Test
Correlational designs and regression
Este test tiene tipos de pregunta que se abren en el reproductor de siempre.
Preguntas del test (46)
1Correlational designs [blank_start]cannot[blank_end] claim causality, it [blank_start]can[blank_end] however make inferences about, for example, covariation
- Acannot
- Bcan
- Ccan
- Dcannot
2If a correlational design involves interval data you can use [blank_start]Pearson's product moment[blank_end]. If it involves non-parametric data you use [blank_start]Spearman's[blank_end].
- APearson's product moment
- BT-test
- CSpearman's
- DANOVA
3To find the average co-variation score you must calculate how much each score deviates from the [blank_start]mean[blank_end] on [blank_start]both[blank_end] axis and [blank_start]add[blank_end] these together and [blank_start]divide[blank_end] by n
- Amean
- Bmedian
- Cmode
- Dvariance
- Eboth
- Fone
- Gadd
- Hmultiply
- Idivide
- Jmultiply
4Co-variation is [blank_start]never[blank_end] larger than SD, higher SD indicates [blank_start]higher[blank_end] variation
- Anever
- Balways
- Chigher
- Dlower
5What does Pearson's r tell us? (2)
- AMagnitude of relationship
- BDirection of relationship
- CSignificance of relationship
- DLength of relationship
6An r score of 1 indicates what?
- APerfect, positive relationship
- BLarge, positive relationship
- CPerfect, negative relationship
- DNo relationship
7An r score of -0.8 indicates what?
- AVery large, negative relationship
- BPerfect, negative relationship
- CVery large, positive relationship
- DSmall, negative relationship
8An r score of 0.3 indicates what?
- AMedium, positive relationship
- BWeak, positive relationship
- CLarge, positive relationship
- DNo relationship
9An r score of 0 indicates what?
- ANo relationship
- BSmall, negative relationship
- CSmall, positive relationship
- DPerfect, positive relationship
10r measures..
- AEffect size
- BSample size
- CSignificance
- DSphericity
11To use Pearson's r you must meet the following assumptions: (4)
- ANormal distribution of variables
- BIndependent observations
- CInterval data
- DOrdinal data with 7 or more categories
- EManipulate one or more IV's
- FHave more than one DV
12To perform Pearson's r on SPSS:
- AAnalyse - correlate - bivariate
- BAnalyse - general linear model - univariate
- CAnalyse - regression - linear
- DGraphs - chart builder - scatter/dot
13How would you report Pearson's r?
- Ar = ____, N = ____, p < ____, ___-tailed
- BY = a + (b x X) + error
- CF = (___,___) ____, p ____, n2 = ____
- D(x1 - x2) / SD
14What graph would you use to illustrate the r results?
- AScattergram
- BBox plot
- CError bar chart
- DTable
15Data = model + error
- AVerdadero
- BFalso
16In the equation 'data = model + error', what is used as the model and estimate if only one variable is being investigated?
- AMean
- BRange
- CVariance
- DSD
17In the equation 'data = model + error', what is used as the error?
- ASD
- BMean
- CRange
- DMode
18A smaller SD indicates what?
- AMore accurate model
- BMore inaccurate model
- CEffect size
- DPower of test
19If you had a small SD, what would the points look like in comparison to the mean on a scattergram?
- AClosely clustered to mean line
- BDistant from mean line
- CRandomly dotted on scattergram
- DThere would be no points
20If there were 2 or more variables being investigated (i.e. a correlation), what would be used as the model?
- ARegression line
- BMean
- CSD
- DRange
21Annotate the below regression equation:
- AIntercept (start of line)
- BGradient (of line with 1 point increase)
22Using the equation 'Y = a + bX + error', what would someones score be if the intercept was 20, gradient 2, and X was 10?
- A40
- B202
- C50
- D0
23Using the equation 'Y = a + bX + error', what would someones score be if the intercept was 20, gradient 2, and X was 15?
- A50
- B40
- C302
- D600
24In a simple regression, what is used as the error?
- AANOVA
- BSD
- CMean
- DRegression line
25Why do we test the error of a simple regression? (2)
- ATo see if the regression line adds more info than the mean model
- BTo identify if the regression model is significant
- CTo identify the direction of the relationship
- DTo identify the magnitude of the relationship
26What is the ANOVA equation used when testing the error of a simple regression?
- AF = variance in DV explained by IV / variance in DV not explained by IV
- BF = mean square between / mean square within
- Ct = mean difference / SE difference
- Doutcome = predictor + error
27R squared indicates what?
- AHow much predictors account for outcome
- BNumber of predictors
- CDifference between 'number of participants - 1' and 'number of predictors'
- DAnalyse - regression - linear
28What is the difference between simple and multiple regression?
- AMultiple regression involves considering multiple predictors on one outcome
- BMultiple regression involves more than one outcome
- CMultiple regression is a more simple model
- DMultiple regression is always less accurate
29Multiple regression involves [blank_start]more[blank_end] predictors on [blank_start]one[blank_end] outcome/s, this makes it a [blank_start]more[blank_end] complex model, but maybe [blank_start]more[blank_end] accurate
- Amore
- Bless
- Cone
- Dtwo
- Emore
- Fless
- Gmore
- Hless
30Annotate the below multiple regression equation:
- Aintercept (start of line)
- Bgradient of line
- Cintercept of line for predictor 1
- Dgradient of line for predictor 1
- Egradient of line for predictor 2
- Fintercept of line for predictor 2
- Ggradient of line for predictor 3
- Hgradient of line for predictor 1
- Ierror
- Jintercept of line
31Each variable added in a multiple regression should:
- AReduce error
- BIncrease error
- CReduce significance
- DReduce effect size
32If we considered the relationship between hours studied AND self efficacy ON exam score, we would conduct:
- AMultiple regression
- BSimple regression
- CCorrelation
- DANOVA
33If we knew the constant (intercept) was 35.60, gradient for hours studied was 0.82, gradient for self-efficacy was 2.16, and we knew pupil 1 studied for 10 hours and had self-efficacy 6, what exam score could we predict pupil 1 would achieve?
- A56.76
- B62.12
- C52.15
- D50.80
34Using the previous example, which 1-point increase would be easier to identify? Self efficacy gradient 2.16 or hours studied gradient 0.82?
- ASelf efficacy
- BHours stuided
35Which part of SPSS output, when conducting a multiple regression, tells us the unit change in the predictor expressed as SD which allows all predictors to be in the same scale to identify largest predictor?
- ABeta values - standardised coefficients
- BBeta values - unstandardised coefficients
- CR
- DR squared
36What does the R score indicate?
- ACorrelation between observed values and outcome
- BMaximum proportion of variance accounted for by the model
- CMinimum proportion of variance accounted for by the model
- DWhich predictor has largest impact on outcome by expressing all predictors in terms of SD
37What does R squared indicate?
- AMaximum proportion of variance accounted for by the model
- BMinimum proportion of variance accounted for by the model
- CCorrelation between observed values and outcome
- DWhich predictor has largest impact on outcome by expressing all predictors in terms of SD
38What does R squared adjusted indicate?
- AWhich predictor has largest impact on outcome by expressing all predictors in terms of SD
- BMinimum proportion of variance accounted for by the model
- CMaximum proportion of variance accounted for by the model
- DCorrelation between observed values and outcome
39What is the ideal R squared score?
- A30%
- B50%
- C10%
- D45%
40Identify the Hierarchal method of regression:
- AExperimenter decides which order in which to enter variables into the model, entering known predictors first so they are accounted for
- BAll predictors entered simultaneously
- CPredictors selected using their semi-partial correlation with outcome, SPSS adds all variables and removes those adding little to the model
41'Predictors selected using their semi-partial correlation with outcome, SPSS adds all variables and removes those adding little to the model' describes which method of regression?
- AStepwise
- BHierarchal
- CForced entry
42Identify the Forced entry method of regression:
- AAll predictors entered simultaneously
- BExperimenter decides which order in which to enter variables into the model, entering known predictors first so they are accounted for
- CPredictors selected using their semi-partial correlation with outcome, SPSS adds all variables and removes those adding little to the model
43How would you run a regression on SPSS:
- AAnalyse - regression - linear
- BAnalyse - compare means - t-test
- CAnalyse - general linear model - univariate
- DAnalyse - plots
44On the SPSS output for a regression, what table indicates which predictors significantly predict the outcome?
- ACoefficients table
- BANOVA table
45The ANOVA table produced once running a regression indicates:
- AWhether the regression model adds more info than mean model
- BWhich predictors significantly predict the outcome
46Assumptions of regression
Outcome variable [blank_start]must be[blank_end] continuous
Predictors [blank_start]must not have[blank_end] zero variance
Must be a [blank_start]linear[blank_end] relationship
Outcome values must be from [blank_start]independent, different[blank_end] participants
Predictors cannot be highly correlated ([blank_start]multicollinearity[blank_end])
Error must be normally distributed- Amust be
- Bmust not be
- Cmust not have
- Dmust have
- Elinear
- Fnon-linear
- Gindependent, different
- Hdependent, similar
- Imulticollinearity
- Jsphericity
Información
- Descripción
- RDA2 - Quantitative Test sobre Correlational designs and regression, creado por Róisín Farmer el 04/01/2017.
- Autor
- Róisín Farmer
- Idioma
- English
- Visitas
- 4
- Historial
- Creado el 4 ene 2017 · actualizado el 10 ene 2017
- Etiquetas
- Valoración
- Aún sin valoraciones Valora este test