Term
|
Definition
| the probability of correctly failing the null hypothesis when there is no effect |
|
|
Term
|
Definition
| the power of an experiment, the probability of correctly rejecting the null hypothesis where there is an effect |
|
|
Term
|
Definition
| rejecting the null hypothesis incorrectly |
|
|
Term
|
Definition
| retaining null hypothesis incorrectly |
|
|
Term
|
Definition
| setting it at .05, .01 or less |
|
|
Term
|
Definition
| our assumptions about how much of a difference might exist in the world is referred to as the EFFECT SIZE, assuming different effect sizes can show how much power our study has, allowing us to better evaluate the failure to reject the null hypothesis |
|
|
Term
|
Definition
| the proportion of the distribution under p-real that falls into the rejection region for the null hypothesis (p-null) |
|
|
Term
|
Definition
| a probability of success that represents the effect size (the true difference in the world) |
|
|
Term
|
Definition
| the probability of success under the null hypothesis (almost always 0.5) |
|
|
Term
|
Definition
| when a variable is influenced by many small independent factors, its distribution will approach the normal distribution as the size of the sample increases |
|
|
Term
|
Definition
| Have a population, draw a sample of size N, calculate a statistic on that sample, repeat the process over and over, this describes the values of the statistic and how often these values occur, the distribution of the statistic depends on the characteristic of the underlying population and the sample size (N) |
|
|
Term
|
Definition
| obtain a date sample, calculate a test statistic, understand the sampling distribution of the test statistic under the null hypothesis, use this understanding to identify critical values, test the likelihood of getting the observed sample under the null hypothesis by comparing the test statistic to the critical values, reject or fail to reject the null hypothesis |
|
|
Term
|
Definition
| used when the mean and standard deviation of the population, specify the sampling distribution of the mean for your sample size, draw sample, test hypothesis the mean could have occurred by chance given the null hypothesis |
|
|
Term
|
Definition
| the number of scores that are free to vary in calculating that statistic (Mean has N df, standard deviation has N-1 df) |
|
|
Term
|
Definition
| developed to be used when the standard deviation has to be estimated from the sample |
|
|
Term
|
Definition
| absolute value of the difference between sample and population mean in standard deviation units |
|
|
Term
| Different uses of statistics |
|
Definition
| Descriptive, Hypothesis testing, size or effect, point estimation |
|
|
Term
|
Definition
| quantify the degree of accuracy or precision of a point estimate, values that bound the upper and lower end of the confidence interval are confidence limits |
|
|
Term
| Three applications of t-test |
|
Definition
| t-test for a single sample, t-test for correlated samples, t-test for independent groups |
|
|
Term
| t-test for a single sample |
|
Definition
| compare a mean to a hypothesized value |
|
|
Term
| t-test for correlated samples |
|
Definition
| compare means when the data is paired within subjects (pre-post) or between subjects (matched pairs) |
|
|
Term
| t-test for independent groups |
|
Definition
| compare means between two groups formed by experimental (random assignment) or observational (e.g., gender) independent variables |
|
|
Term
|
Definition
| sampling distribution of the ratio of two variances |
|
|
Term
|
Definition
| the between groups variance divided by the within groups variance |
|
|
Term
|
Definition
| you need not correct alpha for multiple comparisons, typically only make a few critical comparisons, compare each pair of means using t-test, the within groups variance is the best estimate of the variance to use in estimating the standard error of the difference between two means |
|
|
Term
|
Definition
| you should adjust the alpha level for multiple comparisons, experiment-wise error rate (control your hypothesis testing so that the error rate given all post hoc tests is alpha), comparison-wise error rate (control multiple comparisons by adjusting the comparison wise error rate (alpha is set for each comparisons)) |
|
|
Term
|
Definition
| pooled estimate based on the variability around the group mean within each group |
|
|
Term
|
Definition
| variance of the group means around the grand mean |
|
|
Term
|
Definition
| increases with N, increases with effect size (greater differences between the group means), decreases as the population variance increases |
|
|
Term
|
Definition
| occurs when the effect of one factor is not the same at all levels of the other factor |
|
|
Term
|
Definition
| x2, asymmetric distribution, df is (r-1)(c-1), used for hypothesis testing, only do non-directional alternative hypothesis except when df=1, considered a two-tailed, non directional test |
|
|
