# How does the effect size inform the design (or analysis) of an NHST?

Cross Validated Asked on January 3, 2022

Consider this answer on how to design an NHST. I don’t quite understand what exact process one is supposed to follow to determine the minimum sample size once we have:

• A null hypothesis that is falsifiable. The choice of this null hypothesis will inform the assumed distribution of the data under the null.
• A minimum effect size that we want to be able to observe.
• A maximum false positive rate, i.e. the significant level $$alpha$$ against which we’ll compare a computed $$p$$ value.
• The maximum false negative rate = 1-$$beta$$, i.e. where $$beta$$ is called the power or sensitivity of the experiment.

If simpler, let’s assume a 1D Normal distribution for the Null.

My recollection is that one is supposed to use the CDF of the Null somehow for the test, but how? E.g. how does the chosen value for the minimum effect size inform the sample size?

## Related Questions

### Prediction Intervals (Conformal Predictions) for Regression Problems

0  Asked on December 29, 2021 by bioinformatics_student

### Why not use % change in regression instead of log diff?

0  Asked on December 29, 2021 by tjaqu787

### Estimating expected values for correlated data using random effects models

4  Asked on December 29, 2021 by nicolas-molano

### Should multiple testing correct with bonferroni ever reduce a p value’s size?

1  Asked on December 29, 2021

### Compute Mean of a Clipped Normal Distribution

2  Asked on December 29, 2021 by ahsan

### What is an example of perfect multicollinearity?

3  Asked on December 27, 2021 by tsteatime

### Should I use a seasonal arima or stl decomposition and model residuals only?

1  Asked on December 27, 2021 by string_is_hard

### Endogenous controls in linear regression – Alternative approach?

2  Asked on December 27, 2021 by sgtbp

### Variable selection in logistic regression model

3  Asked on December 27, 2021

### EM Algorithm Derivation, Discrete Case

1  Asked on December 27, 2021

### Should we really do Re-Sampling in Class Imbalance data?

2  Asked on December 27, 2021 by baktaawar

### How to determine data size is statistically efficient?

0  Asked on December 27, 2021 by 1111ktq

### (Non-limit) distribution of maxima from different univariate, discrete and stationary time series

1  Asked on December 27, 2021

### Confused about stationarity and ARIMA processes

2  Asked on December 27, 2021

### How to remove correlated features?

1  Asked on December 27, 2021 by ichait

### Between- and within-person level effects when using multilevel modelling for longitudinal data in R

1  Asked on December 27, 2021 by af1402

### How can I separate the abundance factor of the incidence results?

0  Asked on December 27, 2021

### Chi Square Test on non-numeric data in R

0  Asked on December 27, 2021 by ruffybeo

### How to make correlation test with compositional data?

0  Asked on December 27, 2021

### What are the parameters in signal recovery? Whether source of these parameters are the sampling property of impulse response?

0  Asked on December 27, 2021 by lakshman