click below

click below

Normal Size Small Size show me how

# I/O Psych 542

### Exam 2 Part 2

Question | Answer |
---|---|

Main Effect | a result occurring in a factorial design when one of the independent variables has an influence on the dependent variable |

Interaction | the statistical result achieved in a factorial design when two independent variables have an effect in combination that we do not see when we examine each independent variable on its own |

Mixed Design ANOVA | ANOVA w/both fixed/random factors. Within-subjects factor n repeated-measures is a random factor. W/infinite # of time points to use– use depends on conclusions we wish to draw. we don’t use all levels of the within subjects factor, its a random factor |

MANOVA (What it is and how it compares to ANOVA | Multivariate Analysis of Variance |

ANCOVA (What it is and how it compares to ANOVA) | Analysis of Covariance |

MANCOVA (What it is and how it compares to ANOVA) | Multivariate Analysis of Covariance |

Confidence Interval | an interval estimate, based on the sample statistic, that includes the population mean a certain percentage of the time, were we to sample from the same population repeatedly |

Effect Size | a standardized value that indicates the size of a difference with respect to a measure of spread but s not affected by sample size |

Cohen's D | a measure of effect size that assesses the difference between two means in terms of standard deviation, not standard error (we can only use this in situations in which we have two means – the same situations in which we would use a z test or a t test) |

R2 | statistic used for an ANOVA.Its the proportion of variance in the dv thats accounted for by the iv. Sometimes researchers use a very similar measure of effect size, eta squared. We figure out r squared by making a ratio: SSbetween / SStotal |

Statistical Power | a measure of our ability to reject the null hypothesis given that he null hypothesis is false |

Alpha | also the p level, & this is the chance of making a Type 1 Error. Increasing alpha will increase the statistical power (taking a p level from 0.05 to 0.10); however, this has the side effect of increasing the probability of a Type 1 error from 5% to 10% |

interpret an ANOVA table | |

Be able to describe the problems with small samples and why t-Tests may be important with small samples | As sample size gets smaller, we r less certain about what the pop. distr. really looks like, & the t distributions become flatter and more spread out. However, you can’t do a z test with small samples, the t test will give you some idea, at least (???) |

Be able to explain what ANOVA is and how it is used | An ANOVA is a test that can run more complex statistics than just a t-test. This is the test used for three or more different samples |

Be able to explain why ANOVA is used instead of using multiple t-Tests | It would be too many tests to run w/multiple. The more ind. tests run, the more likely u will have a Type I Error. its more complicated to have a three-group (more group) comparison, so it requires a distribution that can accommodate that complexity |

Be able to explain how the F statistic is calculated. | divide the between-groups variance by the within-groups variance. If the between-groups variance (the numerator) is much larger than the within-groups variance (the denominator), then we can infer that the sample means are different from one another. Ho |

Be able to use software to analyze data using a One-Way ANOVA | |

Be able to explain why Post-Hoc Tests may be necessary | We only run Post-Hoc Tests if there is statistical significance. not every pairing is statistically significant, so the post-hoc tests will tell us WHICH paring is statistically significant. tests are run after the ANOVA is calculated. ex.Tukey HSD. |

Be able to explain what a Two-Way ANOVA is and how it differs from a One-Way ANOVA | 1Way ANOVA has 1 nominal iv w/^ than 2 levels & a scale dv vs a 2W ANOVA has 2 nominal iv, regardless of their #s of levels, & a scale iv(1Way ANOVA –5 part.&their ratings of a Cheap, Mid-Range & High-End Car. 2WANOVA – 2 diff. drinks- grapejuice & water |

Between-Group Effects | Between-Group Effects – partitioned into Main Effects & Interaction Effects. For Factor A, take each marginal mean and multiply the squared deviation score by n and b (the # of levels of Factor B). For Factor B, take the marginal mean for B and multip |

Within Group Effects | Within-Groups Effects (Error??) – calculating the variability of individual scores around their group means, creating an estimate of variance |

2 types of Interactions | Quantitative and Qualitative |

Quantitative Interaction | an interaction in which one independent variable exhibits a strengthening or weakening of its effect at one or more levels of the other independent variable, but the direction of the initial effect does not change. |

Qualitative Interaction | an interaction of two (or more) independent variables in which one independent variable reverses its effect depending on the level of the other independent variable. |

Main Effects | occurs in a factorial design when one of the independent variables has an influence on the dependent variable |

Effect vs. interaction | 2Way ANOVAs make 3 F stat.: 1 for 1st iv, 1 for 2nd iv, & 1 for interaction between the 2 v. The F stat.for each of the 2 iv r main effects. •The effect of each iv on the dv is the main effect •The combined effect of 2 or more iv on dv is an interaction |

Be able to interpret table for a Two-Way ANOVA | |

Statistical Significance | The difference in the group means is likely NOT due to sampling error. |

Practical Significance | Are the differences large enough to have real meaning? |

Be able to distinguish between statistical significance and practical significance | Ex: men & women's IQs may differ if enough samples are used to rule out test error (Statistical Sig.), but is the diff. in IQs enough to have real meaning (Practical Sig.)? Practical Sig. is affected by the size of the difference in Statistical Sig. |

Be able to describe the relationship between Effect Size and Statistical Significance | Statistical Significance – shows that there is a significant and we can reject the null hypothesis. There is a chance. Effect Size – indicates the size of the difference. We can’t figure out effect size if it’s not statistically significant |

Statistical Power | probability that we will reject the null hypothesis when we should reject the null hypothesis – the probability that we will not make a Type II Error |

Be able to describe how Statistical Power relates to Type I and Type II Errors | TypeII Error-fail to reject null hypo. but its false– say we arent pregnant when we are). ^ the stat. power (alpha level of 0.05 to 0.10) has side effect of ^ chance of Type 1 error from 5% to 10%. (Type I Error- reject null hypo. but its correct |