Results of the combined diagnostic tests
Double-click in any of the pop cells in the table above to enter or edit data directly (either after first entering or uploading as a file).
How to cite this page:
Tool developed by Alberto Gómez-Buendia (University Complutense of Madrid) and Simon Firestone (University of Melbourne), with coding support from Poppy Schlaadt (Epi-interactive).
Until our paper specific to this tool is available, please cite:
Cheung, A., Dufour, S., Jones, G., Kostoulas, P., Stevenson, M., Singanallur Balasubramanian, N., Firestone, S.M., 2021. Bayesian latent class analysis when there is an imperfect reference test. Rev Sci Tech Off Int Epiz.
We also make use of R, JAGS and the R packages R2jags and mcmcplots. Please check their websites for how to appropriately cite their contributions.
Please use our tool taking due consideration of this quote from Chapter 1.1.6 of the WOAH Terrestrial Animal Health Manual:
Because Bayesian latent class models are complex and require adherence to critical assumptions, statistical assistance should be sought to help guide the analysis and describe the sampling from the target population(s), the characteristics of other tests included in the analysis, the appropriate choice of model and the estimation methods should be based on peer-reviewed literature.
Prior probabilites
Double-click in any of the cells in the table above to adjust either shape parameter for any of the priors.
* If any of your populations are considered to be possibly disease free, then select them at the top and we will use a mixture prior for them which requires two parameters (PrevX and TauX). TauX is the prior proability that populationX is diseased at all (or 1 minus the prior probability of disease freedom). PrevX is the expected prevalence in populationX if it is actually diseased (akin to the other Prev priors).
Our online implementation of beta-buster can be used to derive the shape parameters for prior distributions. See details on that page for interpretation of the strength of priors which should be considered with respect to the sample size.
See also the reporting standard (STARD-BLCM) for detail on the assumptions and reporting of studies, specifically with respect to the need for prior sensitivity anlayses to ensure appropriate consideration of the influence of the priors used on the inference.
Covariance terms
Disease
Non-Diseased
Select which covariance terms you wish to include in your model.
Please note that you cannot select higher order covariance terms (like Tests 1 & 2 & 3) if a nested lower order term is not included (like Tests 1 & 2). Selection of which terms to include should be based on consideration of biological plausibility, the data available and whether convergence is achieved, the stability and importance of the terms in output of repeated model runs. Comparing DICs is an unreliable basis of making such decisions. Typically following a manual forwards selection approach recommended over starting with an over-parameterised model and working backwards, due to issues of stability of the outputs.
Initial values
Double-click in any of the cells in the table above to adjust any of the initial values for either chain.
MCMC settings
Click on the button below after uploading the data and setting all parameters of your model:
How to cite this page:
Tool developed by Alberto Gómez-Buendia (University Complutense of Madrid) and Simon Firestone (University of Melbourne), with coding support from Poppy Schlaadt (Epi-interactive).
Until our paper specific to this tool is available, please cite:
Cheung, A., Dufour, S., Jones, G., Kostoulas, P., Stevenson, M., Singanallur Balasubramanian, N., Firestone, S.M., 2021. Bayesian latent class analysis when there is an imperfect reference test. Rev Sci Tech Off Int Epiz.
We also make use of R, JAGS and the R packages R2jags and mcmcplots. Please check their websites for how to appropriately cite their contributions.
Please use our tool taking due consideration of this quote from Chapter 1.1.6 of the WOAH Terrestrial Animal Health Manual:
Because Bayesian latent class models are complex and require adherence to critical assumptions, statistical assistance should be sought to help guide the analysis and describe the sampling from the target population(s), the characteristics of other tests included in the analysis, the appropriate choice of model and the estimation methods should be based on peer-reviewed literature.
Click the button twice the first time (once to initialise, ignore any error below the first time)
Model summary
For each parameter, n.eff is a crude measure of effective sample size and should be above 200. Rhat is the potential scale reduction factor (at convergence, Rhat must be below 1.01). If those requirements are not met, consider re-running the model increasing the number of iterations. Download your Model Code Download your ResultsHow to cite this page:
Tool developed by Alberto Gómez-Buendia (University Complutense of Madrid) and Simon Firestone (University of Melbourne), with coding support from Poppy Schlaadt (Epi-interactive).
Until our paper specific to this tool is available, please cite:
Cheung, A., Dufour, S., Jones, G., Kostoulas, P., Stevenson, M., Singanallur Balasubramanian, N., Firestone, S.M., 2021. Bayesian latent class analysis when there is an imperfect reference test. Rev Sci Tech Off Int Epiz.
We also make use of R, JAGS and the R packages R2jags and mcmcplots. Please check their websites for how to appropriately cite their contributions.
Please use our tool taking due consideration of this quote from Chapter 1.1.6 of the WOAH Terrestrial Animal Health Manual:
Because Bayesian latent class models are complex and require adherence to critical assumptions, statistical assistance should be sought to help guide the analysis and describe the sampling from the target population(s), the characteristics of other tests included in the analysis, the appropriate choice of model and the estimation methods should be based on peer-reviewed literature.