[Bug]: Possible incorrect labeling of sample size ratio N1/N2 in Power Module

[arnokimeswenger]

JASP Version

0.98.1

Commit ID

No response

JASP Module

Power

What analysis are you seeing the problem on?

Independent Samples T-Test

What OS are you seeing the problem on?

Ubuntu

Bug Description

Needed parameters for computing the power:

Sample size per group: N (e.g. 20)
Sample size ratio: N1/N2 (eg. 2)

First, the role of N is not clear (at least in my opinion). Is it N1, N2 or something else?
Second, when looking at the answer "A Priori Power Analysis", I see that Jasp uses N1=20, so the role of N is N1. And N2=40. But then the ratio (2 in my case) is not N1/N2 but N2/N1.

bug.zip

Expected Behaviour

In my opinion:
N should be N1
N1/N2 should be replaced by N2/N1

Steps to Reproduce

The problem can be reproduced without data. Maybe the issue occurs for other statistical tests too

Log (if any)

No response

More Debug Information

-------- Application Info --------
JASP Version: JASP 0.98.1
Build Branch: HEAD
Build Date: Jul 6 2026 11:03:35 (Netherlands)
Last Commit: 8fe3677d2e6aeea71f7135f48281e92aa88f38ba

-------- Basic Info --------
Operating System: KDE Flatpak runtime
Product Version: 6.10
Kernel Type: linux
Kernel Version: 6.17.0-35-generic
Architecture: x86_64
Install Path: /app/bin
Platform Name: xcb
System Local: en_US

-------- Module Library Info --------
{
"arch": "Flatpak_x86_64",
"developerMode": false,
"font": "SansSerif",
"installedModules": {
"jaspAcceptanceSampling": "0.95.5-release.12",
"jaspAnova": "0.95.5-release.12",
"jaspAudit": "0.95.5-release.12",
"jaspBFF": "0.95.5-release.12",
"jaspBain": "0.95.5-release.12",
"jaspBfpack": "0.97.2-release.0",
"jaspBsts": "0.95.5-release.12",
"jaspCircular": "0.95.5-release.12",
"jaspCochrane": "0.95.5-release.12",
"jaspDescriptives": "0.96.5-release.0",
"jaspDistributions": "0.97.2-release.0",
"jaspEquivalenceTTests": "0.95.5-release.12",
"jaspEsci": "0.95.5-release.12",
"jaspFactor": "0.97.1-release.0",
"jaspFrequencies": "0.95.5-release.12",
"jaspJags": "0.95.5-release.12",
"jaspLearnBayes": "0.96.5-release.0",
"jaspLearnStats": "0.95.5-release.12",
"jaspMachineLearning": "0.95.5-release.12",
"jaspMetaAnalysis": "0.97.5-release.0",
"jaspMixedModels": "0.95.5-release.12",
"jaspNetwork": "0.96.3-release.0",
"jaspPower": "0.95.5-release.12",
"jaspPredictiveAnalytics": "0.95.5-release.12",
"jaspProcess": "0.95.5-release.12",
"jaspProphet": "0.95.5-release.12",
"jaspQualityControl": "0.97.1-release.0",
"jaspRegression": "0.95.5-release.13",
"jaspReliability": "0.95.5-release.12",
"jaspRobustTTests": "0.95.0-release.12",
"jaspSem": "0.95.5-release.12",
"jaspSummaryStatistics": "0.95.5-release.12",
"jaspSurvival": "0.95.5-release.12",
"jaspTTests": "0.95.5-release.13",
"jaspTimeSeries": "0.95.0-release.12",
"jaspVisualModeling": "0.95.5-release.12"
},
"language": "en",
"theme": "light",
"uninstallableModules": [
"jaspDistributions",
"jaspFactor",
"jaspLearnBayes",
"jaspMetaAnalysis",
"jaspQualityControl"
],
"version": "0.98.1"
}

Final Checklist

• I have included a screenshot showcasing the issue, if possible.
• I have included an encrypted JASP file (zipped) or data file that causes the crash/bug, if applicable.
• I have accurately described the bug, and steps to reproduce it.

[tomtomme]

@arnokimeswenger
thx for the report.
Both bugs were reported before, see below. Sry for taking so long for a fix. Lets hope that someone finds the time soon

Do you agreee that these two are the same?
#4353
#3728

[arnokimeswenger]

Yes, your are right!

[julianwuth]

I created a PR that addresses this issue.

However, how do we want to handle the explanatory text? The text for the paired samples t-test mentions: "A design with a sample size of 20 can detect effect sizes of [...]". This refers to the sample size within each group or, in other words, the number of pairs. Should we clarify this for the paired samples case? @EJWagenmakers

[EJWagenmakers]

Yes, good idea