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
Stats quiz 3 part 3
| Question | Answer |
|---|---|
| Parametric tests | lie 2 sample t-test which compares means |
| Nonparametric tests | Like the Mann-Whitney test, which compares medians |
| Independent (non-paired) populations | are those in which there is no natural pairing or link between specific individuals in one population and specific individuals in the other population. In other words, the data are not paired by observation across the two populations |
| Dependent (paired) populations | are those where there is some sort of natural pairing or link between specific individuals in one population and specific individuals in the other population. In other words, the data are paired by observation across the two populations |
| Which two tests are used for independent populations | 2-sample t-test and the Mann-Whitney test |
| Testing for normality | When our sample size is small (n<25) we need to verify whether the data come from a normally distributed population before deciding which test to use |
| Null hypothesis for normality | the data are from a normally distributed population |
| alternative hypothesis for normality | the data are NOT from a normally distributed population |
| Tests for normality | Shapiro-Wild test, Anderson-Darling test, Kolmogorov-Smirnov test |
| When to use 2 sample t-test | If both populations are normal or the sample sizes are both greater than or equal to 25. The samples are independent (non paired) |
| When to use the Mann-Whitney test | If either population is not normal (this is a non parametric alternative) |
| Hypothesis test for 2-sample t test | We are interested in comparing two population means so the null and alternatives will be statements about the difference of the two means in the population. Generally we are testing to see if the difference is 0. |
| Confidence interval for the Difference of Means | If 0 is contained in the interval, then we can conclude no significant difference in the population means |
| for t table we always round | down for df |
| Mann Whitney (AKA Wilcoxon Ranks Sum Test) Conditions | The samples are independent. The sample data are at least ordinal. Use when the population distribution is NOT normal. Deals with MEDIANS |
| Null and Alternative for Mann Whitney | In terms of the population medians. Testing the difference between the two. |
| Populations are dependent when | There is a natural pairing between observations. Each observation in one sample is linked to a corresponding observation in the other sample |
| Examples of dependent populations | before and after, matched pairs, repeated measures of the same subject |
| When to use paired t-test | data are approximately normal and data is dependent. There are at least ordinal |
| When to use Wilcoxon Signed Rank Test | Data are not normal or ordinal, there are dependent populations |
| Matched Pairs test (paired t test) | a parametric test used to carry out hypothesis tests of a difference of two dependent population means. |
| Conditions of Matched Pairs Test | The samples are dependent. The differences of the paired data are approximately normal, the number of pairs is greater than or equal to 25. RUN shapiro on differences |
| Does the matched pairs test deal with means or medians | population means |
| Confidence interval for the mean difference (paired t-test) | The interval provides a range of plausible values for the true average difference. If it includes 0, there is no significant difference between the two population means |
| Does the Wilcoxon signed rank test deal with means or medians | median differences. Generally testing to see if the median difference is 0 |
| Important note for Wilcoxon signed rank test | ignore the differences that equal 0, rank absolute differences |
Created by:
user-1996284