Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
psy400ch14p364-467
Unequal Variances
| Term | Definition |
|---|---|
| Statistical tests of equality of variances | Levene's test or F test of equality of variances |
| Levene's test and F test of equality of variances | A test of the equality of two variances. |
| Unequal Cell Sizes in Factorial ANOVA Designs | can lead to confounding of experimental effects |
| Lack of Sphericity with Repeated Measures | may increase the probability of a type I error. |
| Mauchly's test of sphericity | tests the sphericity assumption in repeated-measures analysis of variance. |
| In the face ofsphericity violations, you may analyze your data by applying | statistical adjustments such as the Greenhouse-Geisser correction |
| Greenhouse-Geissor correction | adjusts to account for violation of sphericity in repeated measures analysis of variance. |
| Robust statistical methods: | tests that are not greatly affected by conditions that violate the assumptions of standard parametric tests |
| example of assumptions that robust tests are not greatly affected by | nonnormal distribution, the presence of outliers, skewed distributions, or unequal variances |
| Example of a robust statistical method | nonparametric versions of statistical tests such as the Mann-Whitney U test to analyze rank-order data |
| Robust statistical tests may be either | nonparametric or parametric |
| Trimmed mean | computed after removing some fixed percentage (often 10% or 20%) of the largest and smallest values in a data set |
| parametric approaches which estimate specialized parameters, such as a difference in trimmed means | require special procedures for computing important values such as standard errors and confidence intervals |
| BAYESIAN DATA ANALYSIS | prior distribution concerning a hypothesis is combined with experimental data to construct a posterior distribution for the hypothesis. |
| Prior distribution: In Bayesian data analysis, information the experimenter has concerning a statistic or effect prior to running | the experiment. Such prior information may come from previous experiments or a meta-analysis of the relevant literature. |
| In the stroop test a prior distribution might be based on previous experiments. For example, we may have good reason to believe | the difference in mean response times between the interference group and the no-interference group is between 0.5 and 2.0 seconds (likely), — 0.05 and 0.5 (less likely) or 2.0 and 2.5 (quite unlikely) |
| Objective or minimally informative prior: In Bayesian data analysis, a prior distribution that reflects relative ignorance | concerning the value of an experiment effect. For example any difference of mean values from -30 to +30 equally likely |
| Posterior distribution: In Bayesian data analysis, the distribution around a statistic (e.g., mean, difference of means, standard | deviation) that incorporates both prior Information and experimental data |
| Credible interval: In Bayesian data analysis, a range of values from the posterior distribution of a given quantity (e.g.,mean, | difference of means) for which you have a given level of credibility (e.g., 95%) concerning the true value of the quantity |
| Unlike confidence intervals, the 95% credible interval tells us that, based on our prior distribution and our collected data, | there is a 95% chance that the actual difference of means in the population falls between in the interval |
Created by:
james22222222