Confounding Adjustment Disclosure

When Licklider requires covariate adjustment to be disclosed, what the confirmation involves, and how it affects claim-bearing export.

When a regression model includes more than one predictor, at least one of those predictors is functioning as a covariate — a variable included to adjust for its potential confounding effect on the relationship of interest. Reporting that covariates were used, and which ones, is a standard requirement in scientific publications.

Licklider confirms that covariate adjustment has been acknowledged before allowing claim-bearing export of multivariable regression results.

When the confirmation appears

The covariate adjustment confirmation appears when all of the following are true:

  • The analysis is an ordinary linear regression with two or more predictors
  • The regression diagnostics check has been resolved (or does not apply)
  • The figure is being prepared for claim-bearing export

For exploratory analyses, the adjustment is recorded automatically without requiring confirmation. The confirmation is required only when the intent is confirmatory or publication-ready.

Note: Logistic regression, GLMM, and Cox regression are not currently covered by this check. Support for these model types will be added in a future release.

What you are asked to confirm

When the confirmation appears, two options are presented:

Covariates disclosed in methods

You confirm that the covariates included in the model are reported in the methods section of the output. Select this when the covariate list will appear in the methods text.

Covariate adjustment not applicable

You confirm that adjustment was not required for this analysis — for example, because all predictors are of equal scientific interest and none is functioning as a confounder. Select this when disclosing adjustment would misrepresent the analysis intent.

Both options resolve the check and allow claim-bearing export to proceed.

Effect on export

When the check is unresolved — that is, when neither option has been selected — claim-bearing export is blocked for confirmatory and publication-ready analyses. The Inspector will indicate that the covariate adjustment disclosure is pending.

For exploratory analyses, the result can be exported without confirmation. If the figure is later used in a claim-bearing context, the confirmation will be required at that point.

Relationship to other checks

The covariate adjustment confirmation is separate from the regression diagnostics check, which evaluates whether the predictor structure (sample size, collinearity) supports reliable estimation. The regression diagnostics check must be resolved first — or must not apply — before the covariate adjustment confirmation appears.

For batch and plate confounding, which evaluates whether the batch structure is confounded with the experimental groups, see Batch and Plate Confounding.

What this page does not cover