O GoConqr está de visual novo! Conte para nós o que você achou.

Registre-se grátis

Flashcards

Sem título

James Lafferty Psychology Inglês 108 flashcards

Flashcards do conjunto (108)

  1. 1Discrete VariablesWhere the variables are separated in to clear categories
  2. 2Continuous VariablesWhere the variables can be on a continuum, including 0-infinity
  3. 3IV, DV + CovariateIV = The variable you manipulate
    DV = The variable you measure
    Covariate = A variable that tried to explain external factors that could impact on your relationship between your IV and DV
  4. 4What are the four levels of measurement for variables (Scales)1) Nominal
    2) Ordinal
    3) Interval
    4) Ratio
  5. 51) NominalHave two or more categories, but no kind of order or numeric value
  6. 62) OrdinalVariables can be ranked in order.
  7. 73) IntervalThe scale has values of equal intervals which mean something. E.g. temperature or IQ scores
  8. 84) RatioThe same as interval scale, except zero means =does not exist. E.g. weight in kg
  9. 9interquartile range (not THE range)The different between the upper and lower quartile of a data set
  10. 10Standard DeviationThe average deviation of scores in your data set from their mean score
  11. 11Type I ErrorRejecting the null hypothesis when its true in reality
    When you conclude you've got a stat. sign. effect when you haven't
  12. 12Type II ErrorFailing to reject the null hypothesis when its false in reality
    When you conclude you have NOT got a stat. sign. effect when you have
  13. 13One-tailed hypothesisRelationship between variables is in a specific direction only
  14. 14Two-tailed hypothesisRelationship between variables in in any direction
  15. 15Standard ErrorSame as SD, only difference is that it applies to the SAMPLE, rather than POPULATION
  16. 16Kolmogorov-Smirnov testStatistical test of Normality
  17. 17SkewnessThe extent to which a data distribution deviates from symmetry.
  18. 18Z-ScoreA measure of the dispersion of the data.
    Describes a raw scores distance from Mean. Its a Standardised score [Standardised Normal Distribution]
    Converts to Standardised Norm. Dist.
  19. 19Degrees of freedomThe number of values in a data distribution that are free to vary, whilst keeping the result the same.
  20. 20T-ScoreLike z-scores, tells you how many standard devs. a score is from the mean, but is used when you DONT know the population SD/
  21. 21Effect sizeThe size of your effect
  22. 22Correlation Coefficients (r)Reported on a scale between 0 and 1, with +/- indicating the direction of the relationship
    0.1-0.3 = small effect
    0.3-0.5 =medium effect
    0.5+ =large effect
    standardised way of expressing direction and magnitude of the relationship
  23. 23Cohen's DAn effect size used to indicate the standardised difference between two means.
  24. 24ANOVAAnalysis of Variance
    Used when examining diff. between 2+ sets of scores
  25. 25eta2 ["n2"]The effect size explained in an ANOVA
    Estimates proportion of variance in the DV explained by the IV
  26. 26One way ANOVAANOVA with one IV
  27. 27Two Way ANOVAANOVA with 2 x IVs
  28. 281 or 2 way ANOVA, Between or Within, but what about a Mixed ANOVA?Mixed ANOVA = when one level of your experiment is between whilst the other is within
  29. 29Partial Eta2 [Partial "n2"]In 2 way ANOVAS and with 2 IVs, more accurate than the n2
    [One Way ANOVA] = n2 = partial n2
  30. 30Things that are good for statistical power?1) Larger effect sizes = Higher power
    2) Having an a.level closer to 0.05
    3) Fewer SDs away from mean
    4) One tailed Ho Tests ensure more power than 2-tailed
    5) Larger sample sizes
  31. 31Calculating sample sizes1) Base it on previous research/studies
    2) Aim for 0.8 effect size
  32. 32simple Linear RegressionA test used to check how your IV predicts your DV
  33. 33Residual/sDifferences between the actual scores and predicted scores
  34. 34The regression line1) The line of best fit
    2) A line that passes through the data points incurring the smallest residuals
  35. 35sum of squares regression/least sum of squares regressionMost common SPSS use of linear regression; by squaring all the residuals then adding them, you minimise the impact of adding all the residuals together.
  36. 36the y-interceptWhere the regression line touches the y-axis
  37. 37The regression equation: yThe DV [ on the y axis]
  38. 38The regression equation: xThe IV/ predictor variable [on X axis]
  39. 39The regression equation: aa = the y-intercept (where the regression line touches the y-axis)
  40. 40The regression equation: bThe slope or gradient of the regression line
  41. 41r2 (r squared)R squared, so therefore represents the amount of shared variance
    Tells you amount of variance in DV explained by the IV; the amount of the DV score that is explained by the IV variable
  42. 42RCorrelation coefficient: between your two variables. Measures how good the regression model is. +/- 0-1
    Tells you how close to the data points your model is
  43. 43adjusted r2 (r squared)R2 adjusted for number of predictor variables
  44. 44standardised cofficent BETAAllows you to predict the increase of the DV in Stan. Devs/ if the IV increases by 1 Stan. Devs.
  45. 45Multiple linear regressionLinear regression with multiple predictor variables/ IVs
  46. 46Simple linear regression equationy = a +bx
  47. 47Multiple linear regression equationy = a + b1x1 +b2x2 + b3x3 [ 3 predictors example]
  48. 48Unstandardized coefficient "B" "Constant"Tells you the value of "a" in the equation
  49. 49Unstandardized coefficient "B" [underneath]Tells you the value of b in the equation!, one for each predictor variable
  50. 50The required assumptions of Regression analysis? (5)1) Normality
    2) Linearity
    3) Outliers
    4) Multicollinearity
    5) Type of data
  51. 51Linearity?Assumption: Needs to be a linear relationship between your IVs and DVs
  52. 52Outliers assumption?Data needs to be rid of any influence outliers which warp the regression model. They can be removed
  53. 53Absence of Multicollinearity?When predictor variables are highly correlated with other PVs in the model
    It means that we can't tell which one is causing the variance
  54. 54VIF?Variance inflation factor; measure of multicollinearity
    VIF over 5 = problems with multicollinearity
  55. 55HomoscedasticityThe residuals should be equally distributed along the regression line
  56. 56Type of data for regressionDv must be measured at interval or ratio level OR dichotomous ( a or b only)
  57. 57Discourse AnalysisResearch method: Studies use of language; aims to identify PURPOSE and EFFECTS of different types of language
  58. 58Narrative TherapyTherapy based at looking at how to change a problem-saturated self-narrative of a client to a PRODUCTIVE self-narrative. Looks at reassigning their beliefs and values
    Identity is formed by a narrative of a clients life
    tries to disentangle "problem saturated" narratives
  59. 59Acceptance and Commitment TherapyTherapy that uses acceptance & Mindfulness and commitment strategies to improve psychological flexibility
  60. 60Pearsons Chi Squared ( X2)Calculates how likely the association found in a contingency table occurred by chance
    Can tell you if theres a stat. sig. association between two variables
  61. 61Calculating Expected Frequency(Row total x column total) / grand total
  62. 62Yates continuity correctionSlight adjustment to chi squared (X2), but only used on 2 v 2 contigency tables
  63. 63Fischers exact testSlight adjustment to chi squared (X2), but only used on 2 v 2 contigency tables
    Specifically used when expt. freq. of a contingency in the table is less than 5
  64. 64Cramers V coefficientAfter Chi square (X2) has checked sig. association between variables, Cramers V is a post hoc test than determine strengths of association
    Between 0 and 1, 1 being a strong association
  65. 65Odds ratioOne odds divided by the other
    Tells you likelihood of one happening instead of other
  66. 66Phi coefficientCalculates the strength of the association, between 0 and 1
    Takes in to account sample size
  67. 67Cramers V vs Phi ?Cramers V is similiar BUT can be used for tables with more than 2 x 2 columns/rows
  68. 68McNemar testChi square X2 test used on a 2 x 2 table but when the data represents change over time
  69. 69Independent groups designWhen you're interested in difs. between independent groups
  70. 70T-testParametric test to compare independent means
  71. 71Mann-Whitney test/ wilcoxon-mann-whitney test ("U")Non-Parametric test to compare independent means, when data is measured ordinally
  72. 72Correlational designWhen you observe the relationship between existing variables as they occur naturally. E.g. correlation or regression
  73. 73Experimental designWhen you manipulate 1 or more IVs to observe influence on DVs
  74. 74Repeated measures designWhere youre interested in change within the same group of participants
  75. 75Quasi-experimental experimentsExperiments that use naturally occuring groups rather than specific or random allocations
  76. 76Levene test for equality of variancesA test of homogeneity of variances: tests that all groups have the same or similiar variance
  77. 77Assumption of Homogeneity of varianceAssumption that the variance of your DV should be similar for both groups
  78. 78Repeated measures design is an other name for......within-groups design
  79. 79F-RatioBetween groups mean square/ within groups mean square
  80. 804 assumptions of 1 way ANOVA1) Homogeneity of variance
    2) Normality
    3) IV = discrete variables
    4) DV = interval or ratio measurement
  81. 81Krushal-Wallis TestUsed to test the difference between 2+ groups, usually 3+ or more
    Extension of the Mann-Whitney test
    Non-parametric test
  82. 82A Factorial ANOVA is another name for ......Two way between groups ANOVA
  83. 83Main EffectsImpact of each IV on the DV
  84. 84Interaction effectThe impact of the combination of both IVs on the DV
  85. 85Multiplicity and the problem it can causeWhen you conduct several sig. tests to test a single hypothesis
    Increases chances of type 1 Error!
  86. 86Scheffe TestPost Hoc Test: Comparing how +3 means differ (e.g. post hoc for ANOVA)
  87. 87Tukey TestPost Hoc Test: Comparing how +3 means differ (e.g. post hoc for ANOVA)
  88. 88Dunnett TestPost Hoc: Planned comparison normally, compares every mean to a control mean
  89. 89Paired t-testUsed when you want to check if the scores from the same participants have changed between two time points
  90. 90Wilcoxon signed-rank testNon-parametric equivalent of paired t-test
  91. 91Mauchlys test checks for..Spherecity: that the testing between each testing session as approx. equal variances
    If sig. = Assumption is violated, then interpret greenhouse-Giesser
  92. 92Friedman testNon-parametric equivalent of one way within groups ANOVA
  93. 93W. Kendalls coefficientcoefficient; estimate of effect size
  94. 94Bonferroni Test + Least Sig. Dif testPost Hoc tests for within-groups ANOVA
  95. 95Drawbacks of Bonferroni post hoc test?It multiplies each p number by number of corrections made, to control for multiplicity. Less likely to get a sig result!!!
    Less likely to get a type 1 error though
  96. 96Sidak testPost hoc test for within groups ANOVA
  97. 97Box's MTests homogeneity of variance co-variance: Only applied for Mixed ANOVA
    Assumption that the covariance between variables should ALSO be evenly distributed
  98. 98Extraneous VariableAll variables that are NOT the IV , but could explain the outcome of the experiment
  99. 99External validityThe strength of the sample studies generalisability to the population
  100. 100Internal validityThe extent to which a study establishes a cause and effect relationship
    i.e. that it CANT be explained by external factors
  101. 101Threats to External validity (6)- Ecological invalidity
    - sampling bias
    - Historical consequence
    - Experimenter effect
    - Hawthorne Effect
    - Situation effect
  102. 102Threats to Internal validity(4)- Historical consequence
    - Maturation
    - Instrumentation
    Attrition
  103. 103Experimenter EffectAn experimenter acting in a different way between treatment groups , causing an impact on their performance
  104. 104Hawthorne EffectWhen participant performance changes because they know they are being watched.
  105. 105Situation effectWhen the situation impacts the results
  106. 106AttritionWhen participants drop out of the study
  107. 107MaturationWhen changes over time impact the performance of participants in the study
  108. 108InstrumentationWhen different assessment tools are used for pre and post tests

Informações

Descrição
Psychology FlashCards sobre Sem título, criado por James Lafferty em 21-01-2021.
Criado por
James Lafferty
Visualizações
1
Avaliação

Fale Conosco

Sobre o que é sua mensagem?

O que você acha do novo visual?

Você gostou?