Research Methods
Answer: b - As per (a) above but don't divide the participants into two groups; use the actual depression scores in the analyses and see if there is a relationship between depression and biting nails
Suppose you wanted to conduct a study to see if depressed individuals bite their nails more than non-depressed individuals. Which of the following would be the best way to proceed? a) Measure participants' depression with a questionnaire and ask them to give a rating of how much they bite their nails. Then classify participants as 'depressed' or 'non-depressed' on the basis of their questionnaire scores. We could then see if there was a difference in how much they bit their nails b) As per (a) above but don't divide the participants into two groups; use the actual depression scores in the analyses and see if there is a relationship between depression and biting nails c) This sort of study is impossible to carry out and so we couldn't proceed with it d) None of the above
Answer: a) Driving errors The criterion (in SPSS) is termed the dependent variable, as the analysis is aimed to infer the effect that the predictors might have on it. Answer b) 67% (rounded up to the closest whole percent) The R Square value (R2)—which is the square of R—provides the answer for the SAMPLE (66.7% can be rounded up to 67% as the nearest per cent); the amount of variation in driving errors that can be explained by the amount of alcohol. It is not the 'Adjusted R Square', which is the explained variance estimated for the population.
Table 3 shows part of the SPSS output for a linear regression analysis. a) What is the criterion/outcome variable? b) What is the variance explained for this sample in percentage?
Answer: b - Beta Beta is so-called standardised coefficients as it converts the unstandardised B (different variables are likely to be measured by different scales with different ranges) into standardised units (standard deviations) for comparisons. In the example in 8) above, where the Beta for attitudes towards EU is 1.142 and that for British identity is -.345. This means as attitudes towards EU increases by 1 SD, attitudes towards the Euro increases by 1.142 SD, while as British identity increases by 1SD, attitudes towards the Euro decreases by .345 SD.
The best parameter to use for comparing different predictors' contribution towards predicting the criterion is: a) B b) Beta c) t-value d) p-value
Answer: 2 - F(2,30)=12.05, p<.001 The Greenhouse-Geisser row's figures are taken as the test of sphericity for cook x dish is significant. Rounding up the df's of 1.566 and 29.753 will give the figures of 2 and 30 to fill in the brackets.
The contest results from the audience scoring have come in for the different cooks and their mains and desserts. Below are part of the SPSS output of the descriptives, main ANOVA and post hoc tests on the cooks. Refer to Tables 21 to 24 How would you write up the cook x dish interaction in text? 1) F(2,38)=12.05, p<.001 2) F(2,30)=12.05, p<.001 3) F(2,32)=12.05 p<.001 4) F(1,19)=12.05, p=.003
Answer: Yes If look under the Sig. column against all of Cook, Dish and Cook*Dish in the Tests of Within-subjects Effects table, all the p-values are smaller than .05
The contest results from the audience scoring have come in for the different cooks and their mains and desserts. Below are part of the SPSS output of the descriptives, main ANOVA and post hoc tests on the cooks. Refer to Tables 21 to 24 The main effects of cook and dish and the effect from the cook x dish interaction are all significant. Is this true?
Answer: Cook partial eta sq. is .736, larger than that of the other two, and meaning that nearly 74% of the variance in scores given by the audience was accounted for by the different cooks
The contest results from the audience scoring have come in for the different cooks and their mains and desserts. Below are part of the SPSS output of the descriptives, main ANOVA and post hoc tests on the cooks. Refer to Tables 21 to 24 Which of the above effects has the largest effect size?
Answer: Cook and dish
The cooking contest has become an event where the studio audience also have an input in the contest's outcome. In the semi-final, each of the three cooks (cook A, cook B, cook C) has to cook two dishes (main course and dessert) at the same time in front of the live audience. In this part of the contest, the audience (n=30) will each give a score out of 10 for each dish (a score for main, a score for dessert for each cook). The contest's chair is interested to see whether there are differences in scores between the mains and desserts and between cooks. The dependent variable here is obviously the score. What are the independent variables/ factors?
Answer: 3 - 3 by 2 repeated measures 3 by 2 as cook has three conditions/levels (cook A, cook B, cook C) and dish has two (main course, dessert). The numbers reflect the number of conditions/levels.
The cooking contest has become an event where the studio audience also have an input in the contest's outcome. In the semi-final, each of the three cooks (cook A, cook B, cook C) has to cook two dishes (main course and dessert) at the same time in front of the live audience. In this part of the contest, the audience (n=30) will each give a score out of 10 for each dish (a score for main, a score for dessert for each cook). The contest's chair is interested to see whether there are differences in scores between the mains and desserts and between cooks. The independant variables are the cook and the dish. What kind of design would this make? 1) 2 by 2 repeated measures 2) 3 by 3 repeated measures 3) 3 by 2 repeated measures 4) None of the above
Epistemology concerns the theory or nature of 'knowledge' (as in what it is) —whether it is already 'out there' to be found and measured (a stance taken by the 'quantitative' camp), or to be 'constructed' by human meaning-making (the 'qualitative' stance).
The epistemology, not just methods, of qualitative and quantitative research is often contrasted. What is epistemology?
Answer: a - Refers to how consistent the test is within itself assuming that it consists of multiple items measured on a scale
The internal reliability of a psychologist test: a) Refers to how consistent the test is within itself assuming that it consists of multiple items measured on a scale b) Can be estimated by splitting the test into two halves and calculating how the two correlate with each other c) Can be estimated by the Cronbach's Alpha value, which is a measure of how far the scores vary between the items in the test in relation to how far scores for the whole test vary d) All of the above
Answer: Yes There is a main effect of judge in the between-subjects effects table (p<.001)
The scores for the cooks have come in for the contest. Refer to Tables 27-30 and Graph 31. From the ANOVA results, have the audience and chefs given significantly different scores to each cook?
Answer: Cook B Although the audience tend to give the cooks higher scores than the chefs, they are nevertheless in agreement with them on which cook is better (as it is clear from the plot as well as the descriptives table) and that is Cook B.
The scores for the cooks have come in for the contest. Refer to Tables 27-30 and Graph 31. Which cook is rated as the better cook?
Answer: Judge Its partial eta sq: .321
The scores for the cooks have come in for the contest. Refer to Tables 27-30 and Graph 31. Which of the main effects has a larger effect size?
C (from descriptives)
The scores for the three cooks' dishes have come in from the judges. Refer to Tables 15 to 20 for SPSS output for the results. The dish by which cook has the highest score?
Answer: 20 The mean is calculated by the sum of the scores divided by the number of scores: (10+10+15+15+20+50) ÷ 6 = 20 (can easily do it in head)
The scores from a sample of participants on a memory test were: 10, 10, 15, 15, 20, and 50, where 50 is maximum possible score for the test (full mark). The mean score for this sample is:
Answer: True. Indeed that's what (co-)create discourses and where/how they function.
They are located in and by society, culture and history and lead to the construction of knowledge and assumed 'realities'. True or False?
Answer: False. They can refer to an organised way of speaking about a phenomenon.
They are loosely ordered and constructed differently by people even when talking about shared phenomena. True or False?
Answer: False. They are not neutral, but are not necessarily prejudicial/ biased either.
They are never neutral, but always loaded with potentially prejudicial worldviews that are taken for granted. True or False?
Answer: False Although qualitative data may be COLLECTED, the analysis may not be done 'qualitatively'. It is possible to 'quantify' textual (or behavioural/ video) data, for instance, by content analysis ('counting' how many times something has been said (or how long episodes of behaviours have taken place), and then analyse the pattern. This concerns the 'big' and 'little' q; how far one applies qualitative research (as a whole approach or way of thinking, or just use it as a method of data collection) - methodology vs. method.
True or False? All qualitative research goes through non-statistical analysis and ends up as textual results.
Answer: False If a qualitative research 'approach' is used to the full, the role of researcher IS important as the researcher (who comes with their own assumptions, interests and knowledge) will be deeply involved from formulating the research enquiry and questions, collecting data and analysing (interpreting) them. This is why so-called 'reflexivity' (reflecting on own role as a researcher and the approach itself) is useful, and it is recommended that such reflections be written about in the research output (e.g. report) to make things more 'transparent'.
True or False? In qualitative approaches, the role of the researcher is not important.
Answer: False Finding out of 'objective facts' to falsify a priori theory/assumptions is precisely what qualitative approaches are NOT about; the school of thoughts advocates that there is no such thing as 'objective facts'. Also, the idea of such research is to see how participants 'construct' the reality (be it a subjective point of view, experience or lived phenomenon) through making sense (meaning-making) of the data, rather than using the data to test pre-exiting theories.
True or False? Qualitative research aims to find objective facts that can either support or refute theories.
Answer: False A small sample is typical as the approach is about making sense of subjective experiences rather than measuring 'objective' generalizable trends, so in-depth data from a few (or even single cases) are sought rather than a large sample
True or False? Qualitative research typically still requires big sample; over 10 as normal/minimal.
Answer: b - In exceptional circumstances where the integrity of the research would be compromised if participants know the full nature of research from the start
Under what circumstances may deception be used in research? a) No circumstance b) In exceptional circumstances where the integrity of the research would be compromised if participants know the full nature of research from the start c) If participants are told which condition they are assigned to, including placebo d) If very young children or individuals with impaired capacity are participants as they cannot really understand the research
Answer: Order effects are consequences of a within-participants design. Results would not be based on the result of manipulating the IV but from practise of the task that participants did or fatigue / boredom, learning what is being tested etc.
What are 'order effects'?
Answer: Seeing if there are differences on the dependent variable between conditions of the independent variable. Unlike experimental designs there is not random allocation of participants to the various conditions of the individual variable.
What are Quasi-experimental designs?
Answer: Simply allocating people to categories. E.g. Gender, Occupation, Ethnicity, Favourite Colour.
What are categorical variables?
Answer: A specific type of extraneous variable that is related to both the main variables that we are interested in.
What are confounding variables?
Answer: Can take on absolutely any value within a range. E.g. Temperature, Car Speed, Typing Speed.
What are continuous variables?
Answer: Where the experimenter manipulates one variable called the independent variable to see what effect that has upon another variable called the dependent variable.
What are experimental designs?
Answer: Variables that might impact on the variables we are interested in but failed to take these into account. E.g. Characteristics of person, gender, time of day test was conducted.
What are extraneous variables?
1. The relationship between the variables should be linear (can be represented by a straight line)...otherwise there cannot be a best-fit line, which is how the regression model works; predictions are generated from this line. 2. The criterion/outcome variable should be measured on an interval/continuous scale.
What are the two basic conditions or assumptions that should be satisfied for one to use a bivariate linear regression?
Answer: Discursive psychology and Foucauldian discourse analysis (FDA). The former is concerned with how persons use discursive resources to achieve interpersonal objectives in social interactions; the latter the kinds of resources/ understanding people draw on to make sense of the world and position themselves within it.
What are the two main strands of Discourse analysis approaches?
Answer: Foucauldian discourse analysis
What does FDA stand for in qualitative research?
Answer: Systematically vary the order in which participants take part in various conditions of the independent variable. This would happen in a within-participants design.
What is 'counter balancing'?
Answer: Interpretative Phenomenological Analysis.
What is IPA short for?
Answer: Group of participants different to each testing of the IV. Larger group of participants, split into groups, take part in different conditioning.
What is a 'between-participants' design?
Answer: Having the same participants in every condition of the independent variable.
What is a 'within participants' design?
Answer: Designs that investigate relationships between variables.
What is a correlation design?
Answer: Can only take on certain discrete values in a range. E.g. Number of cars owned, number of goals scored, number of children in a family.
What is a discrete variable?
Answer: Prediction of how specific variables might be related to one another or how groups of participants might be different from each other
What is a research hypothesis?
Answer: C - Semi-structured interviewing Semi-structured interviewing is often used where there is an interview schedule based on the research questions, which are open-ended to allow the interviewee to expand on their views, while the setting gives the opportunity to probe/ prompt during the interview for more in-depth information or explore explanations, or the direction of questioning might change depending on the interviewee's experience, point of views or strategy
What is most typical data collection method used by researchers aiming to use the approaches of IPA or Discourse Analysis? a) participant observation b) video-taping interactions c) semi-structured interviewing d) asking participants to write a diary of their experiences
A correlation gives the direction and strength of the association between two variables. A regression allows one to predict scores on one variable (the one set as criterion/outcome variable) from scores on another (the one set as explanatory/predictor variable). If applied to data obtained from an experimental situation, it can be used to 'suggest' that the scores on one variable influence the scores on another; i.e., 'infer' causal relationships (Dancey & Reidy, 2011; p. 385).
What is the key difference between correlation and regression?
Answer: 0.1% chance due to sampling error .01 is 1% (1 out of 100); .001 is 0.1% (1 out of 1000) chance due to sampling error
What is the significance level p < .001 expressed as a percentage? (fill in the blank) ________________ % chance due to sampling error
Answer: There are various. These are not limited to the following (as examples): The tendency for FDA (due to its focus on language) to over-rely on people's talk and therefore also their conscious thought and experience, but to overlook certain entities or ideas such as the body and personal agency.
What may be a limitation of FDA?
Answer: The 'rhetorical' strategies used by DP can include any of (but not exhaustive to): construction (how people construct their accounts out of pre-existing linguistic resources as 'interpretative repertoires'), variation (how people are inconsistent in conversations), context (how people vary with context), function (how people manage stake, interest, blame and their position and ward of accusations, etc.).
When applying one of the above approaches in analysis, the researcher pays attention to how people manage their stake and interest in conversations as 'rhetorical' strategies on a micro/local level. Name two of such strategies.
Answer: d - All of the above In a normal distribution, the mean, median and mode will coincide on the same mid-point value
When scores are normally distributed, the most typical value is: a) median b) mode c) mean d) all of the above
Answer: b - The value of a is 3
When we perform a linear regression analysis, the equation that represents how two variables are related is: Ŷ = a + bx. If we know the value of the constant/intercept is 3, then: a) The value of Ŷ is 3 b) The value of a is 3 c) The value of b is 3 d) The value of x is 3
Answers: B - Does ethnic identity influence self-esteem, mental health and risk-taking? D - To what extent does youth disaffection predict substance abuse? The phrasing of both suggests that the enquiries will involve measuring or quantifying an operationalised concept, or inferring cause-and-effect (e.g., does A 'influence' B or 'to what extent' does A 'predict' B). Qualitative approaches are not concerned with these ideas when they ask questions in a more open-ended way - here, 'how' (esp. in combination with how people 'construct') and 'in what ways' - so that the depth or complexity of the issues could be captured.
Which TWO of the following research questions are NOT suitable for a qualitative approach? a. How do working-class white youth construct their 'English' identities? b. Does ethnic identity influence self-esteem, mental health and risk-taking? c. In what ways do youth become religiously radicalised? d. To what extent does youth disaffection predict substance abuse?
Answer: c - Whether a discourse may have universal relevance or application FDA researchers are primarily interested in what discourse does or can do - which includes (but are not exhaustive to) such functions in 1, 2 and 4, but they are not concerned about discourse demonstrating 'universality' (with discourse being inherently not universal).
Which aspect of discourse are FDA researchers NOT particularly interested in? a) How people's identities and worldviews are produced through discourse b) What people may achieve with their use of discourse c) Whether a discourse may have universal relevance or application d) How objects, situations, experiences, or 'realities' are shaped by discourse
Answer: d - All of the above
Which of the following are problems associated with dichotomising continuous variables? a) Loss of experimental power b) Spurious effects may occur c) There is a serious loss of information d) All of the above
Answer: d - All of the above
Which of the following are problems associated with within-participants designs? a) There is an increased likelihood of practice or fatigue effects b) Participants are more likely to guess the nature of the experiment c) They cannot be used with quasi-experimental designs d) All of the above
Answer: d - All of the above
Which of the following are true of correlational designs? a) They have no IV or DV b) They look at relationships between variables c) You cannot infer causation from correlations d) All of the above
Answer: c - Temperature
Which of the following constitute continuous variables? a) Number of times a score of 180 is achieved in a darts match b) Gender c) Temperature d) All of the above
Answer: d - All of the above
Which of the following could be considered as categorical variables? a) Gender b) Brand of baked beans c) Hair colour d) All of the above
Answer: c - Correlational design
Which of the following designs is least likely to enable us to establish causal relations between variables? a) Experimental design b) Quasi-experimental design c) Correlational design d) Within-participants design
Answer: A - Questionnaire Scores Pre-planned quantitative/numerical data criterion is not qualitative
Which of the following does NOT consist of qualitative data? a) questionnaire scores b) news bulletins c) observations filmed by a mobile phone d) messages on an Internet discussion forum
Answer: d - Students' ranking in a class Interval or scale data fall into a continuous numerical scale where the distance between every two consecutive points is be the same along the scale. IQ (for most of the tested psychological constructs) falls into one such scale, and so do reaction time and age (in units of second or years—they take on a continuous scale). Ranking is an ordinal variable where the participants are 'ordered' according to their places compared against one another rather than measured in relation to one continuous scale (in fact, ranking the data is how the non-parametric tests operate as you have learnt).
Which of the following does NOT represent interval/scale data? (choose ONE) a) students' IQ test scores b) reaction time c) participants' age d) students' ranking in a class
Answer: a - Withholding the full nature of the study until participants are debriefed
Which of the following is NOT necessarily a breach of research ethics? a) Withholding the full nature of the study until participants are debriefed b) Telling participants they would forfeit their reward for participation if they do not complete the procedure c) Recording participants' responses without their consent even though they have agreed to be interviewed d) Tell a child participant how well they have performed in the task compared with another participant as part of debriefing
Answer: c - Sharing participants' data with another researcher, who has proposed to carry out a different study, but its aims overlap somewhat with your own
Which of the following is a serious breach of research ethics? a) Asking a consented group of children to produce a piece of writing, explaining to them that the writing will be entered into a fictitious 'competition' to examine whether the very anticipation of competition will alter performance (measured by the length of writing and spelling accuracy) b) Filming the behaviour of secondary school children in their classroom with a video camera after the children and their parents have signed consent to take part in such a study c) Sharing participants' data with another researcher, who has proposed to carry out a different study, but its aims overlap somewhat with your own d) Giving child participants confectioneries as rewards for their participation
Answer: A - Cross-case analysis Cross-case analysis involves making connections: spotting for patterns across cases (as 'shared understanding' of experiences/meaning), with themes that may be evidenced by each participant. This would be difficult if there is just one person (though different experiences from the same person might evidence the same set of themes).
Which of the following may be difficult with a single-case study? a) cross-case analysis b) grouping super-ordinate themes c) initial noting d) writing emergent/constituent themes
Answer: d - All of the above
Which of the following might be suitable IVs in a quasi-experimental study? a) Gender b) Whether or not someone had Generalised Anxiety Disorder c) Students versus non-students d) All of the above
Answers: A - What is it like to go through a divorce? & C - An exploration into the experience of living in a three-generational household The topics are very much about lived experience and/or meaning-making (exploring a phenomenon; the perspective from those going through/having gone through or living in a certain situation; what something is 'like'). The other topics or statements refer to inferred cause-and-effect or prediction (would something affect something else... how much would it do; do these things predict something else... how much would they explain it) and so pertain to quantitative approaches.
Which of the following research questions or enquiries lend themselves well to the use of qualitative approaches? a) What is it like to go through a divorce? b) Does a child's temperament type have an impact on future personality? c) An exploration into the experience of living in a three-generational household d) The effect of early intervention on autistic children's social development. e) Predictors of bullying in Cyberspace: The role of personality type, social class and academic achievement.
Answer: d - Both (a) and (b) above
Which of the following statements are true of experiments? a) The IV is manipulated by the experimenter b) The DV is assumed to be dependent upon the IV c) They are difficult to conduct d) Both (a) and (b) above
Answer: b - Predictive validity - how well the scores of a (particularly newer) test might correlate with those of related constructs or predict future/outcome variables
Which one of the following is a form of 'criterion validity': a) Construct validity - a broad idea regarding the very existence of the construct being measured as a variable in testing b) Predictive validity - how well the scores of a (particularly newer) test might correlate with those of related constructs or predict future/outcome variables c) Content validity - whether the items of a test are representative of the variable being studied according to experts or prior research findings d) Face validity - whether it is obvious to researchers and test-takers what a test is measuring
It was developed by Jonathan Smith
Who pioneered IPA?
Answer: a - There is not necessarily a causal relationship between getting up early and amount of work done
You have conducted a study that shows that the earlier people get up, the more work they get done. Which of the following are valid conclusions? a) There is not necessarily a causal relationship between getting up early and amount of work done b) People who get up early have a need to get more work done c) Getting up early is the cause of getting more work done d) Both (b) and (c) above
Answer: False Not necessarily. The F and associated p that is significant only means that there is a treatment effect (the effect of the between-groups IV) on the whole. It does not specify which group(s) differ from which other(s). For instance, it may be that only group 1 differs significantly from group 2, but it does not differ significantly from group 3 and group 2 does not differ significantly from group 3 either. Still, the significant difference between groups 1 and 2 is enough for the overall F to be significant. Any conscientious researcher concerned with the detailed effect of the treatment/IV will do follow-up/post-hoc tests in this situation to find out precisely whether every two groups differ significantly from each other.
A 'significant' (F) result from a one-factor between-groups ANOVA must mean that all groups are significantly different from each other. True or False?
Answer: b - Able to take any value within a range of scores
A continuous variable can be described as: a) Able to take only certain discrete values within a range of scores b) Able to take any value within a range of scores c) Being made up of categories d) None of the above
Answer: a) Scoresfor the dishes given by judges. Answer: b) The judges/gourmet chefs as they give the scores (the DV). Answer: c) The problem concerns order effect: judges' impression of the second dish (always by contestant B) may be influenced by that of the first dish (always by contestant A), or impression of the third by that of second). The order of those dishes to be presented may be counter-balanced between the judges so that three of them get the dish cooked by A first, another three get the dish cooked by B first, and final three get the dish cooked by C first. (if a sample of judges) is large enough, one may even randomise the presentation so that by chance each will have a roughly equal probability of getting any of the three dishes.)
A cooking competition is happening in which, in the first round, a panel of nine gourmet chefs as judges will be presented with the same famous dish cooked by three different amateur contestants (A, B and C). The dishes will be shown in the same order (A, B, C) to the judges, who will each give an overall score to each dish. a) What is the dependent variable? b) Who are the 'participants'? c) The audience argue that there is a problem with the order of the dishes' presentation. What is the problem (in research design terms) and what may be done about it?
The problem concerns order effect: judges' impression of the second dish may be influenced by that of the first dish (or that for the third by that for first/second). The order of the dishes to be presented may be counterbalanced between the judges so that three of them get the dish by cook A first, 3 get dish by cook B first and 3 get dish by cook C first. (if the sample (of judges) is large enough, one may even randomise the presentation so that by chance each will have a roughly equal chance of getting any of the 3 dishes.)
A cooking contest is happening where in the first round a panel of 9 gourmet chefs will be presented with the same famous dish cooked by three different contestants (A, B and C). The dishes are shown to the judges in the same order (A, B, C) who will give an overall score out of 10 to each dish. The audience argue that there is a problem with the dishes' presentation (always showing A first, then B & C). What may be done about it?
Score for the dishes given by judges
A cooking contest is happening where in the first round a panel of 9 gourmet chefs will be presented with the same famous dish cooked by three different contestants (A, B and C). The dishes are shown to the judges in the same order (A, B, C) who will give an overall score out of 10 to each dish. What is the dependent variable?
Gourmet chefs; they give the scores (DV)
A cooking contest is happening where in the first round a panel of 9 gourmet chefs will be presented with the same famous dish cooked by three different contestants (A, B and C). The dishes are shown to the judges in the same order (A, B, C) who will give an overall score out of 10 to each dish. Who are the 'participants'?
Answer: a - Reviewing all research involving humans (and non-human animals by some) conducted by persons or groups accountable for the committee's institution
A research ethics committee is responsible for: a) Reviewing all research involving humans (and non-human animals by some) conducted by persons or groups accountable for the committee's institution b) Issuing professional warnings and fines to unethical researchers c) Vetting suitability of individual researchers for the institution's projects according to the individuals' research experience and credentials d) Vetting suitability of individual participants for the institution's projects according to their capacity to understand the nature of research and to give informed consent
Answer: Mann-Whitney. As the non-parametric alternative to the independent t-test.
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. If a non-parametric test needed to be used, which test would it be?
Answer: Between. Two groups of children are tested: one drink, the other not.
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. What design (within- or between-participants) design does this study use?
Answer: Maths test score. This is the V supposed to be affected (by drinking water).
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. What is the dependent variable (DV)?
Answer: Independent-samples (unrelated) t-test. For between-participants/groups.
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. What kind of parametric test would be suitable for testing this difference?
Answer: (Formally) The statement may be like this: It was predicted that children that had drunk water would have higher scores in the maths test than children that had not drunk water. (Often, one writes up the research when it is complete, thus past tense is used.)
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. What would a directional (1-tailed) hypothesis be for this study?
Answer: d - Box plot It shows the means as well as the percentile variation around them. Histogram is better for focusing on the distribution even though the means can be presented. Scatterplot is for exploring the relationship between two interval/scale variable. A line graph can show the means, but not spread, but useful for 'interactions' (see later learning materials in the module).
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. Which of the following type of graph is best for initially calibrating the difference in means and spreads between two variables? a) Histogram b) Scatterplot c) Line graph d) Box plot
Answer: Water or drink (or similar). The variable that is manipulated and is predicted to have an effect (on DV) is whether children drink water, so the IV term ought to summarise that fact.
A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. What is the independent variable (IV) in this study?
Answer: b - Quasi-experimental variable Participants' religion is a quasi-experimental V as it is not manipulated by an experimenter (i.e., not an experimental variable)—it 'comes with' the participant, but is expected to affect their attitudes towards drinking and the amount that they drink (the DVs). It is not a correlational V as it falls into different categories—a correlational design is used for looking at the relationship between two interval/scale variables.
A researcher has collected data on participants' religion (Catholic, Jewish, Hindu, no religion, etc.) and drinking habits and attitudes towards drinking (alcohol), because he expects that different religious groups may drink different amounts of alcohol and may have different attitudes towards drinking. In this case, participants' religion is a: (choose ONE) a) Experimental variable b) Quasi-experimental variable c) Correlational variable d) Dependent variable
Answer: a - The amount of alcohol drunk was related to the home team's ability to score goals but we cannot say it caused the team to score the goals
A researcher has just conducted a correlational study investigating the relationship between the amount of alcohol drunk by fans of the home team before a football match and the number of goals scored by the home team. They found that there was a relationship between the two variables. Which of the following statements are valid? a) The amount of alcohol drunk was related to the home team's ability to score goals but we cannot say it caused the team to score the goals b) The home team's ability to score goals is related not to the amount of alcohol but to the amount of cheering by the drunken fans c) The increase in the amount of alcohol drunk caused an increase in the number of goals scored d) All of the above
Answer: a - the standard deviation will be greater in the amateur group than in the top-20 group It is likely that SD will be greater among the amateurs than the top-20 as the latter will comprise the fastest runners in the race that deviate slightly around the mean of their elite group (indeed in international marathons, men that place top 20 often complete within 2½ hours). Members from the general public have a much wider 'spread' in time, especially when they do different amounts of training and differ widely in fitness, unlike the elites that do it for a living. Put simply, the top-20 are not like the average man as far as Marathon running is concerned.
A researcher is collecting data on the completion times of the world's top-20 male Marathon runners and those from 20 male amateur runners randomly selected from the public that have entered themselves for the next London Marathon. It is likely that: (choose ONE) a) the standard deviation will be greater in the amateur group than in the top-20 group b) the standard deviations will be equivalent in both groups c) the standard deviation will be greater in the top-20 group than in the amateur group d) none of the above
Answer a: .89 The R value gives the correlation between all the predictors entered (here, British identity and attitudes towards the EU) and the criterion. Answer b: Yes The F-value (52.05) and its associated probability (p-value) of <.001 means that this result is highly unlikely (less than 1 in 1000) to be due to sampling error, assuming the null hypothesis to be true (the null hypothesis is that the variables are unrelated). In other words, the prediction for attitudes towards joining the Euro from the combination of British identity and attitudes towards the EU is significant (F (2, 27) = 52.05, p < .001).
A researcher used a questionnaire to obtain information about participants' age, national (British) identity score, attitudes towards the EU, and attitudes towards (joining) the Euro, with the aim to predict the attitudes towards the Euro from the other variables using multiple regression. Before that, she looks at the correlations between these variables. The following is her SPSS output: Refer to Table 4 Above, the researcher goes on to conduct a regression analysis with British identity and attitudes towards the EU as the predictor variables, and attitudes towards joining the Euro as the criterion/outcome variable. Below is the SPSS output of the regression model summary and ANOVA: Refer to Tables 5 & 6 a) What is the correlation between the criterion and the other variables? b) State if it is true or false that the ANOVA shows that the regression model based on attitudes to the EU and British identity is better than would be expected by sampling error/ chance?
Answer: c - Enter only British identity and attitudes to EU as predictors into the regression analysis
A researcher used a questionnaire to obtain information about participants' age, national (British) identity score, attitudes towards the EU, and attitudes towards (joining) the Euro, with the aim to predict the attitudes towards the Euro from the other variables using multiple regression. Before that, she looks at the correlations between these variables. The following is her SPSS output: Refer to Table 4 Based on the correlation results, the researcher should: a) Enter all of the predictors into the regression analysis b) Enter only attitudes to EU as predictor into the regression analysis c) Enter only British identity and attitudes to EU as predictors into the regression analysis d) None of the above
Answer: False. Reflexivity is important in both approaches to render more accountability.
A researcher's reflexivity of his/her role/ experiences/ views etc. is even more important when applying FDA than IPA. True or False?
Answer: Idiography Idiography refers to the focus on what is 'distinct' in individuals (versus what is a general trend among a whole group). This is why more details about required in the data and the context.
A small sample is typical of research using IPA. Even one participant, as single 'case study', is favoured. What theoretical influence underpins this practice?
One confounding variable is order effect. Some Ps might be better off doing tasks with caffeine and some without and at the moment all of them perform without coffee first. One may 'counterbalance' the order so that half of the Ps do the task without coffee first then with coffee whilst the other half do it with coffee first, then without. There will need to be a gap in-between though as it takes time for the caffeine to wear off for the latter group. Also, although two different trials of word-picture matching are used, they are still the same task so Ps have had practice from the first time round. One may use two different variants of the task for each condition to reduce the practice effect somewhat, but nevertheless as long as both require the same skills, Ps will get practice. Another confounding is demand effect. Even without explicitly telling Ps that the study is about testing the effect of caffeine, they can easily tell, by doing tasks after no coffee and after coffee that they are being tested about effects of taking coffee. Also, it might not be the effect of caffeine, but 'knowing' that they have taken coffee (a psychological effect) that makes them perform in a different way. One way to resolve this is to add a 'placebo' condition where a drink that tastes like the instant coffee (e.g. decaf coffee) is consumed before Ps perform the task (if Ps perform comparably in this condition to how they do in the caffeine condition, then there is a placebo effect).
A study tests the effect of caffeine on reaction time. A group of participants perform on a word-picture matching task with randomised trials without coffee. Then they drink a cup of instant coffee before the same matching task. The differences in reaction time between the two groups were analysed. Give two confounding variables that might have been present in the study? Give suggestion to how they might be eliminated or effects minimised.
Answer: b - Within-participants design It was a within-participants design as each participant performed both conditions (coffee or no coffee)
A study tests the effect of caffeine on reaction time. A group of participants perform on a word-picture matching task with randomised trials without coffee. Then they drink a cup of instant coffee before the same matching task. The differences in reaction time between the two groups were analysed. This study used a: a) Correlational design b) Within-participants design c) Between-participants design d) Repeated measures design
Answer: c - 67%
According to Streiner (2002), how efficient are studies that dichotimise continuous variables compared with studies that do not? a) 100% b) 95% c) 67% d) 50%
Answer: C - Initial noting Initial noting: involves reading and re-reading (getting very familiar with) in detail each transcript at a time before then noting down (on the margin) initial themes of potential interest/importance to the research questions, using words (comments) that appropriately describe the script.
Although not a prescribed process, what is recommended to do FIRST when analysing data using IPA? a) cross-case analysis b) grouping super-ordinate themes c) initial noting d) writing emergent/constituent themes
Homogeneity of variance between conditions, and that the dependent variable should be on an interval scale (a continuous variable).
Apart from the data satisfying normal distribution, what are the other two most basic conditions or assumptions required for ANOVA to work well?
Sphericity. It means a measure of the correlation between scores in different conditions. This is again because of the fact that the same participant is taking all the conditions available in the study. Therefore, there is a chance that their performance/score in one condition is related to nother/others for the factors/ independent variables tested
As between-participants ANOVA, the basic assumptions of normal distribution and homogeneity of variance are still important, but there is further an extra assumption. What is that called?
Answer: False. Neither is seen as such because achieving 'objectivity' is neither's goal.
As far as qualitative approaches go, FDA is generally perceived as more 'objective' than IPA. True or False?
Answer: Sphericity It is a measure of the correlation (a relationship) between scores in different conditions. This is again because of the fact that the same participant is taking all the conditions available in the study. Therefore, there is a chance that their performance/score in one condition is related to another/others for the factors/IVs tested.
As in between-participants ANOVA, basic assumptions of normal distribution and homogeneity of variance are still important, but there is an extra condition in repeated-measures ANOVA...what is that called?
Answer: True The F ratio = Between-groups variance Within-groups variance. So, indeed if the bottom element is increased (due to higher individual differences between the participants within each group/condition, for instance), this will offset the treatment effect of the IV (between-groups) and the F ratio will be smaller; the effect will be weaker.
As the ANOVA's test statistic, F ratio, is a ratio of between-groups variance to within-group variance, a larger within-groups variance (greater 'spread' of scores within each group), while the between-groups variance the same, would mean the F ratio will be smaller. True or False?
Answer: Cook A Clearly in the graph, all parts of the line is higher than the other lines.
As there is a significant cook x dish interaction in the ANOVA above, a line plot is presented and simple-effects t-tests to follow up. Refer to Graph 25 and Table 26 Going by the plot, which cook has his/her two dishes scored similarly while the other two cooks have one dish judged more favourably than the other?
Answer: Only Cook C A Bonferroni correction needs to be made as three t-tests have been used (acceptable p = .05 ÷ 3 = .0167). Cook B's p=.019 is still larger than the acceptable alpha so the difference is not significant. Only Cook C's two dishes are rated significantly differently by the audience.
As there is a significant cook x dish interaction in the ANOVA above, a line plot is presented and simple-effects t-tests to follow up. Refer to Graph 25 and Table 26 Which cook(s) has/have their two dishes rated significantly differently by the audience?
Because the analysis is a repeated measures/within-participants design (each cook and his/her dishes are judged by the same participants, so an equivalent t-test for this design is used).
As there is a significant cook x dish interaction in the ANOVA above, a line plot is presented and simple-effects t-tests to follow up. Refer to Graph 25 and Table 26 Why has a 'paired samples test' (a variant of t-test) been used as the simpleeffects test?
Answer: a - Either quasi-experimental or experimental
Between-participants design can be: a) Either quasi-experimental or experimental b) Only experimental c) Only quasi-experimental d) Only correlational
Answer: c - Participants performing or behaving in ways that correspond with what they believe to be the research or researcher's purpose or expectations
Demand characteristics in research refers to: a) Participants avoiding questions that are too sensitive or intrusive to them b) Participants wanting to present themselves in a good light by avoiding or lying in items that are about sensitive issues c) Participants performing or behaving in ways that correspond with what they believe to be the research or researcher's purpose or expectations d) Factors in the research that bias or mislead the participants towards a certain pattern of responding
Answer: a - Participants behave in the way they think the experimenter wants them to behave
Demand effects are possible confounding variables where: a) Participants behave in the way they think the experimenter wants them to behave b) Participants perform poorly because they are tired or bored c) Participants perform well because they have practised the experimental task d) None of the above
Answer: c - Random allocation of participants to conditions
Experimental designs are characterised by: a) Fewer than two conditions b) No control condition c) Random allocation of participants to conditions d) None of the above
The unstandardised betas are the ones to be put in the regression equation (see section The equation in the Regression analysis chapter). In general terms, a multiple regression's equation is: Ŷ = a + b1x1 + b2x2 + b3x3 + ... ... where the x's are the different predictors and b's their associated 'slopes'. These b's are in fact the unstandardised betas (B) shown in the coefficients table computed by SPSS. In the 'Euro' example above, criterion Ŷ is attitudes towards the Euro, and there are two predictors, attitudes towards the EU and British identity (as x1 and x2). From the SPSS output, B's for attitudes towards the EU and British identity are 1.461 and -.367, and B for the constant (a) is also known: -33.395. We can substitute the figures to make a regression equation to predict attitudes towards the Euro: Ŷ = a + b1x1 + b2x2 (Predicted) attitudes towards the Euro = -33.395 + 1.461 (attitudes towards the EU) - .367 (British identity) The equation allows us to work out a predicted value for the criterion variable that we are interested in given the predictors values of an individual
For what can the unstandardised beta coefficients (B) be used in regression?
Answer: Phenomenology Phenomenology is the branch of philosophy that advocates for a detailed exam of lived or subjective experience that are important for the person/ people being studied, while hermeneutics, the philosophy of 'interpretation', advocates for a robust or dynamic way of making sense of the experience of the person/ people being studied. Thus the IPA approach, drawing from these influences, involves a detailed exam of the experience (or phenomenon or point of view) of interest by vigorously making sense, back-and-forth, of (interpreting) the person/ people experiencing it.
Hermeneutics is one of two key philosophical strands for IPA; what is the other?
Answer: Social constructionist As it is concerned with how people use language to socially 'construct' their reality (also 'relativist' and 'critical realist').
IPA is concerned with understanding the 'lived experience' of a phenomenon (as realist). Discourse analysis is concerned with how language is used to 'construct' psychological/ social reality and as such is... (complete with epistemology)
Answer: Homogeneous Homogeneous means participants should share certain significant experience, characteristics or context in which the researcher is interested. As s/he samples with this purpose in mind, this is also 'purposive', as the sample is 'purposefully homogeneous'.
If a sample (more than one) is drawn, what should the sample be like?
Answer: d - Then its combined score may be obtained by either its items' average or sum score after the relevant/ negative items have been reverse-scored
If a test contains reverse-scored items: a) Then it is not possible to obtain a combined score of that test b) Then the combined score of the test must be its items' average score and not the sum of their scores c) Then the combined score of the test must be the average or sum of its items' scores after they have all been reverse-scored d) Then its combined score may be obtained by either its items' average or sum score after the relevant/ negative items have been reverse-scored
Answer: The Greenhouse-Geisser is recommended (in fact, Dancey and Reidy even recommend using this set in any case, as there is always 'some' correlation between conditions—it is just that they are not necessarily significant'). The adjusted figures take into account any greater chance of Type 1 error (that a statistical test finding that there is a significant effect when there really is not due to inflated sampling error) by adjusting the degrees of freedom attached to the conditions when the calculation is conducted.
If there is significant sphericity between the conditions, then when reading the results of a repeated measures ANOVA, which set of figures should be taken? a) Sphericity Assumed b) Greenhouse-Geisser c) Huynh-Feldt d) Lower-bound
Answer: b - Greenhouse-Geisser The Greenhouse-Geisser is recommended (in fact, Dancey and Reidy even recommend using this set in any case as there is always some correlation between conditions—it is just that they are not necessarily significant). The adjusted figures take into account any greater chance of Type 1 Error (that a statistical test finding there is a significant effect when there really is not due to inflated sampling error) by adjusting the degrees of freedom attached to the conditions when the calculation is conducted.
If there is significant sphericity between the conditions, then when reading the results of a repeated measures ANOVA, which set of figures should be taken? a) Sphericity Assumed b) Greenhouse-Geisser c) Huynh-Feldt d) Lower-bound
Answer: b - Within-groups variance In a between-groups ANOVA, where the independent variable (factor) is a betweengroups / participants V—we are interested in whether this V has an effect on the dependent variable (the 'treatment effect'). So, we are interested in the between-groups variance; i.e., how the DV varies due to differences between the groups. The variation within-groups is a result of individual differences within each group and other random factors such as sampling error; i.e., this variance is not due to the independent variable and is referred to as the 'error' variance.
In a one-way between-groups ANOVA, the 'error' term is about: a) Between-groups variance b) Within-groups variance c) Degrees of freedom d) None of the above
Answer: Individual differences This is a problem for between-participants ANOVA as individual participants WITHIN each condition can differ widely regardless of which condition they are in—which can confound the effect of the IV itself. In within=groups/repeated measures design, they are not different participants in different conditions, but the same participants taking part in all conditions, and their scores across the conditions are compared. So, no worries about individual differences in the different conditions and thus repeated measures design and associated tests are seen as more powerful and sensitive.
In a repeated-measures ANOVA, each individual participant performs in each condition of the independent variable(s) (unlike in between-participants design where they are each assigned to just ONE condition). So, as the F ratio is still = Between-groups Variance Within-groups Variance, what is taken out of the within-groups variance?
Answer: c - Between participants
In a study with gender as the manipulated variable, the IV is: a) Within participants b) Correlational c) Between participants d) None of the above
Answer: a - Order effects
In a within-participants design with two conditions, if you do not use counterbalancing of the conditions then your study is likely to suffer from what? a) Order effects b) Effects of time of day c) Lack of participants d) All of the above
Answer: Residuals The residuals are important in evaluating how good (how fitted) an regression model is that is based on the best-fit in SPSS.
In linear regression, the differences between the predicted scores on the best fit line and the actual scores obtained and plotted are called...?
Answer: c - All persons who have legal responsibilities for potential research participants that are not legally competent or vulnerable due to age or diminished capacity to give informed consent
In psychological research, 'duly authorised representatives' of 'vulnerable' groups or individuals refer to: a) Parents, carers or guardians of minors under 16 years old being invited to be participants in research b) Health authorities and/or residential homes for those with disabilities, learning difficulties or mental health issues or other clinical groups under such care c) All persons who have legal responsibilities for potential research participants that are not legally competent or vulnerable due to age or diminished capacity to give informed consent d) Psychologists that have procured the Disclosure and Barring Service (DBR) clearance and gone through relevant research and ethics training before the research with vulnerable populations
Individual differences. This is a problem for between-participants ANOVA as individual participants WITHIN each condition can differ widely regardless of which condition they are in—which can confound the effect of the independent variable itself. Here, in within-groups/repeated measures design, they are not different participants in different conditions, but the same participants taking part in all conditions, and their scores across the conditions are compared. So, no more worries about individual differences in the different conditions, and thus repeated measures design and associated testing are usually seen as a more powerful and sensitive.
In repeated-measures ANOVA, each individual participant performs in each condition of the independent variable(s) (unlike in between-participants design where they are each assigned to just ONE condition). So, as the F ratio is still: Between-groups Variance ------------------------------------ Within-groups Variance What is taken out of the Within-groups Variance?
Judges (studio audience vs. gourmet chefs)
In the finale of the cooking contest, both the studio audience and a panel of gourmet chefs will act as judges to score the final two cooks' (cook A, cook B) signature dish, scoring out of 10 for it. The programme's producer wants to see whether the two kinds of judges will differ in their scoring of the cooks. There are two independent variables/factors here. What is the between participants variable/factor?
Answer: Cooks (or their dish) This is because both kinds of judges/ ALL judges would give a score to the cooks' dishes; i.e., each judge would have a score for Cook A and a score for Cook B
In the finale of the cooking contest, both the studio audience and a panel of gourmet chefs will act as judges to score the final two cooks' (cook A, cook B) signature dish, scoring out of 10 for it. The programme's producer wants to see whether the two kinds of judges will differ in their scoring of the cooks. What is the repeated measures/within-participants IV/factor?
A 2 by 2 (2 x 2) mixed ANOVA (as each IV/factor has 2 levels, and one is a within- and one is a between-participants factor)
In the finale of the cooking contest, both the studio audience and a panel of gourmet chefs will act as judges to score the final two cooks' (cook A, cook B) signature dish, scoring out of 10 for it. The programme's producer wants to see whether the two kinds of judges will differ in their scoring of the cooks. What kind of design do the above make the analysis? (a ? by ? )
Answer: a - Participants get tired in later conditions
In within-participants designs, order effects occur when: a) Participants get tired in later conditions b) Participants perform equally well in all conditions c) Participants have trouble obtaining their drinks at the bar d) None of the above
Transparency, which concerns how 'transparently' the researcher manages to show how his/her data were gathered and analysed towards their proposed themes/ results. If things could be measured in a standard way (in quantitative psychology), then if the procedure can be replicated, the trends too should be replicated (i.e. good reliability). This cannot be applied to qualitative due to the method of research, but also the position that the approaches take (not about generalising to populations).
Instead of having reliability (whether data patterns can be replicated) or validity (whether design tests the target constructs, and results can generalise) as criteria for 'good' quantitative research, how is the quality of qualitative research judged?
Answer: C - super-ordinate themes This is recommended the most by qualitative handbooks and/or manuals. Initial codes or comments are very descriptive (and there will be many of them), summarising the 'gist' of the transcript. At the next stage the researcher attempts to depart from these by using less descriptive 'themes', and there will be fewer of them as these might group those initial codes where they share meaning (though the researcher will still check against the transcript from time to time to make sure these 'higher'-level themes proposed can still 'capture the essence', and/or remain faithful to, what the participant had said). It is from connecting and clustering these emergent themes that the next stage ('core' analysis) may go further to the superordinate themes (usually only a few). Hence, 'emergent'...as the stage 'emerging' from initially staying close to raw data towards eventual themes that are the more over-riding and abstract 'umbrella' concepts.
It would be useful to write a summary/list for which set of items below during IPA? a) initially noted comments b) emergent/constituent themes c) super-ordinate themes d) all of them
Answer: d - What language does through productive, rhetorical and metaphorical use The idea of discourse above is tied in with various aspects of language use (which include those listed in 4) in its own right, but less about using language to reflect or create things about or from people.
On the role of language, in FDA researchers are concerned about: a) How different people differ in their language use b) What language may reflect about the speakers c) How language 'socially construct' the speakers d) What language does through productive, rhetorical and metaphorical use
The positivist stance, which upholds that 'knowledge' should be readily and directly observable and thereby measurable in making up the 'objective' facts. Therefore, for things that are not readily observable or hard to quantify (points of views, emotions, etc.)—often common interests of psychology—this branch at least tends to gauge aspects of them that Can be measured and objectified (such as behavioural measures or self-reporting scores) to infer about them.
Qualitative research is often based on a 'realist' or 'relativist epistemological position. On what epistemological position is quantitative research based?
Answer: d - Both (a) and (b) above
Quasi-experimental designs have which of the following? a) An IV and a DV b) Non-random allocation of participants to conditions c) No IV or DV d) Both (a) and (b) above
Answer: b - Are different in that tests consist a set of scales, which are necessarily closeended, whereas questionnaires may not even be close-ended
Questionnaires, scales and psychologist tests: a) Differ in terms of how many points on the point-scale they enable participants to use when responding to questions b) Are different in that tests consist a set of scales, which are necessarily closeended, whereas questionnaires may not even be close-ended c) Are very different in that all psychological tests have taken a long time to build up, develop and refine to the way that they are d) Are essentially the same thing
Answer: False The Beta value of attitudes towards the EU (1.142) is much higher than that of British identity (.345); thus attitudes towards the EU contributes more to predicting attitudes towards the Euro than British identity.
Refer to Table 7 British identity is a stronger predictor than attitudes to the EU. True or False?
Answer: True The significance level/p-values associated with both predictors are lower than .05.
Refer to Table 7 Each of the predictors independently and significantly predicts the criterion, attitudes to Euro: True / False
Answer: c - F(2,27)=3.53, p=.043 A decent write-up would include making a statement of what test was performed on what and whether it found a significant result, followed by the presentation of the figures above. For example: A one-way between-groups ANOVA was performed with between-groups IV as group and DV as time and found a significant effect of group, F(2,27)=3.53, p=.043 In the bracket must be the df (degree of freedom) of the IV (here group: 2) then (after comma) the df of error (here 27) in this order. The F and p (sig.) values can be looked up from the table itself. Here, it is also acceptable to use p<.05 (as .043 is smaller than .05)
Refer to Table 8 A one-way between-groups ANOVA has been conducted. The following SPSS output shows the ANOVA results. How should you write the ANOVA result in text (say, within a paragraph in a report)? a) F(2,27)=3.53, p=.043 b) F(1,27)=19.24.68, p<.001 c) F(2,27)=3.53, p=.043 d) F(2,29)=3.53, p=.043
Answer: a. Effect size —how much the IV/factor has an effect on the DV Answer: b. Anything from .14 it is large so .207 is a large effect. Answer: c. Expressed as a percentage figure, it is the variance in the DV accounted for / explained by the IV/factor. If in an experiment about the effect of music group (say, no music, soft music and loud music) on memory, IV is music group and DV is memory, then 21% of the 'spread'/variation in participant memory can be explained by the music group they have been assigned to. [you may turn it around and consider that it also means that nearly 80% of the variance in memory is NOT explained by IV/music (despite effect size being socalled 'large'). Consider the results take into account the 'error' variance (e.g., individual differences; there are variations in memory between people regardless of whether they listen to music/which type of music).
Refer to Table 8 In the SPSS output, under the last column 'Partial eta squared' across from 'group', the figure is .207. a. What does partial eta squared show? b. Is .207 large, medium or small? c. Expressed as a percentage (21%), what does the figure mean?
Only Group 1 and group 2 differ significantly from each other (with associated significance of p=.035). The other paired comparisons (groups 1 and 3 with p=.273; Groups 2 and 3 with p=.543) have significance level higher than .05 so these other pairs do not differ significantly from each other.
Refer to Table 9 The post hoc (Tukey) test results from the above ANOVA are here: There are 3 groups: group 1, group 2 and group 3. Which groups significantly differ from each other?
Answer: Yes The hypothesis was a 1-tailed (directional) hypothesis that 'water' group would perform better than 'no water' group. Their means suggest it, and the t-test sig. (p-value)—if converted to 1-tailed—is .04 (less than .05, so less than 5% chance due to error), thus the hypothesis is supported.
Refer to Tables 1 & 2 A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. Going by the research hypothesis, is the hypothesis supported by the results here? Yes / No? Why is / isn't the hypothesis supported?
Example: An independent t-test was conducted and found that children in the 'water' group/condition had a significantly higher maths score than children in the 'no water' group/condition, t(28)=-1.82, p=.04 (1-tailed). (normally start with a statement of how/what was tested and found, then the stats)
Refer to Tables 1 & 2 A researcher aims to study whether drinking water improves children's performance in numeracy. One group of children (N=15) are given a 200ml glass of water to consume before doing a 15-minute maths test. Another group (N=15) are not given a drink with the same test. The researcher then compares the scores from the test between these two groups. How should the t-test results be written up formally?
Answer: Yes. The Levene's Test is not significant, meaning the difference in the variances of the two conditions/groups is not significant (i.e. they have a roughly equal variance).
Refer to Tables 1 & 2 Do the data satisfy the t-test assumption of homogeneity of variance? Yes / No?
Answer: Water That group has a higher mean maths score than 'no water' group.
Refer to Tables 1 & 2 Just judging from the descriptive statistics, which group of children performed better on the maths test? (water or no water)
Answer: a) Cook C. By the descriptives. Answer: b) No C's (3) score was not higher than B's (2) in Pairwise Comparisons. Answer: c) A - F(2,16)=6.40, p=.009 As Sphericity is not violated, we can use the Sphericity Assumed figures. It is ok to use the Greenhouse-Geisser figures too (however, the degrees of freedom in the brackets of second choice are wrong...if rounded up, they should be 2,14)
Refer to Tables 10 - 14 The scores for the three cooks' dishes have come in from the judges. Below are some of the SPSS outputs for the results: a) The dish by which cook has the highest score? b) Is the score of this cook's dish significantly higher than those of both of his competitors? Yes or No? c) How should you report the ANOVA result in text? A) F(2,16)=6.40, p=.009 B) F(1,13)=6.40, p=.013 C) F(1,8)=6.40, p=.035 D) None of the above
Answer: 1 - F(2,16)=6.40, p=.009 As Sphericity is not violated, we can use the Sphericity Assumed figures. It is ok to use the Greenhouse-Geisser figures too (however, the degrees of freedom in the brackets of second choice are wrong...if rounded up, they should be 2,14).
Refer to Tables 15 to 20 How should you report the ANOVA result in text? 1) F(2,16)=6.40, p=.009 2) F(1,13)=6.40, p=.013 3) F(1,8)=6.40, p=.035 4) None of the above.
No (not higher than B from the pairwise post hoc)
Refer to Tables 15 to 20 Is the score of this cook's dish significantly higher than those of both of his competitors?
Answer: a - Whether items within a test or measure are consistent among themselves and whether the test results are consistent across testings
Reliability in a psychological testing refers to: a) Whether items within a test or measure are consistent among themselves and whether the test results are consistent across testings b) A test actually measuring the variable or construct it is supposed to measure c) The Cronbach's Alpha value among the items in a test d) The idea that a test can discriminate between different participants or groups, say, between a group drawn from the clinical population and one drawn from the general population
Answer: c - Any form of harm to which participants and the researchers are exposed
Researchers in psychology have to assess the risks involved with potential 'harm' from the research. Such risks refer to and include: a) Primarily physical harm to the participants b) Both physical and psychological harm to participants c) Any form of harm to which participants and the researchers are exposed d) Those identified by the researchers, their supervisors and their institution's research ethics committee (REC)
Answer: True. Semi-interviewing is a very common method for both camps of researchers.
Semi-structured interviewing as a method of data collection can be suitable for either IPA or FDA. True or False?
Answer: a - Participants tending to present themselves in a good light by not responding truthfully, particularly in attitudinal topics or items that are relatively sensitive
Social desirability in research refers to: a) Participants tending to present themselves in a good light by not responding truthfully, particularly in attitudinal topics or items that are relatively sensitive b) The use of a 'lie scale' to resolve participants' dishonesty c) The practice of avoiding topics or items that are too sensitive being presented in a scale to avoid participants lying d) None of the above
Answer: d - All of the above
Some common ambiguities or biasing factors in psychological research include: a) A limited range of responses b) Item statements being too long or double-barrelled c) Use of emotive language or intrusive questions d) All of the above
Answer: a - Having tests administered to different samples over time and the scores also generally give rise to a distribution approximating the normal distribution
Standardisation of testing refers to: a) Having tests administered to different samples over time and the scores also generally give rise to a distribution approximating the normal distribution b) How test scores that have conformed with the normal distribution have largely measured naturally occurring variables c) How any variable can be measured by constructing and refining tests that can yield scores that confirm with the normal distribution d) All of the above
Answer: As both concern the role of language, both tend to use method with 'texts', and most commonly semi-structured interviews. Yet while the purpose of DP means that as much detail as possible is transcribed of 'naturally occurring talk' (which includes intonation, pauses, interruptions, hesitations, etc.), FDA focuses more on the content of conversations (but both attend to interviewer and interviewee).
State one key similarity and one key difference between the above strands.