Hypothesis Direction Disclosure

When Licklider requires hypothesis direction to be disclosed, what one-sided testing means for your results, and how the disclosure appears in exported output.

A statistical test can be run in two ways: as a two-sided test, which asks whether there is any difference between groups, or as a one-sided test, which asks whether one group is specifically higher or lower than the other. The choice affects the p-value and the interpretation of the result.

When a one-sided test is used, or when the hypothesis direction was specified in advance, Licklider requires this to be disclosed before a confirmatory or publication-ready result can be exported.

Two-sided vs one-sided tests

Two-sided testing

The default. Tests whether the groups differ in either direction. The p-value reflects the probability of observing a difference this large or larger in either direction under the null hypothesis.

Most analyses should use a two-sided test unless there is a pre-registered, domain-justified reason to test in only one direction.

One-sided testing

Tests whether one specific group is higher (or lower) than the other. The p-value is approximately half that of the equivalent two-sided test, which makes the result appear more significant.

One-sided testing is only appropriate when the direction was specified before the data was seen — for example, in a pre-registered hypothesis that the treatment group will show higher values than control. Using one-sided testing after seeing the data direction inflates the false positive rate.

When the disclosure appears

The hypothesis direction disclosure is required for confirmatory and publication-ready analyses when:

  • A one-sided test was requested
  • A specific hypothesis direction was set in the analysis intent

For exploratory analyses, the direction is recorded but no confirmation is required.

The disclosure does not appear for standard two-sided analyses where no direction was specified.

What the disclosure confirms

When the confirmation appears, you are asked to acknowledge that:

  • The test direction was pre-specified before the data was examined
  • The direction is consistent with the pre-registered hypothesis

This acknowledgment is recorded in the figure's disclosure and included in the exported output.

Effect on export

If the hypothesis direction has not been disclosed for a confirmatory or publication-ready analysis that used a one-sided test, claim-bearing export is blocked. The Inspector will indicate that the direction disclosure is pending.

Relationship to analysis intent setup

The hypothesis direction is set in the analysis intent during study setup. The disclosure here is the reporting-side confirmation that the direction used in the analysis matches what was pre-specified.

For how to set the analysis intent and hypothesis direction before analysis → see Hypothesis Direction and Sidedness.

What this page does not cover