stud
bud
Sign up
Log in
Home
Library
Explore
SAT
More
PRM Regression Quiz Flashcards | StudBud
Explore
/
PRM Regression Quiz
Removing outliers from the dataset:
Multiple regression includes:
PRM Regression Quiz
46 terms
·
StudBud Administration
·
33 plays
Regression
Flashcards
46 cards
F
Practice quiz
P
Host live
L
Questions
46
Answers
1
Multiple regression includes:
Use of several simple regressions to produce a mean
The assessment of multiple Y's on X
Assessment of the relationship of multiple predictors on Y
Assessment of multiple predictors on multiple DVs
2
Removing outliers from the dataset:
Reduces strength of the association
Increases strength of the assocation
Reduce ability for significant finding due to smaller sample
Is inappropriate as you are altering the data set
3
Sample statistics differ from population parameters as:
Sample statistics cannot be used in statistical formulae
Sample statistics are represented by Greek letters
Population parameters refer to a subset from the sample
Sample statistics are represented by Roman letters
4
In multiple regression, the effect size of the prediction is given by the:
squared multiple correlation coefficient
sum of individual squared correlations between the IVs & DV
sum of squared beta coefficients
multiple correlation coefficient
5
Balancing nuisance factors neutralisesthe potential effect on internal validity, but can still:
Decrease both sensitivity and external validity
Increase both sensitivity and external validity
Decrease sensitivity and increase external validity
Increase sensitivity and decrease external validity
6
Outliers:
increase the strength of a correlation
Have no impact on the relationship between variables
Increase, reduce, or have no impact on the strength
Reduce the strength of a correlation
7
'Fanning' of the data points to one side of the scatterplot may indicate:
Homogeneity of variance
Heterogeneity of variance
Linearity
Non-normality
8
The advantage of reporting standardised statistics is:
Explains less variance than unstandardised form
Tells us if the variables are significantly related
Explains more variance than unstandardised form
A person unfamiliar with variable can interpret effect size
9
If the
R
2
change is not significant in a hierarchal regression, we can say that:
None of the variables were related to the DV
The N was too small
Group of predictors in this step did not improve the model
The predictors shared too much variance
10
Looking for evidence that 3 psychological techniques predict sport perf, we would hope that:
The DV is not too highly correlated with the IVs
All 3 IVs are uniquely related to the DV
The 3 predictors are highly correlated
The tolerance stat of several IVs is <.10
11
A scatterplot:
assesses if two variables are significantly correlated
used as a last resort to identify outliers in the data set
Uses a formula to calculate the p value
identifies heterogeneous subsamples and outliers in data set
12
An ad hoc inference is:
A scientific judgement
A 'gut' or intuitive inference based on past experience
As reliable as statistical inference
Based on scientifically captured data
13
In terms of the Pearson product-movement correlations (
r
), which is the strongest relationship?
-.70
.01
.69
-.10
14
Hierarchical Regression
is not useful
for:
Hypothesis Testing
Removing the effects of nuisance variables
Reducing predictors in a data set
Covariance Analysis
15
A multiple regression
may not
be statistically significant due to:
The p value being too small
The sample being too large
The R2 being too small
The R2 being too large
16
The assumptions that need to be satisfied for Pearson correlation are:
Heteroscedascity and independence
That the data comes from a valid and reliable sample
Bivariate normality and heteroscedascity
Bivariate normality and independence
17
Cohen's
d
suggests that:
.2 is a small effect
.8 is a small effect
.5 is a large effect
scores >1 are not possible
18
A Type 1 Error refers to:
Cases when the alternate hypothesis is correctly accepted
When the null hypothesis is correctly rejected
Making an error in calculating the statistic
When the null hypothesis is incorrectly rejected
19
When a regression line is fitted to the data, the
lack
of fit is best indicated by:
A random distribution of the residuals
A small unstandardised regression coefficient
A negative correlation coefficient
A non-random distribution of the residuals
20
Heterogeneity of subsamples is
best
assessed by:
Studying separate residual plots for each subgroup
Using HierRegr and assessing if slopes differ from parallel
Using HierRegr and assessing if intercepts differ
By assessing if slopes 'cross-over' each other
21
IQ: mean of 100 and SD of 15. IQ between 85 and 155 is normal; What % of population is normal?
70%
50%
90%
30%
22
You have been exposed to various forms of regression so far, including:
Reduction Regression
Logistic Regression
Poisson Regression
Stepwise Regression
23
Use of a control group, plus random assignment, would allow us to avoid:
Effects arising from testing
Selection Bias
Attrition
Experimenter Bias
24
Multicollinearity is:
A problem as highly correlated DVs violate assumptions
An assumption that needs to be satisfied
A problem as highly correlated IVs violate assumptions
A necessary element to proceed with multiple regression
25
Which one of the following does
not
increase the power of a test of
r
between X and Y?
Conducted a one-tailed t-test instead of a two-tailed t-test
Adopting a larger Type 1 error rate
Increasing the magnitude of r
Standardising the scores of X and Y
26
A correlation coefficient is:
Unable to determine the strength of the relationship
Bounded between -1 and 1
Used only to determine the direction of the relationship
determines A causes B
27
A scatterplot of residuals is NOT used to check for the:
Relationship between variables
Normality of distribution of errors
Presence of influential cases
Homogeneity of variance errors
28
It is
NOT
possible to use a measure as an independent variable if the format is:
LaTrobe University, Murdoch University, Melbourne University
Height in millimetres
Voting Yes/No
<20 years old, 20-29 years old, 30-39 years old
29
In correlational analysis, skewed variables are transformed to normally distributed because:
This might invalidate the test
This removes the effect of influential points
This decreases the relationship between variables
This increases the relationship between variables
30
A 'dummy variable' is:
A variable created to recode a polytomous predictor
A variable that cannot be used in regression
A polytomous predictor
A scalable polytomous variable
31
X is the independent variable and Y is the:
mediator
predictor variable
dependent variable
slope variable
32
In a MR of X1, X2 and X3 predicting Y, the tolerance value of X1 is the variance in X1 that is:
shared with X2 and X3 but not with Y
not shared with X2 and X3
shared with Y but not with X2 and X3
shared with X2 and X3
33
When performing ForwardRegression compared to standard MR, using the same number of predictors:
The same N would be required
A smaller N would be required for the Forward Regression
A larger N would be required for the Forward Regression
A smaller p criterion is set
34
Which case will not be considered an influential regression of Y on X?
A case with a small X and a mean of Y
A case with a large X and a small Y
A case with a mean of X and mean of Y
A case with a mean of X and a large Y
35
R
2
is derived from the:
Combination of unique &shared contributions of the Xs with Y
Combination of shared variances between all Ys with the X
Combination of unique contributions from each predictor
Combination of shared variances between all Xs with the Y
36
Simple linear regression:
Is best used for estimating curvilinear relationships
Involves measuring the association of multiple IV with DV
Does not indicate strength of relationship between variables
Gives an indication of strength and is used for linear
37
Residual analysis is conducted in regression to:
Get a more significant statistical result
Reduce the influence of misrepresentative cases
Get a higher r value
Reduce the sample size
38
The
p
value associated with an
r
value:
Tells us if the p value and r value are related
Tells if the direction of the relationship is stat significa
Tells us if the strength of the relationship is strong
Tells if the r value significantly differs from nullhypothes
39
You are more likely to attain a higher R
2
when:
Your IVs are highly correlated
Your IVs capture substantial unique components of the DV
Your DV is not correlated with the IVs
Your set of IV share substantial variance with each other
40
Three types of statistical regression are:
Hierarchical, Forward, Backward
Backward, Stepwise, Forward
Stairwise, Forward, Reverse
Reverse, Forward, Statistical
41
Which of the following is most likely a polytomous variable?
Age
Marital Status
Income
Gender
42
The defining characteristic of a true experiment is:
The systematic manipulation of independent variables
A high degree of control in testing
The operationalisation of dependent variables
Random assignment of subjects to conditions
43
In standard multiple regression where X1 and X2 are used to predict Y,
R
2
is:
Variance of Y uniquely by X1 plus variance Y uniquely by X2
proportion of variance uniquely&jointly predicted by X1 & X2
the sum of squared correlations between Y & X1, and Y & X2
the sum of the correlations between Y and X1 and Y and X2
44
If we found R
2
= .41, F(3,25),
p = .04,
we could state:
There were 25 independent variables
The finding was non-significant
There were 4 independent variables
There were 29 scores in data set
45
The external validity of an experiment can be improved by:
using a representative sample of the population
holding constant the effect of nuisance variables
balancing the effect of nuisance variables
random assignment of participants to treatment groups
46
The ρ symbol refers to:
The probability of the r value being significant
The population r*
The direction of the relationship between variables
The sample correlation value
