group's numbers are not shown, so that nobody can be picked out. A statement reworded after that round shows nothing from it: no earlier numbers, no earlier rating, and no shared comments. A new statement says it is new in this round.
The live panelist block does not draw a histogram. Show a histogram is a Round design switch. The setup preview can draw bars. That preview is a different screen from the page a panelist opens.
What panelists never see
Peers do not see who gave a rating. Response rows do not store a name. Ordinary staff screens after enrollment show a pseudonym. A written reason can reveal an enrolled person's name, and that reveal is audited. This page does not call that arrangement absolute anonymity. The limit is on Identity and pseudonyms.
Panelists are not shown the staff classification labels Retain, Exclude, or Appropriate. Anchors on a new study do not say those words.
Next step
Open a round only after the approval on Running rounds. The panelist's own wording is What a Delphi panel is.