Quiz
Regression
This quiz has question types that open in the usual player.
Questions in this quiz (20)
1Which of these could be an equation for the line of best fit / least squares regression?
- Ay = mx + c
- By = A0 + A1x
2In the equation y = mx + c, the mx is the [blank_start]gradient[blank_end], and the c is the [blank_start]intercept[blank_end].
- Agradient
- Bintercept
- Cintercept
- Dgradient
3The non-parametric measure of correlation, whereby data is converted into ranks and the correlation between these ranks is assessed, is called:
- ASpearman's rank correlation
- BPearson's correlation coefficient
4The range for Pearson's correlation coefficient is:
- Ap = -1 to 1
- Bp = 0 to 1
- Cp = 0 to 10
- Dp = 0 to 100
5In Pearson's correlation coefficient:
- Agradient doesn't matter, only correlation does
- Bcorrelation doesn't matter, only gradient does
- Cgradient doesn't matter, only intercept does
- Dintercept doesn't matter, only gradient does
6What are ANOVA, ANCOVA and linear regression examples of?
- AEmbedded models of feature selection
- BGeneral linear models
- CHypothesis testing
- DCorrelation coefficients
7What do general linear models assume about error?
- Athat error is normally distributed
- Bthat error is not normally distributed
- Cthat there is no error
- Dthat error is significantly different for every data point
8What is the test statistic for general linear models?
- AF ratio
- BP value
- Cy value
9How is the F ratio calculated?
- Atreatment mean square / error mean square
- Berror mean square / treatment mean square
- C1 - treatment mean square
- D1 - error mean square
10The treatment mean square describes the explained variation
- ATrue
- BFalse
11The error mean square describes the unexplained variation which is the 'signal'
- ATrue
- BFalse
12What is meant by the residual sum of squares?
- Asum of all the squared deviations between the data points and the fitted line
- Bsum of all the squared deviations between the data points and the group mean
- Csum of all the squared deviations between the data points and the total mean
13What does the R2 mean?
- AHow much variation can be explained by the model
- BHow much variation cannot be explained by the model
14What is meant by Simpson's Paradox?
- Awhen true correlations are hidden
- Bwhen false correlations are given
15The original data, what has actually been measured, is called:
- Ainterpolation
- Bextrapolation
16The predicted data from outside the original range is known as:
- Aextrapolation
- Binterpolation
17If the relationship is not a straight line, which of the following could work?
- Aappropriate polynomial explanatory variables (e.g. quadratic, cubic)
- Bnon-linear regression (e.g. Random Forest Regression)
- Cnone of the above
18What can be done if a relationship is not well described by polynomial explanatory variables?
- Anon-linear regression
- Blinear regression
- CANOVA
- DANCOVA
- Efeature selection
19Random Forest Regression is an example of non-linear regression.
- ATrue
- BFalse
20Which of the following are disadvantages of Random Forest Regression?
- Acan be slow to make new predictions
- Bcan oversimplify the data
- Ccan overfit the noise
- Dis very conceptually hard
Information
- Description
- MIBTP Statistics Quiz on Regression, created by Bethany Richmond on 04/11/2018.
- Author
- Bethany Richmond
- Language
- English
- Views
- 1
- Tags
- Rating
- No ratings yet Rate this quiz