The sphericity test was violated so now my question is: which results should I report? The BF or the parametric?

Thank you in advance!

I am using Bayesian repeated-measures ANOVAs and Bayesian paired-samples t-tests in JASP and would like to clarify how best to report effect sizes and uncertainty measures within a Bayesian framework.

For the Bayesian repeated-measures ANOVA, JASP provides a “Posterior R²”, described as the posterior distribution of explained variance. I am considering reporting the model-averaged posterior R² together with its 95% credible interval as a model-level measure of effect size/uncertainty.

I would like to clarify a few points:

• What exactly does the model-averaged posterior R² correspond to in Bayesian repeated-measures ANOVA? Is it the proportion of variance explained by the fixed effects, by the full model including random factors, or by the model-averaged set of models weighted by posterior model probabilities?
• Should posterior R² be interpreted as a model-level measure of explained variance rather than as an effect size for a specific factor?
• JASP seems to summarize posterior R² using a posterior mean and 95% credible interval, whereas Bayesian paired-samples t-tests report the posterior median of the standardized effect size δ with a 95% credible interval. Would it be possible to obtain and report either the posterior mean or the posterior median for both quantities? Reporting a posterior mean for R² and a posterior median for δ seems somewhat inconsistent, and I wonder whether a common summary statistic would be preferable for coherent reporting.

Many thanks for your help and clarifications !

Johan

I am currently running a 2 x 2 x 2 repeated measures ANCOVA in JASP. I have 2 within factors (animacy, word pairing) and one between factor (group).

I have found a significant interaction of animacy * group. However, when i use the “simple effects” option in JASP, it yields that both simple effects are not significant. I have plotted the interaction using the estimated marginal means using R; it is a slight crossover interaction. However the crossover is not as intense to make the simple effects non significant I believe. I have run additional pairwise post-hoc comparisons for the animacy by group interaction (using Bonferroni correction). The pairwise comparisions yielded two significant differences, one of which was that the two levels of the animacy factor for one of the groups (community-dwellers) showed a significant difference.

I additionally ran a Bayesian analysis for the ANCOVA and found that the BFincl for the animacy * group interaction was only 0.311, hinting towards the null hypothesis.

Could these issues be due to power? (N = 56; n group 1 = 26; group two = 30). What should I do now? Which additional analysis could I run and which analysis do I report? How do I interpet the interaction?

Thanks in advance!

ANCOVA

Post Hoc & Simple Effects

Bayesian analysis

I am currently running a 2 x 2 x 2 repeated measures ANCOVA in JASP. I have 2 within factors (animacy, word pairing) and one between factor (group).

I have found a significant interaction of animacy * group. However, when i use the “simple effects” option in JASP, it yields that both simple effects are not significant. I have plotted the interaction using the estimated marginal means using R; it is a slight crossover interaction. However the crossover is not as intense to make the simple effects non significant I believe. I have run additional pairwise post-hoc comparisons for the animacy by group interaction (using Bonferroni correction). The pairwise comparisions yielded two significant differences, one of which was that the two levels of the animacy factor for one of the groups (community-dwellers) showed a significant difference.

I additionally ran a Bayesian analysis for the ANCOVA and found that the BFincl for the animacy * group interaction was only 0.311, hinting towards the null hypothesis.

Could these issues be due to power? (N = 56; n group 1 = 26; group two = 30). What should I do now? Which additional analysis could I run and which analysis do I report? How do I interpet the interaction?

Thanks in advance!

I am working with JASP version 0.96.0.0.

In the edge stability plot, more than half of the red sample lines (and their red points) are missing and i wonder why. The grey confidence intervals appear normal.

Any help or fix would be greatly appreciated!

I am conducting a meta-analysis that aims to compare mean values of a continuous variable X between two study types, A and B. The variable ranges roughly from 10 to 100. I also have a few study-level moderators, such as mean age and proportion of females.

I ran a meta-regression using metafor. The model is three-level, with samples nested in studies (although only a few studies include more than one sample). It is a random-effects model, fitted with REML and based on t-statistics:

res <- rma.mv(
  yi = yi,
  V = vi,
  mods = ~ Group + other_moderators,
  random = ~ 1 | Study_ID/Sample_ID,
  method = "REML",
  test = "t",
  dfs = "contain",
  data = dat
)

Group is coded as a factor comparing study type A vs B.

Since I received mostly null findings, I would like to add a simple Bayesian framework to help interpret them. I am thinking about using a BIC-based Bayes factor approximation:

BF_10 ≈ exp((t^2 - log(k)) / 2)

where (t) would be the t-statistic for the group coefficient from my model, and (k) would be the number of samples.

Do you think this is reasonable in this context? Are there any important pitfalls?

Thank you

Any thoughts on why this might be happening?

Otherwise I could expand on what I've tried and their specific results (e.g. "Install Developer Module" method, via Preferences -> Advanced; via renv, building via RStudio project or separately within JASP R console; "Install Module" method with manual bundling of the .jaspmodule; manual copy into %APPDATA%\JASP\JASP\modules). I could also document the checks per method (e.g., renv::status() and library(jaspModuleTemplate) ok, jaspBase and jaspGraphs linked; simple "print-to-file" R functions).

Have used AI to scour JASP documents and check each step and suggest others, and review log files. Their diagnosis is "There is a fundamental breakdown in how JASP handles local developer paths", or "The "Developer Mode" appears to be broken for local builds, as the engine cannot maintain a stable link to the package it supposedly just "installed."" Seems implausible but I should ask before I keep going. Thanks in advance.

PTID creatinine

1 1.5

2 0.8

PTID age

1 55

2 70

PTID creatinine age

1 1.5 55

2 0.8 70

Is there a way to do this in JASP?

Thank you!

Andrew Kramer

I would like to sort them by a key variable, like in SPSS.

Can anyone advise?

I am using Random Forest on JASP for a multiclass classification problem, mainly to obtain variable importance scores. I also ran the same analysis in RStudio using the ranger package, with which I am more familiar. But for some reason, I obtain very different variable importance rankings between JASP and ranger (both for Mean Decrease in Accuracy and Total Increase in Node Purity), and the differences are stable across runs. I am using the same dataset and predictors in both analyses.

I am wondering whether JASP uses different defaults or a different implementation of Random Forest than ranger, and whether this could explain the discrepancy. Does anyone know what might cause such differences, or how I could check whether the model settings are truly equivalent across the two programs?

Thanks in advance for the help!

Gabrielle

Boa noite, necessito de abrir duas bases de dados para fazer um trabalho e não consigo sair da imagem que coloquei em anexo, alguém me consegue ajudar sff?

Obrigada.

Daniela Mendes

Is it new that JASP now open data preview window when you open a CSV file? Can this be turned off!

Per the JASP lover🥰

The RM ANOVA is 3x2, with Treatment (Control/Stress) X Visual Field (LVF/RVF) X Stimuli (Word/Non-Word).

The frequentist approach is revealing main effects of VF and Stimuli but no interaction effects:

When trying to find information about whether to select "Across all models" or "Across matched models" when computing the Bayesian RM ANOVA, I read that "Across matched models" is more similar to a frequentist RM ANOVA as it only compares to models with the same predictors.

When I run the Bayes RM ANOVA selecting "Across matched models", it appears there is substantial evidence in support of the interaction:

I'm surprised to see such a large BF (41) when it did not reach significance in the frequentist analysis (although the effect size appears large).

However, when I select the other option "Across all models", evidence in support of the interaction is anecdotal:

I guess I'm confused how the options can return such different results. I've read a few other posts querying this, but I'm still unclear what the best approach is.

So, my questions are:

1) Can somebody possibly provide a clearer definition explaining the difference between these two options?

2) Is one of these options more comparable with a frequentist RM ANOVA than the other?

Thanks in advance!

I would like to conduct a simple mediation (X --> M --> Y) using JASP classical and bayesian process module. However, I would like to include two other variables as controls, but I struggle to find a straightforward information whether and how this can be done in JASP. I tried to simply add them in the continuous predictors box, but I am nit 100% whether JASP automatically partials them out?

Any suggestions are greatly appreciated!

Thanks in advance,

Mila

I'm running a SEM model with multiple variables, some observed and some latent. I want to obtain factor scores for the latent variables, so I did a separate EFA (1-factor, 3 indicators) and checked the 'Add FA scores to data' box. But now I have 2 questions (for context, I'm not a statistician):

1) I don't really understand how the calculation of this variable works. Does it transform the raw indicator values with (e.g.) factor loadings, controlled for something else?

2) I also did a CFA (1-factor, same 3 indicators) and saved the factor score, but saw these scores are slightly different. The factor loadings differ as well, so I'm guessing it's got to do with that. But now the question is; which of the factor scores should I use, from the EFA or the CFA?

Hope you can help me with my questions, thanks in advance!

Kind regards,

Lars

I ran a Baysian rmANOVA with a factor time and a factor conditions. I am unsure how to interpret these results though. There seems to be string evidence for the main effect of time compared to the null model. But when I look at the post-hoc tests there doesnt seem to be evidence for an effect of time. This is consistent with the descriptive plot:

I was first thinking that it may be because the main effect of time collapses all conditions but they do not seem to be very different even then.

Can anyone help me interpret this output?

Thank you for your help!

Is there any way that we can program a package for such type of statistical problems?.

I am trying to understand the results of a Bayesian repeated-measures ANOVA.

I want to examine if there is a difference between the two groups (with different subjects) that were tested multiple times for an outcome measure (sprint time). I considered the pre-test, post-test and tapering 1 as "repeated measures factors", the pre-test, post-test and tapering 1 performance as "repeated measure cells", and the groups as "between-subjects factors". In addition, in the main options (order), I select ‘Compare to null model’.

Can you please tell me if my interpretation is correct?

"Using a Bayesian RM ANOVA (prior probabilities of each model were equal to 0.5), the Bayes factor indicates that the data is best represented by a model that included time point as the predictor, the group, and the interaction between the time-point and the group. Post hoc comparisons of pre-test vs. post-test and pre-test vs. tapering 1 revealed posterior odds of 20788.04 and 329497.52 against the null hypothesis, which indicates decisive evidence in favour of the alternative hypothesis. When comparing post-test vs. tapering 1, there was anecdotal evidence in favour of the null hypothesis. Additionally, post-hoc group comparisons exposed posterior odds of 20788.04, indicating moderate evidence in favour of the null hypothesis (no differences between groups)."

Thanks in advance,

Nuno Mateus

Thanks

I was looking to add a computed column in my data in JASP. In simple terms the computation is look at this other column and depending on what is in the other column produce a "friendly name" for the region for that record. The GUI for computed column definition looks to mimic the Filter options. That is a good thing as I might need some of the same features in Filter and Computed column.

In general the formula wold look like for the New Column = If Column A Contains 123 anywhere then return Africa, if column A Contains 456 then return Asia, If Column A contains 987 .... etc.

In general for computed columns and maybe filters nice, but simple set of standard text functions like IF, Left, Right, Mid, Contains, AND (boolean), OR (boolean) would be very helpful. I could also see this useful for Filters.

Does something like this exist without hand writing R or doing it in the source before the data lands in JASP?

If it does not exist does anyone know of a related issue? I could not find an issue to follow before making one.

I can see this for both Global Filters and for the proposed option to filter on the visuals/analysis as listed in this issue I keep hoping will land.

Love to hear others thoughts on the above items.

Alan