Comparing means 0.0 / 5 ? PsychologyComparing meansUniversityNone Created by: francesca_321Created on: 25-04-17 15:22 What is the aim of comparing means? The aim of a standard experiment is to assess whether a particular manipulation alters scores 1 of 39 What do we end up comparing? we end up comparing the means of scores in condition A and B, Do females and males IQ differ 2 of 39 What is meant by statistically significant? Unlikely to occur by chance 3 of 39 What is the opposite of statistically significant? non-significant 4 of 39 What does a T test depend on? The size of the difference between the conditions 5 of 39 What is the standard error of the mean? Is the difference 6 of 39 What happens once we calculated T? We can look up the associated p value in the table 7 of 39 What do we need to know? The degrees of freedom 8 of 39 For example? N guests are coming to dinner how many choose their own seat? N-1 9 of 39 What is degrees of freedom closely related to? N but depends on the test 10 of 39 What do degrees of freedom change? The probability distribution of test statistics -> few degrees of freedom and many degrees of freedom 11 of 39 What are degrees of freedom related to/ Number of subjects 12 of 39 What is the relationship different for? different tests (choice of seats depends on seating methods) 13 of 39 What is the DF needed for? To work out p from t 14 of 39 What is an independent samples T test used for? Between groups design 15 of 39 What are three other names for a independent samples t test? Unpaired means t-test, unrelated samples t-test, between grups t-test 16 of 39 What is it based on? The size of the difference between tohe cnditions, the SEM of that difference 17 of 39 What is the equatin to work out the mean of differences? (mean of Grp 1) - (mean of Grp2) 18 of 39 How do you work out the SEM of means ? square root (variance 2/N + variance 2/N 19 of 39 what is t? the mean difference divided by the standard error of the difference 20 of 39 What is the t-test based on? Means and standard deviation assumes that these do adequately describe the data 21 of 39 What is the first? DAta must be on a ratio or interval scale 22 of 39 What is the second? Data must be normally distributed 23 of 39 What is the 3rd? Data in one group must be independent of data in the other 24 of 39 What was the 4th? Data in the two groups must have similar variance 25 of 39 Where is this found in SPSS? Analyse/compare means/ independent samples 26 of 39 What happens if we want to know if p We need to calculate p (alphas is fixed. 1. calculate some test statistic, 2. from the test statistic and d.f claculate p 27 of 39 What is a paired means t test also used for? Within subject design 28 of 39 What is a paired means t test also known as? Related samples t test and within subjects t test 29 of 39 What are subjects used as? Their own control more powerful: likely o find an effect 30 of 39 What is the equation? Mean of the differences for the pairs divided by standard error of differences between pairs 31 of 39 What do you have to do first to calculate T? check to see if data is normally distributed 32 of 39 How? Graphs/legacy dialogs/histogram/boxplot 33 of 39 What is needed? No need to cluster for this data, one column for multiple groups, multiple columns 34 of 39 What is the label cases by? Determine how many outliners are labelled 35 of 39 What does the output window show? How many subjects had a score 36 of 39 What do Box plots show? Data with all non parametric markers 37 of 39 What is the stuff in a box plot show? Bar is median, box is interquartial range, whiskers are full range excluding outliners, cirles are cmoderate outliners, stars and extreme outliners 38 of 39 Where is it found in SPSS? Analyse/comparemeans/paired sample t-test 39 of 39
Comments
No comments have yet been made