Why This Produces More Reliable Answers

Showing an example and asking for step-by-step reasoning both do the same underlying thing: they give the tool more to work with before it commits to a final answer, instead of forcing a single instant guess from a short instruction alone.

The two combine well: show one worked example of how you want a problem solved, then ask it to apply that same reasoning to a new case and show its work. This is especially useful for repetitive tasks with a consistent logic — categorizing items, scoring something against fixed criteria, applying one rule to many inputs.

Neither trick makes the tool infallible — it can still show confident, wrong reasoning, or an example can still be applied inconsistently. "More reliable" means fewer mistakes on average, not zero. The verification habit from level 1 still applies to whatever comes out the other end.

What do "show an example" and "think step by step" have in common as prompting techniques?

Techniques like showing an example or asking for step-by-step reasoning make answers more reliable on ____, not guaranteed to be perfect.

Combining a worked example with a request to "show your reasoning" is especially useful for repetitive tasks that follow one consistent rule, like categorizing a list of items.