- Information
- AI Chat
Lecture Notes Rest of Sem
Abnormal Psychology (PSYC 3303)
University of Colorado Boulder
Preview text
Ch 16 One way ANOVA: one Independent variable w 3 or more levels - Same hypothesis as independent t test but 3 variables (independent variable categorical, dependent continuous) Pr(>F)= predicted value Significant difference only of less than. TukeyHSD-posthoc and Pairwise t test are the same but just depends on when/if you wanna use
Factorial (two way) ANOVA: 2 (or more) independent variables (factors)
Within: sum of squares for each group added together DF: # groups - DF within: df added together
Mean Square: SS/df (own row) F: mean sq/ mean sq residual
For labapalooza: - Pearson's r(r statistic is cor output, compare that to p (whether less than .05), if negative then direction changes) - linear regression(don’t need to look at P values, x value given we find y) (in summary identify y and x and then plug in equation)(y=mx+b, y=given x *number under intercept+ intercept ) - single sample t (has to include t , p , and if means are going opposite way describe that too, still address mean either way) - Paired sample t(include t, p , and compare means), - independent groups t (look at p value)(if significant list t, p, and means) (if not list t and p only) - one way analysis variance (tukeyhsd, or pairwise)(f statistic for significance ), If not significant don’t do any follow up, if it is do pairwise (if there's already a prediction) or tukey HSD (if no
prediction) Statistic? T value, F(in anova) Hypothesis? P
- factorial analysis of variance (dependent variables have to be categorical, f statistic and p value) (look at P values if less than .05=significant) none of the above (people who own a pet are more likely to own an iphone than people who don;t own a pet (two categorical variables)), or we never learned how to do this ( or z score)
CH 17
I think that young adults who are working watch more screen time than young adults who are students. However, I think that this is only true if those working young adults have a high school diploma. If they do not have a high school diploma, I think that working young adults watch as much screen time as students. → Screen Time depends on employment and education (independents) Main Effect: Looks at difference between the levels of one and only one factor, compares marginal means , have as many as u have factors Simple Effect: means inside the cells, You have as many interactions as you have combinations of factors Number of cells in interactions measured by combination of levels (how many means represented) Interactions: the way the factors can be combined (if its 2x2 then theres 4 interactions)
For final: Hypothesis- - Z Score (1 continuous variable (1 person) compared to the population)(ex. John ate 25 nachos, is John weird compared to the rest of the pop?) - Z test (everybody in the group compared to pop (i think CU boulder students eat more nachos than the rest of the pop)) (Mu and sd pop given ) - Binomial (1 categorical variable that has two levels)(this group compared to pop ex. I think CU boulder students are more likely to own a car than the population) - Pearson's r (2 variables, general linear relationship (up or down))(ex. the more hours you watch netflix the more nachos you consume) - T test - Linear regression (raw value x given, predict raw value of Y, ex. I know someone who watches
Lecture Notes Rest of Sem
Course: Abnormal Psychology (PSYC 3303)
University: University of Colorado Boulder
