Quiz
PSY4210 exam
Create a free account to finish this quiz
Save your progress and see your final score.
Already have an account? Sign in
Questions in this quiz (85)
1Which of the following is a possible return for a Boolean value in R?
- ATRUE
- BNA
- CMISSING
- D56
2I am conducting a study into the number of children people have in my hometown. What type of data would I expect to be collecting?
- ANumeric
- BInteger
- CComplex
- DLogical
3Which type of merge would you expect to result in the most amount of missing data (assuming not much overlap between your two datasets)?
- ANatural join
- BFull outer join
- CLeft outer join
- DRight outer join
4Which type of merge would you expect to result in the least amount of rows (assuming not much overlap between your two datasets)?
- ANatural join
- BFull outer join
- CLeft outer join
- DRight outer join
5I am conducting a study and using a Bernoulli distribution for binary regression. In my sample, I find that 75% of people have schizophrenia, and 35% of people have anxiety. Upon further inspection, I realise that all cases of anxiety occur in people with schizophrenia. What is this an example of?
- AThe logit function
- BSeparation
- COR = 1
- DOR = 0
6Which of the following would you not expect to see in the output for a poison regression?
- Ap-value
- Bz-value
- Ct-value
- Db value
7I am going through and cleaning my data and notice that my current data set has time as a variable, so each participant ID has multiple rows associated with it for each time point (e.g., participant A has timepoint 1, 2, and 3 as different rows). What can you say about the format of my data?
- ALong
- BWide
- CSuited to binary regression
- DSuited to poisson regression
8I have a large data set with 10,000 entries from people with varying demographic characteristics. I have just thought of a research question that I could conduct with this data, however I would only be able to use the data from participants under the age of 10. What should I do?
- ACreate a subset of the data including only participants under age 10
- BCreate a polynomial of the data
- CRemove all participants over the age of 10 from the data set
- DCollect new data, as you cannot simply re-use data for the sake of a new research question
9Sandra and Becky were instructed to create a univariate graph of the distribution of stress. Sandra decided to use filled bars for her histogram with a grid background, whereas Becky created a histogram with clear bars and background. Which graph seems to have the lowest data to ink ratio, and which graph would (based on data:ink ratio alone) be preferred by their instructor?
- ASandra; Becky
- BBecky; Becky
- CSandra; Sandra
- DBecky; Sandra
10Which of the following values indicates the highest level of internal consistency?
- Aα = 0.56
- Bα = 0.79
- Cκ = 0.88
- Dκ = 0.31
11I want to make a comparison of the distributions of stress for 65 people who attend university, and compare it with the distribution of stress for 70 people who work full-time. Which univariate plot is best suited to this?
- AViolin plot
- BBar plot
- CDot plot
- DScatterplot
12I am scoring a simple instrument whereby a total score is obtained by summing the scores for each question. Assuming that I have missing data, which function would be most appropriate to score my instrument?
- ASum all items together (e.g., LSS1 + LSS2 +LSS3)
- Buse rowMeans() (e.g., rowMeans(LSS1, LSS2, LSS3))
- Cuse rowMeans() and multiply by the total number of items (e.g., rowMeans(LSS1, LSS2, LSS3)*3)
- DSum all items together and multiply by the total number of items (e.g., (LSS1 + LSS2 + LSS3)*3)
13In linear regression, we want to _____ the distance between the predicted data and the observed data.
- Aminimise
- Bmaximise
- Cextend
- Deliminate
14Which type of model best suits the following equation, assuming all variables are continuous?
stress = 3.7*loneliness + 3.9*happiness + error- AMultiple linear regression
- BSimple linear regression
- CPoisson regression
- DModerated linear regression
15I run some diagnostics on my general linear model and receive the following diagnostic plots. What does it mean in terms of continuing with my analysis?
- AI can continue with the analysis without further action
- BI should remove all outliers before continuing with the analysis
- CI should not continue with the analysis as the assumption of homogeneity of variance is violated and I have multiple outliers
- DI should transform my data as the assumption of normality is violated, and remove all outliers before continuing with the analysis
16What is the difference between a linear model and the generalised linear model?
- AA linear model extends to different outcomes such as continuous, binary, and count variables
- BA linear model extends to different predictors such as continuous, binary, and count variables
- CA generalised linear model extends to different outcomes such as continuous, binary, and count variables
- DA generalised linear model extends to different predictors such as continuous, binary, and count variables
17How many parameters does the normal distribution have?
- A0
- B1
- C2
- D3
18I am trying to explain to my friend how we interpret the notation for simple linear regression. They are very confused about what b0 represents. What is the best explanation?
- AIt is the slope of our predictor variables, where high positive values indicate a steep positive slope, and high negative values indicate a steep negative slope
- BIt is the intercept of the model, or the value of the outcome when all predictor/s are equal to 0
- CIt is the intercept of our model, which tells us about the difference between the expected and predicted values generated by our model
- DIt is the slope of our predictor variables, where low positive values indicate a steep positive slope, and low negative values indicate a steep negative slope
19What sort of transformation can we expect to see when we apply the link function for linear regression (the identity function) to a normal linear regression?
- ANo transformation is expected
- BThe left side of the graph will become unbounded and fall between positive and negative infinity
- CThe right side of the graph will become unbounded and fall between positive and negative infinity
- DAll values will be transformed to fall between 0 and 1, like probabilities
20What is a plausible interpretation of the following output?
- AThere is a statistically significant association between points and salary, whereby one unit higher points is associated with a $10232.50 increase in salary (p < .001)
- BThere is a statistically significant association between salary and points, whereby a one dollar higher salary is associated with a 10232 increase in points (p<.001)
- CThere is a statistically significant association between points and salary, whereby a one unit higher points is associated with a $10232.50 decrease in salary (p < .001)
- DThere is a statistically significant association between salary and points, whereby a one dollar higher salary is associated with a 10232 decrease in points (p<.001)
21In which case would it be appropriate to do a log transformation of the outcome variable in normal linear regression?
- AWhen normality is violated
- BWhen homogeneity of variances is violated
- CWhen there are many outliers present
- DWhen you are wanting to calculate odds ratios
22What is the correct interpretation of the attached output?
- AWhen controlling for years of experience, the association between satisfaction score and salary was statistically significant (p = .057), such that a one unit higher satisfaction score was associated with a -1.19 Lakh decrease in salary.
- BWhen controlling for satisfaction score, the association between years of experience and salary was statistically significant (p = .022), such that one year higher experience was associated with a 2.08 Lakh increase in salary.
- CWhen satisfaction score = 0, salary is expected to be 12.29 Lakh
- DWhen satisfaction score and years of experience = 0, salary is expected to be 12.29 Lakh
23Which effect size should I use in a linear regression if I want to get an estimate of the unique variance in the outcome explained by each predictor, relative to the total variance not explained
- ACohen's f2
- BConditional R2
- CMarginal R2
- DR2
24What is an appropriate interpretation of the attached output?
- AThere was a significant simple effect of self-esteem on stress, such that a one unit increase in self-esteem was associated with a -2.17 decrease in stress (p <.001). This relationship was dependent on the significant interaction term, loneliness (p <.001)
- BThere was a significant simple effect of loneliness on stress, such that a one unit increase in self-esteem was associated with a 0.05 increase in stress (p <.001). This relationship was dependent on the significant interaction term, self-esteem (p <.001)
- CThere was a significant simple effect of self-esteem on stress, such that a one unit increase in self-esteem was associated with a -2.17 decrease in stress (p <.001).
- DThere was a significant simple effect of loneliness on stress, such that a one unit increase in self-esteem was associated with a 0.05 increase in stress (p <.001).
25What is the best conclusion for the outcome variable based on this density plot?
- AIt is a skewed variable and likely needs a log transformation before we can carry on with analysis
- BIt is a count variable, so we cannot use a linear model
- CIt is a binary variable, so we cannot use a linear model
- DIt appears to be normally distributed
26Lucy is telling me about her data analysis for her honours project. I have a look at her R code and am really confused with what she is doing, so to clear things up she tells me that her distribution only has a single parameter, λ (lambda). What can I now conclude about her data?
- AShe is using a Poisson distribution, and has a count variable as her outcome
- BShe is using a Poisson distribution, and has a binary variable as her outcome
- CShe is using a Bernoulli distribution, and has a count variable as her outcome
- DShe is using a Bernoulli distribution, and has a binary variable as her outcome
27How could we best define Eta for a Poisson regression after being transformed by the link function?
- ANever goes below zero
- BFalls only between 0 and 1
- CFalls between positive and negative infinity
- DNever goes above zero
28What assumption is NOT relevant to Poisson distribution?
- AOutliers
- BHomogeneity of variance
- CIndependence of errors
- DLinearity
29What can be concluded from the attached output?
- AA one-unit increase in stress predicted -0.54 times the odds of having high self esteem (p<.001)
- BA one-unit increase in anxiety predicted 0.00 times the odds of having high self-esteem (p<.001)
- CA one-unit increase in stress predicted 0.59 times the odds of having high self-esteem
- DA one-unit increase in anxiety was associated with a 1 unit increase in self-esteem
30In binary logistic regression, what shaped graph would we expect to return if we graphed probabilities against unbounded probabilities
- AA line
- BAn S shape
- CAn inverse S shape
- DA positive slope
31In logistic regression, an OR of 2.7 would indicate
- Athat the coefficient increases the odds
- Bthat the coefficient decreases the odds
- Cthat the coefficient has no effect
- Dthat people are 27x more likely to achieve the outcome
32In a logistic regression predicting stress from self-esteem, a predicted probability of 0.4 can be classified as
- Aa model predicting low levels of stress
- Ba model predicting high levels of stress
- Ca model predicting 40% of stress
- Da model predicting 60% of stress
33What is the correct interpretation of the following output, assuming the outcome variable is loneliness?
- AA one unit increase in self-esteem predicts a -0.35 decrease in the probability of being high on loneliness
- BA one unit increase in self-esteem predicts a -0.35 decrease in loneliness
- CA one unit increase in loneliness predicts a -0.35 decrease in the probability of being high on self-esteem
- DA one unit increase in loneliness predicts a -0.35 decrease in loneliness
34I have decided to conduct a survey on ice cream preferences in my city. I send out my survey via mail, and my participants mail be back their answers. However, 5% of the surveys were lost in the mail. What type of missing data is this?
- Amissing completely at random
- Bmissing at random
- Cmissing not at random
- Dmissing due to error
35My friend tells me that her honours project is on children with ADHD. She is getting them to fill out some questionnaires on their ADHD symptoms, but she notices that some of the kids didn't fill out all of the sections, so she has a bit of missing data. She comes to me for help as she doesn't know the best way to deal with it. What type of missing data is likely occurring here and what is the best way to deal with it?
- AMCAR; listwise deletion of missing responses
- BMNAR; listwise deletion of missing responses
- CMNAR; multiple imputation
- DMCAR; multiple imputation
36Which component of multiple imputation language refers to an estimate of a population value (e.g., mean or regression coefficient) with some uncertainty due to sampling variation, and is calculated in each imputed data set?
- AQ
- BQ-
- CQ^
- DQ$
37I am currently conducting multiple imputations on my data, and I have just generated 30 data sets with no missingness by filling in different plausible values for any missing data. What should I do next?
- APool the results to generate an overall estimate, Q-
- BPerform the analysis of interest on each imputed data set and generate 30 different Q^ and associated standard error
- CExamine my incomplete raw data to see distributions of missingness
- DNothing, this is the final step of the multiple imputation process
38In the attached image, what does B represent?
- Athe variance in the Q^ estimates between imputed data sets
- Bthe total variance
- Cthe variance due to sampling uncertainty
- Dthe variance estimate (e.g. squared standard error)
39When someone says their data has "converged" in the context of multiple imputation, what does this mean?
- AThere is no clear systematic change with additional iterations
- BThe model is able to operate on the most optimal parameters
- CThere is a high amount of variation between iterations
- DThere is no longer any missing data
40Which of the following is NOT a drawback of using GLMs for multiple imputations?
- AThey assume linear relationships on the scale of the link function
- BThey only include interactions between variables when specified by the analyst
- CThey are more prone to errors due to sampling variation
- DIn small data sets, there may be more variables than participants
41What can be concluded from the attached missing data plot?
- AThe most common combination present in the data set is data missing on Ozone and Solar-R, and data present on Wind, Temp, Month, and Day
- BThe proportion of participants who have missing data only on Ozone is 2.29%
- COzone has the lowest amount of missing data
- DSolar-R has around 5% missing data
42I am completing a multiple imputations for a data set with 68 missing variables, and 5635 participants. I am using a GLM and my outcome variable is stress. Which seed number should I use?
- A5635
- B6868
- C1234
- DIt doesn't matter
43According to the attached output, what can we conclude?
- AThe proportion of total variance due to missing data for positive affect is 30%
- BThe between imputation variance for age is 9.4
- CV- for positive affect is 0.00024
- DThe regression coefficient for age is -0.235
44I am studying whether stress increases near exam time. I have decided to employ a longitudinal design to do so, and have asked participants to come in 6 times during their uni semester to fill out a survey on stress levels. I am hypothesizing that the closer the stress surveys are filled out to exam time, the higher the amount of stress will be. To analyse my data, I employ a simple linear regression, with exam proximity as a predictor and stress as an outcome. I will use the B values, confidence intervals, and p values to make conclusions about my data. What is the main issue with my approach?
- AUsed a linear model when I should have used a linear mixed model
- BI have failed to control for covariates
such as assignments that may influence stress - CI should not be using B values, but beta values for my interpretation to ensure standardisation
- DStress should be my predictor, not my outcome variable
45Which of the following describes a scenario in which repeated measures ANOVA is not appropriate?
- AWhere time is a continuous variable
- BIf you have a lot of missing data
- CWhere the predictors are continuous variables
- DAll of these options are correct
46A regression coefficient that differs by person is called a
- Afixed effect/variable
- Brandom effect/variable
- Crepeated effect/variable
- Dbetween effect/variable
47I am conducting a study using linear mixed models. When I run my analysis, I get an intraclass correlation coefficient of 0.88. What does this mean?
- A88% of the total variance occurs between people
- B88% of the total variance occurs within people
- C12% of the total variance is due to error
- D88% of the total variance occurs either within or between people
48I am participating in a study where I need to record my appetite over the course of a week out of 10. The researchers tell me that at the end of the study my average appetite rating was 7/10, and on day 4 my score was -5. What is my appetite rating for day 4?
- A2
- B5
- C7
- D9
49Relating to the last question, the deviations I experienced in appetite rating from my own mean over the course of the week is termed my
- Awithin effects
- Bbetween effects
- Cintervariability
- Dintravariability
50What type of model was fit here, and how many people were included in the study?
- AAn intercept-only LMM of loneliness; 87
- BA random slope intercept-only LMM of loneliness; 87
- CA random slope intercept-only LMM of loneliness; 259
- DAn intercept-only LMM of loneliness; 259
51What is the correct interpretation of the attached output? (ignore the fact that there are no confidence intervals)
- AThe fixed effect intercept revealed that the average loneliness was 2.18 (p<.001). However, there were individual differences in mean loneliness, with the standard deviation for the random intercept being 1.02.
- BThe random effect intercept revealed that the average loneliness was 1.02. However, there were individual differences in mean loneliness, with the fixed effect for the random intercept being 2.18 (p<.001).
- CThe fixed effect intercept revealed that the average loneliness was 2.18 (p<.001). However, there were individual differences in mean loneliness, with the standard deviation for the random intercept being 1.07.
- DThe random effect intercept revealed that the average loneliness was 1.07. However, there were individual differences in mean loneliness, with the fixed effect for the random intercept being 2.18 (p<.001).
52Which level of data do we remove outliers for first in a linear mixed model?
- AWithin person
- BBetween person
- COutcome
- DPredictor
53I am running a study and creating some BLUPs. The mean for my data is 50, and I am aware that I will likely experience some shrinkage for my data. Participant 18 in my data set only filled in 6/10 of the surveys, and their mean was 25, whereas participant 27 filled in all of their surveys, with a mean of 25 as well. Should I expect a difference in shrinkage, and if so, which participant would experience more shrinkage?
- ANo
- BYes; participant 18
- CYes; participant 27
- DMore information is needed to answer this question
54How do we overcome the issue of participants with more observations having weighted means when using egltable() for descriptives?
- ADrop duplicated rows to keep one row of data per ID and remake the table
- BConvert the data to wide format
- CConvert the data to long format
- DAverage all participant means to get the average marginal effects and then remake the table
55Which of the following is NOT an approach to calculating descriptives on multilevel, continuous variables
- Aaverage by person and report descriptives on individual averages
- Breport descriptives on the overall variable to capture the total variance
- Creport descriptives on individual timepoint/assessments in longitudinal data
- Dall of these options are correct
56I am conducting an analysis using an intercept-only linear mixed model. I begin by ensuring that my ICC indicates that I have variance both within and between participants. I then set my estimator to maximum likelihood (ML). Next, I calculate my descriptives, and assess for the presence of outliers, homogeneity of variance, and distributional assumptions. Once that is done, I generate a summary of my model and continue to interpret it. What mistake have I made?
- AI forgot to check for the assumption of a linear association between predictor and outcome
- BI used ML instead of REML
- CI did not need to calculate an ICC for this model as it is intercept-only
- DI forgot to check for independence of assumptions
57What is the best interpretation of the attached output?
- ABoth between and within-person loneliness were statistically significantly associated with stress. On average across people, at the between person level, a one unit higher average loneliness score was associated with a 0.26 increase in average stress scores (p<.05). On average across people, at the within-person level, a one unit higher than usual daily loneliness score was associated with a 0.16 increase in average stress (p<.05)
- BBoth between and within-person loneliness were statistically significantly associated with stress. On average across people, at the between person level, a one unit higher average loneliness score was associated with a 0.16 increase in average stress scores (p<.05). On average across people, at the within-person level, a one unit higher than usual daily loneliness score was associated with a 0.26 increase in average stress (p<.05)
- COnly between person loneliness was statistically significantly associated with average stress. On average across people, at the between person level, a one unit higher average loneliness score was associated with a 0.26 increase in average stress scores (p<.05). Although not significant, on average across people, at the within-person level, a one unit higher than usual daily loneliness score was associated with a 0.16 increase in average stress (p = 0.326)
- DOnly between person loneliness was statistically significantly associated with average stress. On average across people, at the between person level, a one unit higher average loneliness score was associated with a 0.16 increase in average stress scores (p = 0.326). Although not significant, on average across people, at the within-person level, a one unit higher than usual daily loneliness score was associated with a 0.26 increase in average stress (p <.05)
58What is the most robust method of calculating degrees of freedom for LMMs?
- ABootstrapping
- BUsing profile-based confidence intervals
- CAssuming sample size is large enough so we can assume parameters follow a normal distribution
- DAssuming sample size is large enough so we can assume parameters follow a t-distribution
59What is standard deviation replaced with in linear mixed models?
- AThe correlation of random effects with fixed effects
- BThe correlation coefficient
- CCovariance
- DMultivariate standard deviation
60What is the likely value for the covariance between x and y based on this graph?
- A0.8
- B0
- C-0.4
- D-0.7
61Which of the following statements are true?
- AMultivariate normal distribution implies univariate normal distribution
- BUnivariate normal distribution implies multivariate normal distribution
- CMultivariate normal distribution does not imply univariate normal distribution
- DNone of these are correct
62What does this plot signify?
- AThe data follow a MVN
- BThe data do not follow a MVN
- CThe data are skewed to the left
- DThe data are skewed to the right
63Fill in the blanks
- AA = within-person B = between person
- BA = between person, B = within-person
- CA = intravariance, B = intervariance
- DA = intervariance, B = intravariance
64You have run a model with ADHD symptom severity as a predictor and loneliness as an outcome. If you have a correlation between the random intercept and slope of -0.85, what does this mean?
- APeople who have a higher level of loneliness when ADHD symptom severity is 0 also tend to have a more negative slope of the association between ADHD symptom severity and loneliness
- BPeople who have a higher ADHD symptom severity when loneliness is 0 also tend to have a more negative slope of the association between ADHD symptom severity and loneliness
- CPeople who have a higher ADHD symptom severity when loneliness is 0 tend to have a more positive slope of the association between ADHD symptom severity and loneliness
- DPeople who have a higher level of loneliness when ADHD symptom severity is 0 tend to have a more positive slope of the association between ADHD symptom severity and loneliness
65What is a correct interpretation of the attached output?
- AThere was a significant effect of within-person loneliness on average stress, such that on days where people were one unit higher on loneliness than their average, they were expected to have 0.15 higher stress. There were individual differences, with the standard deviation for the random slope being 0.09, indicating there are individual differences in the slope of within person loneliness on stress.
- BThere was a significant effect of between-person loneliness on average stress, such that where people had one unit higher average loneliness, they were expected to have 0.15 higher stress. There were individual differences, with the standard deviation for the random slope being 0.09, indicating there are individual differences in the slope of between person loneliness on stress.
- CThere was a significant effect of within-person loneliness on average stress, such that on days where people were one unit higher on loneliness than their average, they were expected to have 0.26 higher stress. There were individual differences, with the standard deviation for the random slope being 0.86, indicating there are individual differences in the slope of within person loneliness on stress.
- DThere was a significant effect of between-person loneliness on average stress, such that where people had one unit higher average loneliness, they were expected to have 0.26 higher stress. There were individual differences, with the standard deviation for the random slope being 0.86, indicating there are individual differences in the slope of between person loneliness on stress.
66Convergence and _______ issues often surround the _______ of the model
- Asingularity; fixed effects
- Bsingularity; random effects
- Cdivergence; fixed effects
- Ddivergence; random effects
67What is the term given to describe where the variances of one or more linear combinations of effects in LMM are/close to zero?
- AConvergence
- BSingularity
- CMarginal effects
- DConditional effects
68I currently have a data set that was analysed with LMM to determine whether stress can predict happiness. My supervisor mentions to me that it would be interesting to see what the average stress score would be if happiness was 5 points above the average on a given day. What type of prediction would I be making to get this information?
- AMarginal prediction
- BConditional prediction
- COvert prediction
- DCovert prediction
69In LMM notation, regression coefficients without any j subscript indicate
- Afixed effects only
- Bfixed and random effects
- Crandom intercepts
- Drandom slopes
70Which of the following variables/variable combinations cannot be random effects?
- AWithin only variable
- BBetween only variable
- CWithin and between variables
- DContinuous variables
71Which of the following is a plausible example of something that could be represented by the 'x2' in this equation?
- ALoneliness scores measured daily for each of the 50 participants in the study
- BLoneliness scores measured once at baseline for each of the 50 participants in the study
- CA random slope
- DA fixed-effect only slope
72Which of the following can be used as an outcome in a LMM?
- ABaseline stress scores measured for each participant in the study
- BSex of the participants (dummy coded)
- CDaily happiness scores measured for each participant in the study
- DAge of the participants, measured once at the end of the study
73I am conducting some research on whether stress can predict emotional eating, and whether this is moderated by coping skills. I measure coping skills at the beginning of the study, and measure stress and emotional eating every day that the study runs. What type of interaction is this?
- ACross level
- BBetween unit
- CWithin unit
- DMixed effect
74When would using the 25th and 75th percentiles for simple slope creation be especially useful?
- AWhen the data is skewed
- BWhen the data is normally distributed
- CWhen you have over 100 participants
- DWhen you have under 100 participants
75What can be said about the predictor variables involved in the model associated with the attached graph?
- AThey are both continuous
- BThey are both categorical
- COne is categorical and one is continuous
- DYou cannot infer anything about the variable types based on this graph
76Model A predicts happiness from stress and friendship number, and model B predicts happiness only from friendship number. Therefore, these models are
- ANested
- BNot nested
- CCross level
- DBetween unit
77Model A predicts stress from loneliness and anger, and model B predicts stress from loneliness and happiness. What can we use to compare these models?
- ALikelihood ratio test
- BAIC
- CBIC
- DAIC & BIC
78Which of the following is a simple description for the likelihood ratio test?
- AIt is 2x the difference in log-likelihoods of the models being compared
- BIt is the chi-squared distribution of the likelihood of the models being significant
- CIt is calculated on the log likelihood and enforces a penalty for complexity, is harshest for sample sizes greater than 7.39
- DIt is calculated on the log likelihood and enforces a penalty for complexity, is harshest for sample sizes lower than 7.39
79I am comparing two models, which each use different predictors to determine the happiness of participants. My sample size is 150, and my models are not nested. Which model comparison statistic will have the strongest penalty based on the number of parameters used in my models?
- AAIC
- BBIC
- CLRT
- DThey will all have the same amount of harshness
80In the attached image, which model seems to have the best fit?
- AModel A
- BModel B
- CWould need to know the sample size to answer
- DWould need to know whether the models are nested or not to answer
81A quadratic polynomial has how many degrees?
- A1
- B2
- C3
- D4
82If I want to know for my overall model, what proportion of the total variance is attributable to the fixed-effects portion, which effect size should I look at?
- AMarginal R2
- BMarginal F2
- CConditional R2
- DConditional F2
83If I want to know for a specific model parameter, what proportion of the total variance is attributable to the fixed and random effects, which effect size should I look at?
- AMarginal R2
- BMarginal F2
- CConditional R2
- DConditional F2
84If I am looking at the effect size of a random effect only, what should I expect the value of my marginal F2 to be?
- A0
- B0.5
- C1.25
- DThe same as my conditional F2
85A conditional F2 value of 0.45 would constitute a
- Anegligent effect
- Bsmall effect
- Cmedium effect
- Dlarge effect
Information
- Description
- Quiz on PSY4210 exam, created by Rachael Martin on 16/05/2022.
- Author
- Rachael Martin
- Language
- English
- Views
- 219
- Rating
- No ratings yet Rate this quiz