Study Guide for Final Exam Chapter 12 Dependent Samples T tests When to use dependent t Independent o 2 different unrelated groups o Between subjects effects Dependent o Same group of participants tested MORE THAN ONCE o Repeated measures design Within and Between subject designs be able to identify each and know the pros cons of each Independent between o Differences that are based on the effect of our variable o Pro fast o Con cohort effects Dependent within o Differences NOT based on the effect or our variable Chance Error Individual differences o Pro each player serves as his own control o Con takes 9 years to do study Know the steps for executing a dependent t test and making a decision 1 State null and research hypothesis 2 Set the risk level alpha 0 05 3 Select the appropriate test statistic 4 Compute the test statistic called the OBTAINED VALUE 5 Determine the CRITICAL VALUE minimum value needed to reject the null Need to know alpha two tailed test degrees of freedom 6 Compare the obtained to the critical 7 Make decision Reject or not reject null 8 Interpret Degrees of freedom for dependent t n 1 Interpreting reporting dependent t tests Interpreting is a significant difference between the pre and post tests Reporting t 4 4 47 p 05 o T is the type of test o 4 is the degrees of freedom o 4 47 is the obtained value o P 05 tells you null was rejected Chapter 13 One way ANOVA When to use one way ANOVA Simple one way ANOVA o When you have ONE factor with more than 2 levels Repeated measures ANOVA o When you test the same participants more than twice Factorial ANOVA o When you have more than one factor Factors and levels Factor the variable that designates the groups to be compared Level the different groups within a factor F Tests you should understand what an ANOVA is conceptually as well as the steps to calculating it Basic Logic of ANOVA First we find the total variability for the entire set of data Then we try to partition the variance into these 2 sources o Try to figure out how much of the total variance is the effect of our variable and how much of the total variance is due to other factors If all the scores for one group are the same then all of the variability is BETWEEN the groups o If that happened we could probably all agree the difference is meaningful All variance between reject null meaningful difference between groups All variance within null is true no difference between groups The F Tests F variance between groups variance within groups o Dividing the amount of difference between groups by the amount of difference within groups o group effect difference due to chance differences due to chance The obtained value is F F is based on sources of variation rather than mean differences The bigger the F value the smaller the p value o Meaning we reject the null when F is big F is big when o The numerator is much bigger than the denominator The between groups variance should be larger than the within groups variance Which means that there is a big group effect When there s no effect of our variable the numerator and denominator are roughly equal and F Omnibus tests test whether the explained variance in a set of data is significantly greater than is small close to 1 o Null is true Calculating sum of squares You should be familiar with the ANOVA table What does it mean when you say ANOVA is an omnibus test the unexplained variance overall Chapter 14 Factorial ANOVA What is it and when do you use it One Way used for o Single factor experiments only one IV Simplest experiment design Difficult to isolate only one variable that causes a behavior Factorial used with o Factorial designs 2 or more IVs Factorial notation system e g 2x2 2x3 A way to summarize the design of the study numerically 2x2 design o The number of numbers is equal to the number of factors variables o Actual numbers equal to the number of levels for each of the factors Two factors Both have 2 levels 2x3 design o 2 factors o One has 2 levels one has 3 levels Main effects vs interactions What are they conceptually o Main effects The overall effect of a single variable The effect of ONE variable ignoring the levels of the other variable There are potentially as many main effects as there are factors o Interactions When the effect of one factor depends on the level of another factor How can you spot them with the means o Main effects Compute row and column means that combine the data for all cells that are at one level of a variable collapsing Won t tell you it it s significant but will give you a quick idea of whether there is a main effect or not o Interactions Calculate the differences across the rows columns Always do same groups first See whether the differences differ Occur when one line is higher than the other Or one side of graph is always higher than the other i e big slope How can you spot them in graphs o Main effects o Interactions Occur when the lines are not parallel How you can you spot them in SPSS output o Split the variance into sources Possible sources The main effect of any sources The interaction between any 2 of the factors o The p value tells us whether any of those particular effects is significant or not Each source is tested separately to get values Chapter 15 correlations What are correlations and when do you use them Correlations ask how do changes in the value of one variable influence the values of another variable Correlation coefficients are a way of assigning a single number to describe these relationships o Abbreviated as r o Range from 1 to 1 Scatterplots Strength Plot correlations One variable on X axis other variable on Y axis How perfect correlation is the pattern of the relationship Think of variability within groups Closer to 1 or 1 closer to everyone fitting the pattern If r 0 no relationship If r 1 or r 1 perfect relationship Direction Positive 0 1 Negative 1 0 o One goes up other goes up o One goes up other goes down Testing correlations H0 r 0 H1 r 0 o No relationship o There is a relationship Use table to decide o 1 vs 2 tailed use 2 tailed o Set alpha 05 o Set df n 2 Reporting correlations Ex correlation between height and weight R is 0 81 with 20 df R is an obtained value compare to critical r s to see if it s significant n sample size or of pairs of pairs of variables First 2 points will always be perfect correlation every other is free to vary o Critical value at 20 df is 0 4227 o Reject null r 20 0 81 p 05 A …