MCQs Flashcards
What does the F-ratio reduce the risk of?
Type I errors
Which of these statements is true for an ANOVA?
A. Both the dependent and independent variable need to be continuous
B. The DV needs to be continuous and the IV needs to be categorical
C. The DV needs to be categorical and the IV needs to be continuous
D. Both the DV and IV need to be categorical
B. DV - continuous, IV - categorical
Which of these isn’t an assumption of a one-way independent groups ANOVA?
A. Independent sampling
B. DV on an ordinal scale
C. Normally distributed data within levels
D. Homogeneity of variance
B. DV should be on an interval scale (or rational)
What does p<0.05 on Levene’s test mean?
There isn’t homogeneity of variance
What does Machuly’s test test for?
Sphericity: if variance within each level of the IV is similar, and covariance between pairs of levels is similar
How to test for skew and kurtosis?
Divide the raw statistic by the standard error (converts into a z-score). If this is greater than 1.96, then p < 0.05
What information does effect size provide?
What proportion of variance in the DV can be explained by the IV.
Gives an idea of relative scale of a significant effect
Whether they are clinically significant results
Which is the most accurate estimate of likely effect size? Why is it the most accurate?
Omega squared, because it adjusts for random error
What post hoc test should be done if there is no significant main effect in a one-way ANOVA?
None - only do post-hoc if results are significant
If Mauchly’s test isn’t significant, what do you do?
Read from the line below - usually Greenhouse Geisser
What is the family-wise error rate?
Probability of making type I errors
How to reduce the family-wise error rate in post hoc tests?
Bonferroni correction - adjust the threshold for significance
If there are three possible t-tests, using the Bonferroni correction what should the significance level be?
p < .05/3 –> p < .16
What post-hoc test should be carried out if wanting to be conservative and have a lot of comparisons to make?
Tukey
What post-hoc test should be carried out if there are both unequal variances and sample sizes?
Games-Howell
When would you use a planned contrast?
If you have prior predictions about the relationship between levels
What is there a higher risk of when doing a conservative post-hoc test?
Type II error (accepting the null hypothesis when it is false)
When would you do a 2-way ANOVA?
A. When want to compare the mean of a continuous DV between two or more groups in two categorical IVs
B. When want to compare the mean of two continuous DVs of groups in two categorical IVs
C. When want to compare the mean of a continuous DV between two or more groups in two continuous IVs
D. When want to compare the mean of two continuous DVs of groups in two continuous IVs
A two or more groups, in 2 categorical IVs
In a 2-way independent groups ANOVA, what does the Sum of Squares (Model) consist of?
SS A - variability of 1st IV
SS B - variability of 2nd IV
SS AxB - variability that they both share
What is an interaction?
When the impact of a given IV on the DV is conditional on levels within one or more of the other IVs in the model
What should be focused on the most of these results of a 3-way ANOVA: A. Significant 3-way interaction B. Insignificant 3-way interaction C. Significant 2-way interaction D. Significant main effect
A. Significant 3-way interaction - always look at the highest order interaction first, only explore if it is significant
Which of these statements is correct about an interaction plot?
A. DV on the X axis. One IV on the Y axis, other IV defining the lines
B. DV on the Y axis. One IV on the X axis, other IV defining the lines
C. DV defining the lines. One IV on the X axis, other IV on the Y axis
B
If there are parallel lines on an interaction plot, what does this suggest?
No interaction
Why does a repeated measures factorial ANOVA have more power than independent groups?
Because it also takes into account the between subjects variance, whereas the independent groups does not, only considering within-groups variance.
What does ANOVA (3x3) represent?
A. A 2-way ANOVA with 3 IVs
B. A 2-way ANOVA with 3 levels within each IV
C. A 3-way ANOVA with 2 levels within each IV
D. A 3-way ANOVA with 3 levels in each IV
B. 2 IVs, 3 levels in each
What does 2 straight, horizontal lines on an interaction plot imply?
No significant interaction and no significant main effects
What is the equation of the F-ratio?
Mean squares of model/mean squares of residual
What do you need to know to calculate the mean squares of model?
Sum of squares of model and degrees of freedom of model
How is the sum of squares (error) calculated for a repeated measures ANOVA?
SS R = SS W - SS M
SS W = variation within subjects
How many possible interactions are there in a 3-way ANOVA?
4 - one 3-way and three 2-way
What is a 3-way interaction?
The nature of a 2-way interaction differs depending on the level of a third IV
How to do an interaction plot for a 3-way interaction?
2 interaction plots, split by the 3rd IV
What is a covariate?
A factor that manipulates the variability in the DV but isn’t related to the IV
What is the advantage of taking into account a covariate?
Gives a more powerful test, as takes out unsytematic variance that isn’t associated with the model
When are ANCOVAs useful?
A. Quasi-experimental designs
B. When it is impractical to control for covariates in the design
C. When the covariate is related to the IV
A - in quasi-experimental difficult to control for covariates
If covariate is related to the IV, need to enter the covariate as another IV in the design
What assumptions need to be checked in an ANCOVA?
Homogeneity of regression (between CV and DV, on all levels of IV)
No multicollinearity of CV
Why does qualitative research have different approaches to conducting research than quantitative?
Due to the different assumptions about the nature of knowledge
What are the five methods of collecting qualitative data
Interview Naturally-occurring Observational Structured Collaborative
Why do relativists believe that data is generated, not collected?
They consider there to be many different truths, no solid facts, therefore need to co-construct ‘facts’, not collect them
Outline the four methods of analysing qualitative data
Discursive
Thematic
Structured
Instrumental
What is an inductive direction of analysis?
Bottom-up, derive theory/themes from the material. Opposed to theoretical, which starts from theories
What is a latent approach to analysis?
Don’t analyse the actual words, but what can be implied from the text, things that are implicit.
Pick the correct matching pairs:
A. Semantic - explicit, inductive - top-down
B. Latent - explicit, inductive - bottom-down
C. Semantic - explicit, theoretical - top-down
D. Latent - explicit, theoretical - bottom-down
C. Semantic - explicit, theoretical - top-down
What is human science?
The activity of humans studying other humans
What phenomena does human science incorporate?
People may change their behaviour as a result of being studied
Researchers aren’t independent of what is being studied
Which of these is not an aspect of qualitative research as a human science: A. Reflexive B. Inductive C. Situated D. Positivism
D positivism is the prediction, manipulation and control carried out within natural science
When should interviews be used?
When little is known about the topic
To explore quantitative data
For a very specific, detailed account of something
For a complex, not easily quantifiable topic
To explore how people frame their experiences
What is the best interview practice?
A. Letting the participant talk about what they want to
B. Trying to get the answers you want to hear from the participant
C. Not having an structure, and letting the participant guide the interview
D. Keeping the topic relevant, but not influencing what the participant is saying
D - should be specific to research question but not overinfluence what they say
What would mean the highest information power?
A. Case analysis, applied theory and dense specificity
B. Cross-case analysis, applied theory and dense specificity
C. Cross-case analysis, no theory and dense specificity
D. Case analysis, no theory and sparse specificity
A. case analysis, applied theory and dense specificity
Advantage and disadvantage of interviewer knowing nothing about the subject?
Adv - impartial
Disadv - difficult to gain rapport with participant as they cannot relate to each other
Advantages of focus groups?
Releases inhibitions
Activates forgotten details
Widens range of responses
Which of these isn’t an ethical issue regarding interviews?
A. Researcher safety
B. Ownership of the interpretation
C. Informed consent
D. Data isn’t put into sufficient context
D - not an ethical issue
Researcher safety - going into strangers houses
Ownership - researcher owns it, is that right?
Informed consent - don’t know exactly what you’re consenting to
Outline Elliott, Fischer & Rennie’s publishability guidelines
Owning one's perspective Situating the sample Ground in examples Credibility checks Coherencne General vs. specific research tasks Resonating with readers
How does Elliott, Fischer & Rennie say ‘owning one’s perspective’ should be addressed?
Author should specify their theoretical orientations and anticipations in advanced.
Author needs to recognise their values, interests and assumptions - and make them clear to reader
How can credibility checks be done?
Check understandings with the original informant
Compare other qualitative perspectives
Triangulation
Which of these isn’t a contingent problem of interviews (Potter & Hepburn):
A. Deletion of the interview
B. Too detailed transcriptions used
C. Not showing how the claims account for the specifics of the talk
D. Not providing the interview set up
B. too detailed transcriptions - they believe more detailed transcriptions should be used
Deletion of interviewer - not providing interviewer input in extracts
Interview set up - what category participant recruited under, and what participant has been told interview will be about
What is the difference between contingent and necessary problems?
Contingent - not a necessary feature of interviews, can be fixed
Necessary - difficult to deal with
Why is the footing positions of the interviewer and participants important?
Are the participants speaking as an individual, or on behalf or the group?
Is the interviewer the recipient of the answer, or speaking on behalf of them
How should the stake and interests be made clear?
By showing how the stake and interests were introduced in the interview set-up
How is evidence presented in qualitative analysis?
Examples from the data, e.g. quotes
How to chose the qualitative method?
Depends on the research question
What is the aim of grounded theory?
To generate a theory
What is the four stages of analysis in grounded theory?
Codes
Concepts
Categories
Theory
What is theoretical sampling?
Collect data, generate codes and analyse, then decide what data to collect next
What are the two approaches in discourse analysis?
Identifying the different types of discourse (we are observing reality)
Analyse the ways by which discourse is constructed (we are constructing reality)
What is the aim of conversation analysis?
To analyse the structure of verbal interaction
How is conversation structured?
Takes into account: Rules Norms Conventions In a particular culture
What is a nomothetic approach?
For large groups of people, generalises and generates laws about human behaviour
In opposition to idiographic approach
What is attempted to be understood in IPA?
People’s experiences, within the context they experience it
What is a ‘double hermeneutic’ process?
Dynamic, dual interpretation
Researcher - interpreting the participant’s interpretations of their own experience
What are problems with quantifying psychological data?
Quantifying assumes the attribute is ordered, with equal step changes (interval) - not always true
Some phenomena are too complex or unknown to be quantified
What are the features of participatory/action research?
Agenda for change
Emancipatory - help individuals free themselves from constraints
Collaborative - seeing the participants as expert
What is the hermeneutic circle?
Understanding the text both as a whole but also in its constituent parts
What are the two types of hermeneutics did Ricoeur say were needed?
Hermeneutics of faith (obbious meaning of text)
Hermeneutics of suspicion (be aware of influence of culture)
What is reflexivity?
Identifying how the researcher’s subjectivity influences the research - contextualise the researcher’s relationship with the phenomenon
What does reflexivity promote?
Transparency - reader can see how the researcher influenced the research
What is multiplicity? And why is it a problem in qualitative research?
How there are too many ways to speak about things
Means that something is always left out because something more can always be said, and in a different way
What is the downside of contextualism?
Accounts are so bound in context that they can’t be generalised
Why does the constructionist view of subjectivity find a problem with reflexivity?
View says that people aren’t transparent to themselves and can’t accurately report their inner feelings. Therefore, reflexivity isn’t possible
What is the problem with anonymity in qualitative research?
Assumes the participants are vulnerable and less powerful than the researcher, meaning they need protection
Often participants don’t want it to be anonymous, but it is enforced by ethics committee
What are the power issues in qualitative research?
There is asymmetrical power
Interviewer sets agenda and questions
Development of rapport/friendship just to get data
Asymmetrical disclosure
Interview can upset participant
Deletion of the interviewer - don’t mention questions in report
4 principles from the Code of Human Research Ethics (2014)
Respect for autonomy and dignity of persons
Scientific value
Social responsibility
Maximising benefit and minimising harm
Four problems Duncan found with interviews:
Can’t predict questions and responses
Nature of relationship –> sensitive disclosures
Difficult for participants to voice concerns or withdraw
Researcher know’s participant’s identities, complicates boundary issues
How did Bhati think qualitative and quantitative research could work together?
Theory generation by qualitative and theory testing by quantitative
What are repeated contrasts?
Comparison of each level in the variable against the previous level
Level 1 vs. level 2, level 2 vs. level 3 etc.
Sequential
What does the SS W represent?
Sum of squares within participants - variation within participants, used in repeated measures as want to look how participants change in different levels of IV
What is the total sum of squares in a repeated measure ANOVA?
Variation between subjects (SS B) and variation within subjects (SS W)
What validity does the measuring of covariates protect?
Internal
True/false: Covariates are continuous
true