Apps

🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!

Type III Error

Also Known As: Wrong question error
Aspect ID: type_3_error

Definition

A Type III error occurs when a researcher correctly rejects the null hypothesis but draws the wrong conclusion about the nature or direction of the effect — getting the right answer to the wrong question. This can occur when the statistical test is valid but the operationalization of the hypothesis is flawed, or the conclusion drawn from the significant result does not follow from what was actually tested.

Examples

A study tests whether a training intervention 'improves performance' and measures completion time. The intervention significantly reduces completion time. But the training improved speed at the expense of accuracy — the study answered 'did completion time change?' rather than 'did performance improve?'

A public health campaign tests whether a new messaging strategy 'changes smoking rates' and finds a statistically significant effect. Officials celebrate and roll out the campaign nationwide — but closer inspection reveals the campaign increased smoking among teenagers while reducing it among adults. The null was correctly rejected, but the conclusion about the direction of the effect was wrong for the most vulnerable group.

An economics team correctly finds that a minimum wage increase significantly affected employment levels in a region. They conclude the policy reduced employment, consistent with classical theory. In reality, the data show employment rose — the team had miscoded the direction of the employment change variable and drew the exactly opposite conclusion from a real effect.

Verification Steps
Verification Steps
Binary yes/no questions that an AI must answer to detect a reasoning pattern in a text.
Each of the 452 aspects has verification steps — simple yes/no questions designed to systematically detect whether a pattern appears in a text. For ad hominem: "Does the argument attack a person rather than their claim?" For false dichotomy: "Are only two options presented when more exist?" This ensures consistent, reproducible analysis.

Binary (yes/no) questions an LLM must answer to identify this aspect:

  1. 1

    Was the null hypothesis correctly rejected based on the statistical test?

    Type: binary
  2. 2

    Is the reported direction or nature of the effect consistent with the data?

    Type: binary
  3. 3

    Could the significant result reflect a different question than the one the researcher intended to answer?

    Type: binary
  4. 4

    Is the operationalization of the outcome variable well-matched to the theoretical construct?

    Type: binary
Deep Dive
The expandable detail section on each aspect page with examples, psychology, and counter-strategies.
The Deep Dive section provides in-depth information about each aspect: a real-world example showing the pattern in action, an explanation of why it works psychologically, practical advice on how to counter it, alternative names, and links to related aspects.