An inquiry into the quality of a statistical test by calculating the power - the probability of rejecting the null hypothesis given that it is false - under certain circumstances. Power analysis is often used when planning a study to determine the sample size required to achieve a nominal level of ...

learn more… | top users | synonyms

1
vote
0answers
13 views

What, if anything, is the relationship between random intercepts models and compound symmetry?

In the analysis of correlated data, you're often working with replicates where a worker does a task 2 or 20 times, or an individual is measured for blood pressure randomly over several times during ...
0
votes
0answers
25 views

Power calculation for sensitivity and specificity analysis [on hold]

I am conducting a study to determine how good a diagnostic test is at identifying disease positive and disease negative. I want to do a power calculation and I know approximate disease positive as 150 ...
1
vote
0answers
9 views

Is Mann Whitney U test appropriate for samples of really different sizes and distribution?

A Mann-Whitney U test on two samples of different sizes (Group A: 170,000) Vs. (Group B: 200), different distribution shape (as per Kolmogorov-Smirnov Tests ks.test ...
0
votes
0answers
7 views

Power calculation for panel regression

How do you perform a power calculation in a panel regression setting? In particular what do you use as your sample size. The number of individuals or the number of observations (individuals x time ...
1
vote
1answer
23 views

Measures of minimal detectable effect size

Statistical power is defined as the probability of wrongly accepting the null hypothesis for a given sample size, p-value cutoff and effect size. I am after accepted measures of the converse: the ...
0
votes
0answers
14 views

Power analysis for polynomial regression

There are two variables, $X$ and $Y$. I am interested in the shape of the relationship between $X$ and $Y$, specifically whether it is linear, curved, or some higher order polynomial function, because ...
1
vote
0answers
18 views

GAM or spline power analysis in R

There is a package in R called pwr. This is useful to make power analysis when designing the sampling of a project. here are few examples: ...
2
votes
1answer
319 views

How to find correlations between multiple low-frequency variables?

Let's say we have multiple ($k > 2$) binary variables (e.g., genetic mutations and/or disease states), which appear at a given frequency of $f(1)$ to $f(k)$ in a sample. We would like to test ...
2
votes
1answer
32 views

Some questions on multiple logistic regression

Let's say we have k binary (binomial) variables, and a sample size of n observations per variable, where each variable occurs (positive case) at a given frequency/probability f. We would like to check ...
0
votes
1answer
64 views

Power calculations Uniform(0,$\theta$)

I am working on this problem and got about halfway through it, before getting stuck. Could anyone take a look at it? here's my problem... $x_1,...,x_4$ are distributed $U(0,\theta), \theta>0$ ...
0
votes
0answers
65 views

Power Analysis By Simulation

Problem : A simulation of power gives a far more conservative sample size to just using power.prop.test in R (say). Question : Is the simulation we are using a reasonable way to estimate power. Are ...
0
votes
2answers
30 views

Statistical power of binomial vs. paired t-test

I am conducting an experiment and I am deciding whether to use a forced-choice paradigm (participants choose A or B), or a continuous rating paradigm (rate each stimulus A and B on a Likert scale). ...
1
vote
0answers
32 views

Barnard's Exact Test (Boschloo's variant) when distribution is hypergeometric

Let's say, we have a standard $2 \times 2$ contingency table and because of low sample size we're about to use some type of Exact Test, e.g. Fisher's Exact Test (FET) or Barnard's Exact Test. It's ...
0
votes
0answers
27 views

Determining the power associated with the logistic model

Although there has been some detailed discussions about power analysis on this website (for example here and here), the answer provided to this question has outlines the steps to simulating a power ...
0
votes
0answers
17 views

How to determine statistical power when I must use Welch's Robust Test of Equality of Means

I couldn't use a one-way ANOVA on my means comparisons (3+ groups) because I have heterogeneity of variance (and unequal group sizes), so I used Welch's Robust Test of equality of Means. Of all my ...
1
vote
0answers
13 views

Combination of Tests' Power

I have set of tests. It is not difficult to calculate power for each of them (I have normal distributions and know effect sizes). I would like to infer power of combination of tests. I.e., $H_0 = $ ...
0
votes
0answers
8 views

guidelines for repeated measures anova power

I has been requested to assess the power of a repeated measures analysis experiment: 11 subjects, one between subject variable (product, three levels) and one within subject variable (time, four ...
0
votes
0answers
16 views

Power Analysis for Within Subject ANOVA

I am going to be collecting survey data that asks for participants' attitudes towards four groups of people. For the sake of this example, let's say that it's black women, black men, white women, and ...
0
votes
0answers
23 views

Determining sample size of skewed data

I have seen questions asking about sampling with non normal data but did seem to help. I have two datasets, dataset A containing results of students before ...
0
votes
0answers
12 views

Clustering and power: increasing the n but not the number of clusters

I am designing an experimental study with N observations from M sampled clusters, where M << N. Half of the clusters will be assigned to treatment. I can spend more money to increase the number ...
2
votes
2answers
116 views

Post hoc power analysis for a non significant result?

I have submitted a paper to a journal reporting a non-significant finding. One of the reviewers has asked me to include a power analysis on my data to work out how big a sample size I would need to ...
1
vote
1answer
38 views

What do they mean by 'To calculate sample size, I use simulation in all cases.'?

I was looking for a book on sample size calculations and power analysis and I met this phrase 'I use simulation in all cases.'. What does it mean, can i forget about using traditional methods for ...
1
vote
0answers
24 views

Minimum sample size for 1-tailed hypothesis testing

Let's say we have some endangered species of animals, whose population is 10,000 and dwindling due to some disease they all carry. And say I want to be able to test the efficacy of some new drug (with ...
0
votes
1answer
76 views

How to draw the estimated power curve of a test?

The estimated power of a given test can be reported with a table or with a curve in a plot. I need to compare the power of two permutation tests by simulation and I was wondering how to create a ...
0
votes
1answer
17 views

converting a chi-square with more than one degree of freedom into a pearsons r

I want to run a power analysis for a study using correlations between two continuous variables. Previous research with the variable I am interested in used a Kruskal-Wallis test to look at differences ...
0
votes
2answers
13 views

Prospective power using observed difference

In my line of research we often calculate prospective power of a proposed study as a way to determine what sample size we should aim for. However, I'm often studying phenomena, or interactions between ...
0
votes
0answers
19 views

Power of a binomial based test

I have a Bernoulli trial, which by definition can be either success of failure, and I have a Binomial random variable of the number of successes, where I don't know p (I wish to estimate it). I have a ...
0
votes
0answers
11 views

Power calculation for assessing relatively success two different predictors on a continuous outcome, Fisher r-to-z?

I'm thinking about a study where I assess the relatively predictive power of two different measures. Let's say a patient comes to a clinic on 3 different days, and 3 different measures of a noisy ...
0
votes
0answers
13 views

Help with Power analysis with PASS

I'm putting together a Power analysis, which is not my strong suit. I'm using the Poisson Regression tool in PASS. My outcome variable is falls over a five year follow-up period. My exposure is a ...
1
vote
0answers
11 views

Why is a larger sample required for a weaker rate when the response is a binary event?

I am interested in conducting an experiment and am attempting to calculate the sample size required for this experiment. If I hold beta at 0.8, alpha at .95, and am looking for a 5% increase, why do I ...
1
vote
0answers
23 views

Determining the number of operators (random effect) for a two-way chi-square test

I am designing a study to benchmark a new measurement technique with two traditional measurement techniques. Let's call the new technique N and the two traditional techniques T1 and T2. It is believed ...
1
vote
0answers
17 views

Statistical support to aid setting weights in a simple function

I'd like to create some consensus scoring function that incorporates two independent scoring functions $f_1$ and $f_2$. These functions rank items $i$ where there are $i=20,000$ items. $C_i = (f_1 ...
8
votes
1answer
231 views

Consequences of exceeding sample size after determination of sample in power analysis

Challenge In our office, I have been involved in a discussion about sample size and its influence on effect size - can you help me out and explain further? Base When conducting a power analysis, ...
0
votes
1answer
37 views

Power calculation using simulation — should I use empirical distribution?

In most tutorials on power calculation using simulation (e.g. this example in R), the analyst simulates the outcome variable using some convenient distribution such as the normal distribution. It ...
1
vote
0answers
21 views

Inflated Type 1 error in glmer (for main effect but not interaction?)

I am doing simulations of type 1 error, power, and power' (power corrected for anti conservativity) for research on a specific application of (g)lmer, namely to small-N designs of longitudinal ...
1
vote
1answer
37 views

Can increasing the significance level help to avoid detecting minor differences in large datasets?

In the accepted answer of THIS question, it is stated that "With such large sample sizes both tests will have high power to detect minor differences". On the other hand, a voted up comment in THIS ...
0
votes
0answers
37 views

How do I calculate the N to achieve a given precision for a binomial with parameter p?

I want to find out the prevalence of tuberculosis in patients coming to hospital with acute respiratory infection. What should be the sample size that will be needed given a precision of 5% with a ...
1
vote
0answers
28 views

Power calculations with a continuous independent variable in a case-control study with pair-matched data

So I have been struggling with this matter quite a while now and still have not found a solution to this problem. In fact similar questions have been asked here before but no answer was provided to my ...
0
votes
0answers
24 views

standard error for proportions from a bootstrap sample

I am trying to estimate the power (with confidence interval) of one-sample $t$-tests (with sample size $N$) on a parametrized distribution. One approach is to draw $N\times K$ samples from the ...
0
votes
0answers
30 views

How to calculate the required sample size?

Basically I am looking to see if a resistance to a certain drug is associated with a higher level of plasma protein. Resistance is a" yes or no "variable and the plasma protein is a nominal value. ...
0
votes
0answers
8 views

Power Calculation for Gene Significance

So, I think I need to do a power calculation for the following, but I am really not sure if that's the right procedure to use: I am looking at gene types in cancer, and I want to figure out if ...
0
votes
0answers
14 views

Accounting for time series correlation in power calculations

we want to do power calculations for a clustered RCT design household survey. We have a good idea for cross-sectional ICC from a similar earlier survey in the same study area. However, this time we ...
1
vote
1answer
124 views

Power analysis for a multinomial logistic regression

How do I conduct a power analysis for a multinomial logistic regression analysis? I have 1 independent variable (with 3 levels) and 1 dependent variable (with 2 levels). All variables are continuous. ...
1
vote
1answer
242 views

Power analysis for inter-rater reliability study (Kappa) with multiple raters

I've spent some time looking through literature about sample size calculation for Cohen's kappa, and in several studies there are stated that increasing the number of raters, reduce the number of ...
0
votes
0answers
33 views

Sample size for variable selection methods?

I have data in which the response variable is occupational injury rates ( positive continuous variable), and the predictors the job characteristics (standardized on a rating from 0 to 5) which might ...
4
votes
1answer
56 views

Why does significance not equal validity? Why can't I stop a test as soon as it has significant results?

According to this article, there's an issue with concluding a test as soon as it reaches significance (the "peeking effect"). Instead you need to determine the sample size ahead of time, and only look ...
0
votes
0answers
11 views

How many raters do I need to determine reliability of agreement on stimulus classification?

I have a stimulus set comprised of 147 videos, each with one of seven labels (hidden from raters). I want to determine how reliable label assignment is for each stimulus -- that is, I want to be able ...
1
vote
0answers
65 views

Sample size for A/B testing

I'm using Evan Miller's sample size calculator as well as his article "How not to run an A/B test". In my case, I'm using a conversion rate of 6.3 with a minimum detectable effect of 5% (that is, ...
4
votes
0answers
38 views

Power in survival analysis

I just read an article discussing the effect of radiotherapy duration on survival. They state the following: It was expected that changing the treatment from 1 week to 3 weeks would improve ...
0
votes
0answers
17 views

Power calculation to test different intercepts in simple linear regression

I have two groups of data: patients and healthy controls. I perform a measurement and a modelled estimation of this measurement on both groups. I then calculate using simple linear regression ($y = ...