GoConqr hat ein neues Design! Sag uns, wie es dir gefällt.

Kostenlos registrieren

Karteikarten

DATA #3

LAUREN BELLINI Englisch 56 Karten

Karten in diesem Stapel (56)

  1. 1What is a multi factorial ANOVA? how is it different to a normal ANOVA?a multifactorial ANOVA is where there is more than one independent variable.
    it is different to a regular ANOVA as normal ANOVAS only have 1 independent variable
  2. 2why are multi factorial ANOVAS useful?1. makes testing more efficient as testing multiple IV's
    2. can see effects and interactions between IV's that you wouldn't see by analyzing them separately
  3. 3There are 3 types of Two way ANOVAS ... what are they?Two way independent groups ANOVA
    Two way repeated measures ANOVA
    Two way mixed ANOVA
  4. 4Explain when you would use a Two way independent groups ANOVA. What are the assumptions?would use when you have 2 independent variables and different groups of participants experience each of the IV.
    Assumptions
    - homogeneity of variance
    - normality
    - independent cases
    - independent samples
    - DV must be interval or ratio
  5. 5Explain when you would use a Two way repeated measures ANOVA. What are the assumptions?when there is 2 independent variables and one group of participants only who experiences both IV's
    Assumptions
    - homogeneity of variance
    - normality
    - independent cases
    - sphericity
  6. 6Explain when you would use a Two way mixed ANOVA.When there are 2 independent variables and one IV is independent groups and given to different groups of participants and one IV is repeated measures and is given to both groups of participants.
  7. 7what are marginal means?are the means of each column or each row
  8. 8what is a main effect? when do we say we have a main effect?the effect of a single IV on the DV. Does not take into account any other IV's
    we have a main effect when the marginal means differ.
  9. 9what is an interaction? when do we say we have an interaction?Interactions are whether one level of the IV is interacting with the other IV.
    interactions show interactions between the 2 IV's.
    We have an interaction when there is a difference between the cell means.
  10. 10what do we look at to tell whether main effects/ interactions are significant?we look at the F statistic. We want F >1 to reject the null hypothesis which claims that any differences between means is a result of chance.
  11. 11If you have a significant F test for main effects what is the follow up testing you do to? why do you do follow up testing?Do follow up testing to tease out where differences lie within the marginal means when the main effect was significant.
    For main effects the follow up test is Tukeys.
  12. 12If you have a significant F test for interactions what is the follow up testing you do to? why do you do follow up testing?You do the follow up testing to investigate the nature of the interactions and to see which IV's are interacting.
    For interactions we use "simple effect contrasts"
  13. 13what do "simple effect contrasts" compare?used in significant interactions.
    Compare means of one IV for different levels of the second IV.
  14. 14what effect size do we use in main effects/ interactions?partial eta squared
  15. 15what does partial eta squared calculate?it estimates the specific effect size of one independent variable when the effects of the other IV's and interactions are partialed out.
  16. 16What is MANOVA? When is it used? what does a significant result tell us?MANVOA is a multi variate analysis of variance.
    It is used when there are 2 or more dependent variables and those variables are correlated and conceptually related.
  17. 17what are the assumptions of MANOVA?- univariate and multivariate normality
    - univariate and multivariate variance
    - independent observations
    - DV'S MUST BE CORRELATED AND CONCEPTUALLY RELATED.
  18. 18what statistical test do we use for MANOVA? what does a significant result tell us?Wilks Lambda (W)
    when it is significant it tells us that at least 2 of the DV'S are having an effect on the mean that is greater than what we would expect to see by chance alone.
    But it does not tell us which of the DV'S are doing so- need to do follow up testing
  19. 19what is the follow up test for a significant Wilks lambda?one way ANOVA and a bon ferroni correction.
  20. 20what is a bon ferroni correction? why are they used and important?Bon ferroni correction is an adjustment to the alpha level when we are making multiple pair wise comparisons. We adjust the alpha level to control for Type I error.
  21. 21what is regression?is an analysis of the relationship between variables.
  22. 22what is simple linear regression? why is it called "simple"allows us to determine the relationship between two variables X and Y
    called simple because it deals with only one independent variable
  23. 23what is the equation for a simple linear regression?Y = C + BX + ERROR
  24. 24what does the Y C B X ERROR stand forY= dependent variable/outcome
    C = the intercept of X and Y
    B= the regression co efficient
    X= the independent variable/ predictor
    ERROR= the residual
  25. 25what is the symbol used in both simple and multiple regression. In simple regression what does it represent?R
    represents the relationship between
    X and Y
    varies between 0 and 1.
    0= no relationship, 1 = perfect
    interpreted the same as pearsons r
  26. 26R is always positive. How do we determine whether it is a positive or negative relationship?we look at the sign of the regression co efficient (B) for X
  27. 27what does R squared in simple regression represent? what is it similar to in normal bivariate correlation?R squared in simple regression represents the proportion of variance in Y that is accounted for by X
    same as coefficient of determination.
  28. 28what ratio do we look at to determine whether there is a significant linear relationship between variables? what does it tell us?look at the F statistic this tells us that the slope of the regression line is significantly different from zero
  29. 29what does the null hypothesis claim in simple regression?that the population slope will be zero and the independent variable is not accountable for any changes in the DV
  30. 30What are the assumptions of simple linear regression? (4)- DV and IV must have normal distribution
    - linear association with IV
    - residuals normally distributed
    - homoscedasity
  31. 31what is homoscedasity?says there must be equal variance of Y at every level of X
  32. 32Discuss what partial correlation is? (pr)when we have a significant relationship between X and Y it is possible that other confounding variables are the cause of this significant finding.
    pr moves from both X and Y any association with a confounding variable
    and tells us whether variance between X and Y is significantly greater than zero after the overlapping variance of a control variable has been removed.
    Thus a significant pr tells us that any variance in the DV after controlling for the confound must be uniquely explained by the IV alone.
  33. 33In a venn diagram,what does a large overlap indicate?the greater the overlap of the circles the greater the variance in one variable is explained by the other variable.
  34. 34what is multiple regression? what does it assess?allows us to examine how multiple independent variables/ a set of independent variables are related to the dependent variable.
    it assesses the linear relationship between a DV and a set of IV's
  35. 35what are the 4 features that need to be checked before using multiple regression?1. IV's need to be continuous
    2. number of cases needs to be 10+
    3. check for univariate and multivariate outliers
    4. check for multicolinearity and singularity
  36. 36what are the assumptions of multiple regression (4)- DV has normal distribution
    - linear relationship between IV and DV
    - homoscedasity
    - normality of residuals
  37. 37what is the R in multiple regression a reflection of?R quantifies the degree on linear association between the DV and the set of IV's
  38. 38In multiple regression, what is the null hypothesis claiming?that there is no linear relationship between any of the IV's singularly or in combination with the DV.
  39. 39what does a significant F value tell us?that at least ONE of the population regressions in greater than zero.
  40. 40The F is omnibus, it does not tell us which particular IV is significant... what can we do to find out?we look at t scores of each IV. A significant t score tells us that the IV is significant and does explain a proportion of variance in the DV
  41. 41Even when we know which IV's are significant, we don't know how important this significance is... what are the 2 ways we can assess how great/important the effect of IV's are on the DV?we assess importance of IV's through
    1. standardised regression coefficients
    2. squared semipartial correlation
  42. 42For standardised regression coefficients, what do we look at to determine importance?look at the beta scores. The higher the value the more impact that IV is having on the DV. close to zero means IV is having little effect on DV
  43. 43How does squared semipartial correlation help us evaluate importance of IV's?( sr squared )
    squared semipartial correlation corresponds to the proportion of total variance in the DV explained by the IV.
    The larger the sr2 the more variance in the DV is explained uniquely by that IV.
  44. 44what exactly is a significant IV telling us?a significant IV says that there is a linear relationship between the IV and the DV and that the IV can explain variance in the DV that is not just a result of chance.
    If the IV is not significant, it means there is no linear relationship between the IV and DV.
  45. 45what are the 3 types of multiple regression?standard multiple regression
    hierarchal multiple regression
    stepwise multiple regression
  46. 46briefly describe standard multiple regressionAll IV's are entered into the mR equation at the same time
    each IV is assessed in terms of its unique contribution in explaining variance in the DV
    Any variance that is explained by more than one IV is called shared variance
  47. 47briefly describe hierarchal multiple regressioncontrols the order that IV's are entered into the MR equation on a theoretical basis.
    high priority variables entered first
    each other IV entered after this and assessed in terms of how much additional variance they contribute.
  48. 48briefly describe stepwise multiple regressionalso controls the order that the IV's are entered into the MR equation but this is done on a statistical basis.
    IV with best prediction is entered first
  49. 49when do we use discriminant function analysis?when we want to determine whether a set of independent variables (predictors) can correctly classify a sample of participants (who come from different groups) into those groups.
  50. 50what statistical test do we use for discriminant function analysis? what does a sig result mean?we use wilks lambda
    it assess whether the IV can predict grouping better than chance. If it is significant we can reject the null, that the level of prediction obtained occured by chance
  51. 51Every analysis with the word "independent" in it (independent t test, one way independent ANOVA, two way independent ANOVA) has the same 4 assumptions what are they?- homogenity of variance
    -normality
    - independent cases
    - DV must be interval/ ratio data
  52. 52Every analysis with the word "repeated" in it (one way repeated ANOVA, two way repeated ANOVA) has the same 4 assumptions what are they?- Homogenity of variance
    - normality
    - DV interval or ratio data
    - spherecity
  53. 53what are the assumptions of the paired samples t test?homogenity of variance
    normality
    DV must be interval/ratio rata
  54. 54what are the assumptions of MANOVA? (5)DV'S MUST BE CONCEPTUALLY RELATED
    univartiate and multivariate normality
    univariate and multivariate variance
    independent observations
  55. 55what are the assumptions of ANCOVA?- homogenity of variance
    - normality
    - independent cases
    - linear relationship between the DV and the covariate
    - homogenity of the regression slope
  56. 56what are the assumptions of multiple regression?- normality
    - linear relaitonship between IV and DV
    - homoskedacity
    - residuals normally distributed

Informationen

Beschreibung
Karteikarten am DATA #3, erstellt von LAUREN BELLINI am 13/11/2015.
Erstellt von
LAUREN BELLINI
Aufrufe
2
Bewertung

Mehr zu Factorial ANOVA and Interactions

Feedback

Worum geht es?

Wie gefällt dir das neue Design?

Dein Eindruck