QCM PRACTICE TEST QUESTIONS
WITH 100% CORRECT ANSWERS
What is the non-parametric equivalent of one-way ANOVA? - Answer- Krystal-Wallace
test
What is the non-parametric equivalent of one-way ANCOVA - Answer- none
What is the non-parametric equivalent of two-way ANOVA? - Answer- none
What is the non-parametric equivalent of paired t-test - Answer- Wilcoxon matched-
pairs signed-rank test
How do we test normality of residuals? - Answer- histogram of residuals + QQ plot
What is another term for homogeneity of variance? - Answer- homoscedasticity
When P=0.05, what chance is there that all the means are equal? - Answer- 5%
If the overall effect of a categorical predictor is significant, what do we do next? -
Answer- use post-hoc tests to find out which groups differ from one another
What is the purpose of correction methods in post-hoc tests? - Answer- corrects for
multiple testing
Homogeneity of variance - Answer- the error should have the same variance throughout
the range of x values
How do we test homogeneity of variance? - Answer- plot the residuals against the fitted
values
How to test whether x influences y (2) - Answer- test whether the gradient is equal to 0
Test whether the model is better than a null model that does not include X
What is the non-parametric equivalent of the one-way repeated-measures ANOVA? -
Answer- Friedman test
What are parametric tests based on? - Answer- t distribution
, Assumption of parametric tests - Answer- population distribution from which the
samples came is normal
What are four steps of statistical analysis? - Answer- plot data
fit model
check assumption
test hypotheses
when to use correlation? - Answer- when there is no reason to believe that one variable
is the cause of the other
(either could go on x axis)
when to use regression? - Answer- when one variable is the cause of the other
purposes of regression (3) - Answer- quantified relationship between variables
predicting new y values based on x
testing significance of relationship between predictor and response
which variable is the predictor? - Answer- independent variable
which variable is the response? - Answer- dependent variable
what assumption do we make when we use regression? - Answer- there is a linear
relationship between the two variables
why do we assume that there is a linear relationship between variables in regression? -
Answer- allows us to predict x values using y=mx+c
error on a scatter graph - Answer- vertical distance between the furthest data points
above and below the fitted line at each x value
residuals on a scatter graph - Answer- vertical distance from each point to the fitted line
what is the line of best fit? - Answer- line that minimises the sum of squared residuals
what is the R command for regression - Answer- lm(y~x)
what does (y~x) mean? - Answer- response as a function of predictor
what equation is used to test whether the slope is equal to zero in regression - Answer-
y = mx + c + error
what are the four assumptions of GLM? - Answer- independence of data points
linear relationship
WITH 100% CORRECT ANSWERS
What is the non-parametric equivalent of one-way ANOVA? - Answer- Krystal-Wallace
test
What is the non-parametric equivalent of one-way ANCOVA - Answer- none
What is the non-parametric equivalent of two-way ANOVA? - Answer- none
What is the non-parametric equivalent of paired t-test - Answer- Wilcoxon matched-
pairs signed-rank test
How do we test normality of residuals? - Answer- histogram of residuals + QQ plot
What is another term for homogeneity of variance? - Answer- homoscedasticity
When P=0.05, what chance is there that all the means are equal? - Answer- 5%
If the overall effect of a categorical predictor is significant, what do we do next? -
Answer- use post-hoc tests to find out which groups differ from one another
What is the purpose of correction methods in post-hoc tests? - Answer- corrects for
multiple testing
Homogeneity of variance - Answer- the error should have the same variance throughout
the range of x values
How do we test homogeneity of variance? - Answer- plot the residuals against the fitted
values
How to test whether x influences y (2) - Answer- test whether the gradient is equal to 0
Test whether the model is better than a null model that does not include X
What is the non-parametric equivalent of the one-way repeated-measures ANOVA? -
Answer- Friedman test
What are parametric tests based on? - Answer- t distribution
, Assumption of parametric tests - Answer- population distribution from which the
samples came is normal
What are four steps of statistical analysis? - Answer- plot data
fit model
check assumption
test hypotheses
when to use correlation? - Answer- when there is no reason to believe that one variable
is the cause of the other
(either could go on x axis)
when to use regression? - Answer- when one variable is the cause of the other
purposes of regression (3) - Answer- quantified relationship between variables
predicting new y values based on x
testing significance of relationship between predictor and response
which variable is the predictor? - Answer- independent variable
which variable is the response? - Answer- dependent variable
what assumption do we make when we use regression? - Answer- there is a linear
relationship between the two variables
why do we assume that there is a linear relationship between variables in regression? -
Answer- allows us to predict x values using y=mx+c
error on a scatter graph - Answer- vertical distance between the furthest data points
above and below the fitted line at each x value
residuals on a scatter graph - Answer- vertical distance from each point to the fitted line
what is the line of best fit? - Answer- line that minimises the sum of squared residuals
what is the R command for regression - Answer- lm(y~x)
what does (y~x) mean? - Answer- response as a function of predictor
what equation is used to test whether the slope is equal to zero in regression - Answer-
y = mx + c + error
what are the four assumptions of GLM? - Answer- independence of data points
linear relationship
