Selecting a radio option replaces the previous choice within its own group, not every choice on a page. Start by identifying which question the group answers. Layout, language and optional extras may be separate questions even when they appear in one form.
What the group rule means
Native HTML radio buttons with a shared name form a group. Its buttons compete only with one another. Separate groups remain independent, even when they appear beside each other on the same screen. Checkboxes use a different model: each checkbox can be selected or cleared without changing its neighbours. See the MDN explanation of radio inputs for the underlying HTML behaviour.
Worked example: a display form
Hypothetical example: Imagine a fictional display form with a layout group, a language group, and two independent checkboxes. Initially, Compact is selected, English is selected, and Captions is checked while References is clear. The person then selects Comfortable. The resulting state is:
| Control | Before | After selecting Comfortable |
|---|---|---|
| Layout radio group | Compact selected | Comfortable selected; Compact cleared |
| Language radio group | English selected | English still selected |
| Captions checkbox | Checked | Still checked |
| References checkbox | Clear | Clear |
The observation is limited and precise: Comfortable replaces Compact because both belong to the same radio group. English is not cleared because it belongs to another group. Captions stays checked and References stays clear because neither belongs to the layout group. Nothing in this rule requires a person to select at least one option unless a separate form requirement says so.
Practical limit when reading a screen
Do not infer that every circular-looking control shares one selection rule. First identify the named or explicitly declared group, then compare changes within that group. Visual placement alone cannot establish membership. A custom widget may not follow native HTML rules. If choosing a layout unexpectedly changes a language or optional extra, record the exact labels and before/after states. That observation supports a useful report without guessing how the controls were implemented.