Questions tagged [gpower]
G*Power Statistical Power Analyses for Windows and Mac is a program which helps with calculating the power of some statistical tests.
90 questions
5 votes
2 answers
101 views
Power of t-test with unequal sample sizes: GPower and MESS::power_t_test do not converge?
I've created two item sets with 24 items each, that I expect to differ along my outcome variable. To test this assumption before using the items in future studies, I'm planning to recruit human ...
- 53
0 votes
0 answers
56 views
Experiment Design Power Analysis for PhD User Study, Within or Mixed Subjects
I'm designing a user perception study as part of my PhD project, and I'm trying to figure out the sample size I need. I created clips of an avatar talking for 20-30s, and I'm varying the verbal style (...
- 1
0 votes
0 answers
74 views
How to calculate power for a quasi-poisson regression
I ran a quasi-poisson regression (due to overdispersion), now I'm wondering how I can calculate the power of my test. I tried G-Power but it seems that it can only calculate power for Poisson but not ...
- 1
0 votes
0 answers
43 views
G*Power Repeated Measures with multiple items
I am interested estimating a sample size required to detect a small effect (Cohen's f = .15), alpha = .05 and power = .80. If I have a one-way repeated measures design (4 levels of difficulty: easy, ...
- 373
2 votes
1 answer
112 views
Statistical power using mixed models
For computing statistical power of planned contrasts from a (generalized) linear mixed model, I have read that simulation is often used as the design / model is too complex that an analytical formula ...
- 1,173
1 vote
0 answers
85 views
Bootstrapping as an alternative to simulation-based power/sample size calculations
Typically when one has to do a sample size and/or calculation, they might use simple sample size calculations from textbooks or use gpower. However, there are only closed form calculations for very ...
- 729
3 votes
1 answer
368 views
How can I convert an effect size from Cohen's d to f(v)
I am interested in calculating the power I need to compare several groups to one reference group in how much an outcome changes between two timepoints. My understanding is that I should use the ANOVA: ...
- 423
0 votes
1 answer
249 views
Calculating sample size based on a 2 x 2 x 2 repeated measures ANOVA using G Power
I am planning on doing a 2 (A: a1, a2) x 2 (B: b1, b2) x 2 (C: c1, c2)repeated measures ANOVA where A and B are within subject factors and C is a between subject factor. In G Power, I suppose the &...
- 1
0 votes
1 answer
307 views
Guidance on running G*Power to calculate the sample size using Poisson Regression
I am working on a dissertation to determine if there is a difference in the number of specific disciplinary actions (5 count dependent variables) from before remote learning and after remote learning (...
1 vote
0 answers
104 views
Sample size estimation using G*Power
I am trying to calculate an a priori sample size using the G*Power for a project that I intend to do. I will collect reaction time data from two groups of participants (G1, G2) across two time points (...
- 53
0 votes
0 answers
101 views
Mixed ANOVA Power Analysis
So I'm doing a thesis on compassion meditation and I'm confused about the power analysis. When I calculated it in G*power, "total sample size= 34", but it's for a 2x2 mixed ANOVA, so does ...
- 1
2 votes
1 answer
175 views
How do I to correct the α in the power analysis？
The reviewer asked us to use a corrected α when calculating the sample size with G*Power. The literature I found suggests that the Bonferroni adjustment = α/n (Armitage, Peter; Colton, Theodore, 2005)....
- 21
1 vote
1 answer
136 views
Power of One Sided t-test
Let $X_1, \ldots, X_n$ be a sample from $N(\mu, \sigma^2)$ for unknown $\mu \in \mathbb{R}$ and unknown $\sigma > 0$. Fix $\mu_0 \in \mathbb{R}$. The one-sided hypothesis is $H_0: \mu \leqslant \...
- 808
3 votes
1 answer
310 views
Approximate power analysis for simple linear mixed model without random effects
For educational purposes, a repeated measure of all participants in two conditions is modeled as linear mixed model. We're interested in the difference in outcome while controlling for confounders $...
- 247
4 votes
2 answers
361 views
sample size calculation using G*Power -- which model to use based on an already calculated effect size?
I want to perform a sample size calculation using G*Power. I have 2 predictors and want to test how these 2 predictors affect change over time in a specific cognitive measure. The cognitive measure ...
- 41