VOORBEREIDING ARMS
GRASPLE VRAGEN
What is the difference between amount of predictors / independent
simple & multiple linear variables (simple only 1)
regression?
6 assumptions of regression 1. normality of residuals
2. no outliers
3. no multicollinearity
4. homoscedasticity
5. level of measurement: ratio or
interval (or dummy)
6. linearity
In a regression equation, the 1. b0
intersection with the y-axis is also 2. intercept
called... 3. constant
What do we call the difference 1. residual
between y and ŷ? 2. error
How do we calculate this y-ŷ
difference?
If we square this difference, what is least squares method.
this method called? It's used to 'parameters', 'which model fits the
estimate ... of the linear regression data best'
model to see ....
R-squared is a so-called .... number Goodness of fit
meaning it explains the ... of the percentage of variance
response variable that is explained predictor variable
by the ...
What are the boundaries of R2? [0, 1]
What does R2 tell us about new Not much; it could be odd, specific
predictions? patterns only in the dataset
What is R called? multiple correlation coefficient
The assumption of linearity is oval cloud + no clear quadratic
usually met when we see an ... in trend
the point graph of the data
2 definitions of construct validity: 1. did the measurement
instruments succeed?
2. did the intended manipulation
succeed?
Definition of statistical validity: were the results of the analyses
accurate and well-founded?
5 questions to test construct 1. how is the concept defined?
validity: 2. how is the definition
operationalised?
3. does the chosen instrument
measure the intended concept?
4. reliability analysis
5. factor analysis
, Why does a large sample NOT sample must be representative
necessarily indicate high external
validity?
3 questions to test statistical 1. is analysis technique
validity appropriate for research question?
2. are the assumptions to perform
analysis met?
3. are the results of the analysis
reported and interpreted correctly?
2 problems of frequentist NHST 1. publication bias
2. Questionable Research Practices
What is the H0 of an ANOVA? ALL means are the same
What kind of test is an ANOVA? F-test
Another word for explained between variance, because
variance: explained by group membership
Another word for residual variance: within variance, because not
explained by group membership
What is the formula for F? MSbetween / MSwithin (mean
squares)
What does F < 1 mean? within MS (Variance) is bigger so p
is likely not small
What are the 6 assumptions of an 1. normality of DV in each group
ANOVA? 2. no outliers in scores dependent
variable
3. homogeneity of variances DV in
each group
4. scores on DV independent within
group
5. random sampling
6. measurement levels: DV =
ratio/interval, IV = categorical
What is the eta2? percentage of total variance that is
explained by group variable
How to read ANOVA post hoc tests? positive values mean differences:
first-mentioned group scores x
amount higher than other group
How is p bonferroni different? Why adjusted: multiplied by the amount
do we need it? of tests
we need it to adjust for error rate
increase
WEEK 1
Grasple vragen week 1
What is the Bayes theorem? P(A given B) = (P(B given A) . P
(given A)) / P(given B)
How is a prior model probability if each hypothesis is equally likely:
GRASPLE VRAGEN
What is the difference between amount of predictors / independent
simple & multiple linear variables (simple only 1)
regression?
6 assumptions of regression 1. normality of residuals
2. no outliers
3. no multicollinearity
4. homoscedasticity
5. level of measurement: ratio or
interval (or dummy)
6. linearity
In a regression equation, the 1. b0
intersection with the y-axis is also 2. intercept
called... 3. constant
What do we call the difference 1. residual
between y and ŷ? 2. error
How do we calculate this y-ŷ
difference?
If we square this difference, what is least squares method.
this method called? It's used to 'parameters', 'which model fits the
estimate ... of the linear regression data best'
model to see ....
R-squared is a so-called .... number Goodness of fit
meaning it explains the ... of the percentage of variance
response variable that is explained predictor variable
by the ...
What are the boundaries of R2? [0, 1]
What does R2 tell us about new Not much; it could be odd, specific
predictions? patterns only in the dataset
What is R called? multiple correlation coefficient
The assumption of linearity is oval cloud + no clear quadratic
usually met when we see an ... in trend
the point graph of the data
2 definitions of construct validity: 1. did the measurement
instruments succeed?
2. did the intended manipulation
succeed?
Definition of statistical validity: were the results of the analyses
accurate and well-founded?
5 questions to test construct 1. how is the concept defined?
validity: 2. how is the definition
operationalised?
3. does the chosen instrument
measure the intended concept?
4. reliability analysis
5. factor analysis
, Why does a large sample NOT sample must be representative
necessarily indicate high external
validity?
3 questions to test statistical 1. is analysis technique
validity appropriate for research question?
2. are the assumptions to perform
analysis met?
3. are the results of the analysis
reported and interpreted correctly?
2 problems of frequentist NHST 1. publication bias
2. Questionable Research Practices
What is the H0 of an ANOVA? ALL means are the same
What kind of test is an ANOVA? F-test
Another word for explained between variance, because
variance: explained by group membership
Another word for residual variance: within variance, because not
explained by group membership
What is the formula for F? MSbetween / MSwithin (mean
squares)
What does F < 1 mean? within MS (Variance) is bigger so p
is likely not small
What are the 6 assumptions of an 1. normality of DV in each group
ANOVA? 2. no outliers in scores dependent
variable
3. homogeneity of variances DV in
each group
4. scores on DV independent within
group
5. random sampling
6. measurement levels: DV =
ratio/interval, IV = categorical
What is the eta2? percentage of total variance that is
explained by group variable
How to read ANOVA post hoc tests? positive values mean differences:
first-mentioned group scores x
amount higher than other group
How is p bonferroni different? Why adjusted: multiplied by the amount
do we need it? of tests
we need it to adjust for error rate
increase
WEEK 1
Grasple vragen week 1
What is the Bayes theorem? P(A given B) = (P(B given A) . P
(given A)) / P(given B)
How is a prior model probability if each hypothesis is equally likely:
