Lmst

🚨New blog post 📝: Your Study Is Too Small (If You Care About Practically Significant Effects)

#effectsize #precision #poweranalysis #research #Psychology #MCID #SESOI #samplesize

https://open.substack.com/pub/mzloteanu/p/your-study-is-too-small-if-you-care?r=3b457w&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Correlation Power Analysis with a Leakage Model

Eric Brier, Christophe Clavier, and Francis Olivier

https://www.iacr.org/archive/ches2004/31560016/31560016.pdf

#hardwarehacking #hacking #poweranalysis

#statstab #230 Power and Sample Size Determination

Thoughts: Frequentist power is a complicated and non-intuitive thing, so it's good to read various tutorials/papers until you find one that sticks.

#stats #poweranalysis #power #NHST #effectsize

https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_power/bs704_power_print.html

#statstab #192 Statistical Power from Pilot Data

Thoughts: One new use for pilot data - estimating the SE to power your study. If you can follow the formula. By @CarlisleRainey

#pilot #experiments #poweranalysis #standarderror #simulation #methodology

https://www.carlislerainey.com/blog/2024-06-03-pilot-power/

#statstab #189 Post Hoc Power: Not Empowering, Just Misleading

Thoughts: "the observed power is a 1:1 function of the P value" If you need a ref for why "post hoc" power is nonsense (paywalled).

#power #posthoc #QRPs #errorcontrol #poweranalysis

https://www.sciencedirect.com/science/article/pii/S0022480420305023

#statstab #155 Power & Sample Size calculations for ordinal data {Hmisc}

Thoughts: An appropriate sample size is just as important as an appropriate (ordinal) analysis. Nice guide for a 2-group design with a Likert-type DV.

#ordinal #NHST #poweranalysis

https://library.virginia.edu/data/articles/power-and-sample-size-calculations-ordered-categorical-data

code example for computing sample size with ordinal data

Undo Arduino Encryption with an Oscilloscope https://hackaday.com/2024/07/14/undo-arduino-encryption-with-an-oscilloscope/ #SecurityHacks #poweranalysis #ArduinoHacks #cryptography #arduino #crypto #rsa

it is very easy to tell two chips apart based on their startup power consumption. reset is negated at around 6500; after that moment, code starts executing. so the part before it should depend solely on the chip.

#spa #poweranalysis

Check out our reference list for others that cover the U.S., sub-Saharan Africa, and other sectors besides education.

Every time researchers publish their variance components they are supporting the next generation of field trials.

#researchmethods #poweranalysis

source: Gregory Hancock from Quantitude Podcast on X
#PowerAnalysis

Last week I attended the 6th Perspectives on Scientific Error Conference at @TUEindhoven
I learned so much! About #metascience #preregistration #replicability #qrp questionable research practices, methods to detect data fabrication, #peerreview, #poweranalysis artefacts in #ML machine learning...
I'm impressed by the commitment of participants to improve science through error detection & prevention. Thanks to the organizers Noah van Dongen, @lakens @annescheel Felipe Romero and @annaveer

Finishing my migration to #substack here is:

What NOT to do with NON-“null” results Part III: Underpowered study, but significant results https://open.substack.com/pub/mzloteanu/p/what-not-to-do-with-non-null-results?r=3b457w&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
#rstats #statistics #poweranalysis #typeM #falsediscoveryrate #NHST

The labor movement needs the solidarity and organization of tech workers acting in solidarity with warehouse or production workers dependent on their algorithms! What an awesome transformative power!

`https://stansburyforum.com/2024/01/25/tech-workers-deserve-a-union

https://labornotes.org/blogs/2024/01/book-review-power-tech-workers

Labor Power and Strategy by John Womack, Jr. for more discussion of #chokepoints and #poweranalysis
https://pmpress.org/index.php?l=product_detail&p=1298

#techworkers #union #labornotes

Statistical #PowerAnalysis currently dominates #ExperimentalDesign. In this Essay, @itchyshin &co argue that we should move away from the current focus on power analysis and instead encourage smaller scale studies & collaborative projects #PLOSBiology https://plos.io/48Gk8Og

How Good Is The Cheapest Generator On Amazon? https://hackaday.com/2024/01/04/how-good-is-the-cheapest-generator-on-amazon/ #electricalgenerator #poweranalysis #Reviews

Hello #statstodon! A reviewer asks me to perform a post-hoc #PowerAnalysis. I know this is generally not advised because if you replace the a priori effect size by the effect size measured in the experiment, this will introduce an erroneous relationship between the significance level of the test and the measured power.
… but does that mean that there is no proper way of measuring power retrospectively? For example, if you refrain from using the measured effect size and instead simulate a range of “a priori” effect size unrelated to the results of the test, then the dependency of the power to the significance level should not happen?
#stats #statschat @lakens

Had a lot of fun figuring out how modals work in Shiny in order to add this pop-up window of individual interaction simulations to my power analysis web app for interactions/moderation: https://david-baranger.shinyapps.io/InteractionPoweR_analytic/
#rstats #shiny #PowerAnalysis

Screenshot of a web app, where a simulated data set with an interaction is shown, next to panel of results from the regression.

Online #workshop:
Simulation-based power analyses in (generalized) linear mixed models
17.05.2023, 10-12h CEST

The workshop will cover basics of power analysis, linear mixed models, and why the combination of both requires a simulation-based approach.

In my experience, this is for many areas of #HealthSciences and #HRQL research a key problem when designing studies.

Maybe worth a read as well:
https://link.springer.com/article/10.3758/s13428-021-01546-0

#PowerAnalysis #SampleSize #Rstats #mixedpower #ZPID

Simulation-based power analyses in (generalized) linear mixed models

Abstract

The statistical power of a research design is closely linked to the reliability and replicability of empirical findings. Accounting for power while planning a study is therefore crucial and often a requirement for submissions in scientific journals. However, this can quickly become highly difficult in practice – especially for more complex, but very popular analysis procedures like linear mixed models (LMMs). In this workshop, we will briefly cover the basics of power analysis, linear mixed models, and why the combination of both requires a simulation-based approach. We will then focus on the R-package mixedpower and how to use it to estimate power in LMMs. The general aim of this workshop will be to help researchers build intuitions about simulation-based power analyses, and to empower them to set up highly powered research designs when they plan to use mixed-effect models to analyse the resultant data. A prerequisite for this workshop is a basic knowledge of R. Although we will briefly cover the basics of LMMs, familiarity with LMMs and the R-package lme4 is strongly recommended.

Just took part in a neat little workshop on #poweranalysis with #SEM (a technique for #dataanalysis), instructed by @yilinandrewang

Had experience with using his app (for my master's thesis, hopefully to be published next year), but nonetheless it was great to refresh my knowledge on SEM and its power in general.

My personal main takeaways:
1) Decent #measurement is central for SEM power.
2) Highly reliable short scales will give you a much more realistic estimate with decent sample size.
3) (Wasn't mentioned, but I was once again reminded of it): Confidence intervals for parameter estimates are even more important, as hypotheses test always against zero (as is common in regression)!
4) Model saturation is very important for estimation!
5) Multilevel SEM seems like a pain in the behind (guess what my next project is gonna be lmao)

All in all: Andre is a great teacher, is a cool guy and I would recommend his course to anybody interested in doing SEM properly!

#blinkylights #poweranalysis #demo #iot

#PowerAnalysis

Client Info