Random sampling: who gets selected is a separate question

The first question is how something enters the sample. That is different from what happens to it after selection. When a study mentions randomization, we check which of those steps it means.

Two separate actions with a set of cards

Imagine a fictional workshop with forty numbered design cards. A person uses a specified random procedure to select ten cards from the full set. That action selects a sample of cards.

Next, the ten selected cards are assigned to two display boards. Even if that second assignment is random, it does not tell us how the original ten were selected. We need the first procedure to answer the sampling question.

Now change the example: only cards from the top drawer were available for selection. Random selection within that drawer would not give cards in another drawer a chance to enter. The label “random” does not erase the limits of the set from which selection occurred.

Name the selection frame

The frame is the set or list used for selection in this example. A reading note can separate:

• Population the report wants to describe.
• List or set available for selection.
• Selection procedure actually used.
• Number selected and number represented in results.
• Any later assignment to groups or conditions.

Do not substitute “everyone” for a list whose coverage is not stated. Also keep an invitation, a response and a completed observation distinct when the report supplies separate counts.

A family-friendly way to ask

An adult may ask whether volunteers in a study were recruited from one service or from a broader population. A parent might look for the ages eligible for selection. A grandparent can help find where recruitment is described. A teen may notice that “randomly assigned” was repeated in a news story as “randomly selected.”

Those are questions about the documented process, not judgments about the people who took part. There is no need to collect identifying information about participants to understand the distinction.

What the word cannot promise

This page does not certify a study as representative, teach a complete sampling design or draw conclusions about ADHD prevalence. Our card illustration has no clinical meaning and no real participant data.

The practical result is a two-part sentence: “Selection into the study happened this way; assignment within the study happened that way.” If one part is not available, mark it unknown instead of borrowing certainty from the other part.

Source

[1] Australian Bureau of Statistics. Census and sample. Australian Bureau of Statistics. Published February 2, 2023. Accessed September 6, 2026. Random/probability sampling and simple-random-sampling sections; complete relevant page inspected. Supports the adjacent attributed claim. All illustrative scenes, numbers and optional note formats are original inventions, not source findings.

Related guides

• selection bias
• sample size
• randomized controlled trial
• generalizability