click below

click below

Normal Size Small Size show me how

# PSY 121

### Chapter 12 and Chapter 13

Term | Definition |
---|---|

effect size | general term that refers to the strength of association between variables; a type of correlation (pearson r is one indicator) general guide: 0.15 = small effects, 0.30 = medium effects, 0.40 or above = large effects |

standard deviation (SD) | average deviation of scores from the mean. SD = small; scores are close to mean; SD = large; scores are far from the mean. Used with interval and ratio scales |

68% rule | fall within the +/- 1 deviation units from the mean |

bar graph | use a separate bar for each piece of information. Used with nominal or ordinal data. When drawing bar graph, the bars should not touch each other |

polygon (line graph) | uses a line to represent frequencies, useful with interval and ratio scale variables |

pie chart | divide a whole circle into sections that represent relative percentages. Useful with nominal scale data |

scatterplot | used to visualize the relationship between the variables (correlation coefficient of +/- 1.00 - positive and negative) |

null hypothesis (Ho) | population means equal, the observed difference is due to random error. Logic: to be able to reject the null hypothesis --> good thing. Ho = mean of the treatment group = the mean of control group |

research hypothesis (H1 or Ha) | aka alternate hypothesis. Population means are not equal. The IV had an effect on the DV. Logic: to fail to reject (to accept) the research hypothesis. H1 = mean of treatment group will not equal the mean of the control group |

t-value in regions of rejection (ROR) | reject Ho, significant result |

t-value NOT in ROR | fail to reject Ho, not significant |

t test | used for single experiment (one IV and two levels - 2 groups). Used to examine whether 2 groups are significantly different from each other |

F test | ANOVA. More general and common than the t-test. Used to determine if there is a significant difference between 3 or more groups OR to evaluate the results of factorial designs |