other methods - resampling Flashcards
(12 cards)
What is the main advantage of resampling statistics over traditional parametric tests?
Resampling makes fewer assumptions about the data, is flexible, and can be adapted to various situations without requiring complex equations or distribution assumptions.
What are the two main types of resampling methods?
Permutation Tests and Bootstrap Resampling.
What is the purpose of a permutation test?
To test if there is a significant difference between two groups by randomly shuffling the data and comparing the observed difference to the distribution of shuffled differences.
How is a permutation test conducted?
Shuffle the data between groups many times, calculate the mean difference each time, and compare the actual difference to this ‘null distribution’.
What does a low p-value from a permutation test indicate?
That the observed difference is unlikely to have occurred by chance, suggesting a significant result.
What is bootstrap resampling used for?
To estimate confidence intervals or standard errors by resampling with replacement from the original dataset and analyzing the variability of the statistic of interest.
How is a 95% confidence interval calculated using bootstrap resampling?
By ordering the bootstrap results and removing the top and bottom 2.5% of values.
How does within-subjects permutation differ from between-subjects?
Instead of shuffling group labels, the sign of each subject’s difference is randomized to preserve individual variability.
How does sample size affect resampling tests?
Larger samples reduce variability in the null distribution, making it easier to detect small effects as significant.
How can bootstrapping be used in one-sample tests?
By comparing how often bootstrap means fall below or above a hypothesized value, such as a population mean.
Can bootstrapping be applied to model parameters beyond the mean?
Yes, bootstrapping can estimate confidence intervals for any model parameter, like regression slopes.
What are some limitations of resampling techniques?
They require a computer, a large number of resamples, and representative original data to be effective.
