Analysing 3 or More conditions 0.0 / 5 ? PsychologySTAtisticsUniversityNone Created by: francesca_321Created on: 26-04-17 17:35 What are the issues with multiple T-tests? If each test has an alpha of 5% the likelihood of error becomes 15% which is too high 1 of 46 What is the bad solution? Reduce our alpha to 1.6% combined error in 5% but reduces power 2 of 46 What is an ANOVA? extension of the t-test but doesnt reduce power 3 of 46 What are the three types of ANOVA? Analysis of variance, multivariant analysis of variance and analysis of covariants 4 of 46 When are MANOVA's used? When there are multiple DVs 5 of 46 When are ANOVAs used? Correlations 6 of 46 Who was ANOVA invented by/ Fisher 7 of 46 What does it do Compare the amount f variance explained by our experiment with variance that is explained by our experiment with variance that is unexplained 8 of 46 What are the scores? sampled randomly and are independent 9 of 46 What do ANOVAs have? Normal distribution, roughly eual number of pps in the groups, equal variance for each condition 10 of 46 What is the between group ANOVA? The explaine variance is varaince between groups, unexplained is variance within a group 11 of 46 What is the calculation refered to? Calculation is referred to as the mean 2 eror ANOVA is based on the F ratio 12 of 46 What is the F equation? MS between/MS within 13 of 46 What is the MSbetween? The effect 14 of 46 What is the Ms within? The noise 15 of 46 What are the degrees of freedom? Dehrees of freedoms within conditions, residual degrees of freedom 16 of 46 What is ANOVA? Critical values require 2DF values one for each aspect of variance 17 of 46 What is a pair wise correlation? ANOVA tells us whether groups differ or not 18 of 46 How can you tell? Run multiple comparisons, some of these are planned comparisons and some are post hoc tests, correct for multiple comparisons 19 of 46 How do you find a single IVs? Analyse/comparemenas/one-way-anova 20 of 46 How do you test with multiple variables? Analyse/GLM/Unvariate 21 of 46 When do you run Post HOc Test ? If main test found a significant result 22 of 46 What method is used? Bonferroni's method, uses a test and divides alha by some values based on t tests, very strict safe not powerful 23 of 46 When do you run a unvariate ANOVA? One DV 24 of 46 What is variance based on? Variance explained by experiment (the effect) residual variance that cant be explained 25 of 46 What is the variance between groups? Explained 26 of 46 What is the variance within groups? Unexplained 27 of 46 What is a repeated measures also called? Variances between condition, variance between subjects, residual unexplained variances 28 of 46 What is between group? variance between subjects fell under category 'unexplained' 29 of 46 What is between group? Variance between conditions, residual variance 30 of 46 What is repeated measures? REsidual noise, inter subject variance (control between conditions) 31 of 46 In both cases what do we want to know? WE want to know about the ratio of the effect to the noise 32 of 46 What is ratio? Larger after removing inter subject variance 33 of 46 How do you work out MSnoise? Mstotal - mseffect-msindiff 34 of 46 What is MsIndDiff? The variance betweeen subjects within a condition 35 of 46 What is Ms total? The variance of all subjects in all conditions 36 of 46 What is Spss? Plot data, Graphs/error bar plot 37 of 46 What is a simple plot? We only have IV 38 of 46 What are summaries of variables? Vompare collumns enter number of levels 39 of 46 What is multifactorial ANOVA? Factors can all be within subjects, all between group or mixed design, main effect: one of factors affect DV, i 40 of 46 What is interaction? The effect of one factor depends on the presence of another 41 of 46 What is 2X2 ANOVA? multifactoria ANOVA is a single test, 2 levels no poc test needed, 2X2 is a single test 42 of 46 What requires parametric data? Parametric tess: skewed or non normal distribution, continous data, different groups more statistically power 43 of 46 What is non parametric test? Small sample size and ordinal and ranked data 44 of 46 What is the family Wise error? The error rate across statistical tests conducted on the same experimental data 45 of 46 What is the F ratio used to test? The overall fit of a regression model to a set of observed data 46 of 46
Comments
No comments have yet been made