Paper 2: Research Methods Flashcards
Q. What are the 5 sampling methods?
Stratified
Systematic
Random
Opportunity
Volunteer (Self-Selected)
Q. Give an advantage and disadvantage of Stratified sampling.
(+) It is usually representative of the target population as p’s are in proportion to their occurrence and selected randomly.
(-) It is very time consuming
Q. What is the difference between an IV and DV?
IV (Independent Variable) is the variable that is manipulated and is the potential cause that is being investigated whereas the DV (Dependent Variable) is the measured variable and is the effect of changing the IV.
Q. What are the 6 ethical issues?
Privacy
Confidentiality
Deception
Right to withdraw
Informed consent
Protection from harm
Q: Explain what is meant by a Quasi Experiment.
This is an experiment done in any setting where the participants are unable to be randomly allocated because they are automatically assigned to a condition based on a characteristic of them
Q. Explain what is meant by confounding variable
There are a variable other than the IV that affects the DV or participants behaviour. They systematically vary with the IV so affect one condition and not the other. Can be situational or participant variables.
Q. Explain what is meant by Validity.
Validity is concerned with accuracy. It includes whether the study has measured what it intended to measure (internal) and whether results can be generalised (external).
Q. Give an advantage and disadvantage of Volunteer sampling.
(+) Relatively easy as the only preparation needed is to place adverts and then participants contact them to take part.
(-) likely to be biased and not apply to target pop as only helpful people will apply.
Q. Explain what the mean, mode and median are
Mean – is the average calculated by adding all data points up and dividing by the number that there are.
Mode – most frequent score in a data set.
Median – is the middle value in a ranked (from lowest to highest) data set
Q. Explain what is meant by Reliability.
Reliability is concerned with consistency. It includes whether everything has been done consistently in the study and whether the study results are consistent over time.
Q. What are the different types of Observation?
Naturalistic and Controlled,
Structured and Unstructured,
Covert and Overt,
Participant and Non-Participant
Q. Give one advantage and one disadvantage of using the Mean.
(+) It is representative of all data scores as they are all included in the calculation.
(-) it is possible that it can be skewed by an anomaly (very high/low score) and so may not be a true reflection of the data
Q. Give an advantage and disadvantage of opportunity sampling.
(+) It is quick and easy and requires no prior preparation because it uses those available at time of testing.
(-) It is likely to be biased as certain types of people will be unavailable and so may not apply to target population
Q. Explain what is meant by a Laboratory Experiment.
A research method conducted in a controlled setting whereby the research manipulates the IV and measures the DV. EV’s are controlled.
Q. Give one advantage and one disadvantage of using the Mode.
(+) It isn’t affected by extreme values.
(-) It is unrepresentative of all data points as it focuses on only the most frequent value and ignores the others.
Q. Give an advantage and disadvantage of Random sampling.
(+) Usually representative and unbiased as every member of the target population has an equal chance of being selected.
(-) can be difficult and time consuming to collect the list of all member of target pop first.
Q. Explain what is meant by a Field Experiment.
A research method conducted in a natural setting whereby the research manipulates the IV and measures the DV. There is some control of EV’s.
Q. Give one advantage and one disadvantage of using the Median.
(+) It isn’t affected by extreme values.
(-) It is unrepresentative of all data points as it focuses on only the middle value in the data set.
Q. Give an advantage and disadvantage of Systematic sampling.
(+) Can be fairly unbiased as the researcher decides on system before seeing ps – so can’t select particular people.
(-) if participants selected by system don’t want to take part then the system fails and it may make the sample unrepresentative
Q. Explain what is meant by a Natural Experiment.
A research method whereby a researcher makes use of a naturally occurring IV (one that is already happening and is not manipulated by the researcher). There is little control of EV’s.
State what is meant by operalisation
It is turning variables into measurable Forms (e.g using an IQ test to measure intelligence). Being very specific about how to manipulate and measure the variables
Q. Explain what is meant by inter-rater reliability.
This is a check for internal reliability of researchers. It gets two or more researchers (e.g observers) to score participants using the same coding sheets and then their answers are correlated to ensure they have the same data – i.e correlation of 0.8 is needed
Q. Explain what each of the 3 experimental designs are.
Repeated Measures - same p’s do all conditions of IV.
Independent Measures/groups – Different groups of p’s do each condition of IV.
Matched Pairs – P’s are paired on certain variables and then one from each pair does one condition.
Q. Explain what the different types of interview there are.
Structured – The questions are pre-set and all p’s asked the same.
Unstructured – Questions not pre-set, they are developed from p’s previous answers.
Semi-Structured – Some pre-set questions & can develop some as well.
Q. Give any advantages and any disadvantages of independent measures/groups design.
(+) less demand characteristics as they only see one condition.
(+) No Order effects
(-) There could be individual differences (participant variables) that account for the difference in scores.
Q. Give any advantages and any disadvantages of repeated measures design.
(+) There are no participant variables because participants sit all conditions of IV.
(-) There can be order effects
(-) it can increase demand characteristics as more likely to guess what the aim is.
Q. Explain what the difference is between open and closed questions?
Open – participants are free to answer anything.
Closed – Participants are to select from limited answers.
Q. Give any advantages and any disadvantages of matched pairs design.
(+) There are less participant variables because participants are paired.
(+) less demand characteristics as they only see one condition.
(-) there could be other individual differences not accounted for in the paring.
(-) Time consuming to match up
Q. Give 3 examples of situational variables.
Time of Day
Weather
Temperature
Noise
Distractions
Q. How could you minimise/deal with participant variables?
• Repeated Measures (completely gets rid of them)
• Matched Pairs (Minimises some)
• Random allocation
Q. What is mundane realism?
How true to life the study is, particularly how reflective of real life the task given to participants.
Q. What are the two different types of order effects?
Practise – whereby people get better in the second condition as they have done it once already.
Fatigue - whereby people get bored doing 2 conditions so don’t try as hard the second time.
Q. In an observation you can do two different types of sampling to decide what behaviour to observe. What are they?
Time Sampling – recording the behaviour of participants at time intervals (e.g every 20 secs)
Event Sampling – counting the number of instances a behaviour is shown.
Q. What measures of central tendency and dispersion are used with each level of measurement?
Nominal – Mode & Range
Ordinal – Median & Range
Interval – Mean and Standard Deviation
Q. What is the range?
It is a measure of dispersion and looks at the difference between the highest data point and the lowest data point. It is mainly used with ordinal data
Q. What is a confederate?
An actor employed by the experimenter to act a participant during the experiment.
Q. What is social desirability bias?
Social desirability bias is where participants give a “desirable” answer to make themselves appear good.
Q. What are demand characteristics?
Demand characteristics are whereby a participant changes their behaviour in order to help a researcher or deliberately spoil a study.
Q. What is meant by an observation?
It is where a researcher simply watches without interfering and there is no manipulation of an IV.
Q. Explain how order effects can be minimised in a repeated measures design?
Counterbalancing
Some sit condition A first and some sit condition B first.
It balances order effects out, so effect both conditions equally
Q. How could you minimise both investigator effects/bias and demand characteristics at the same time?
Double Blind
Keeps aims hidden from researcher collecting data and participants.
Q. Explain what standardised procedures means
All participants receive the same instructions, researcher, methods etc. Used to avoid issues with internal reliability or extraneous variables.
Q. Give one advantage and one disadvantage of using a correlation.
(+) shows the strength and direction of relationship between variables.
(-) it cannot show a cause and effect relationship between the variables as they could be linked by a separate intervening variable
Q. Give one advantage and one disadvantage of quantitative data.
(+) numerical so it can be analysed with statistics and easily compared with other quantitative data to detect patterns.
(-) it isn’t detailed and cannot explain why the data collected is the way it is.
Q. What is the null hypothesis?
It predicts that the study will not find any difference/relationship or that any found was found due to chance.
Q. Explain what a correlation is.
A correlation is a research method used to investigate if there is a relationship between 2 variables. It can show positive, negative or no correlation between them.
Q. Give one advantage and one disadvantage of qualitative data.
(+) it is detailed and descriptive so gives an in-depth picture of behaviour.
(-) it is difficult to analyse, compare, test for significance or detect patterns and trends.
Q. What is a significance level?
It is the level or point at which it is accepted that results are significant as there is a low enough probability the results were obtained by chance. In Psychology this is usually lower than 5%.
Q. Explain the difference in the graphs of positive and negative skew.
Positive - shift left
Negative - shift right
Q. Explain what is meant by nominal level data?
This is data collected that is put in categories or groups. It is counting the frequency of a set of scores (e.g how many participants are tall medium and short).
Q. What are the two different ways a hypothesis can be written and what is the difference?
Directional – you state which group/condition will score higher/lower.
Non-directional – you state that a difference/relationship will be found but not exactly what way it will do.
Q. What is a correlation coefficient?
A statistical measure of the strength and direction of relationship between 2 variables. It goes from -1 (perfect negative correlation) to +1 (perfect positive correlation).
Q. Explain what is meant by Ordinal level data?
This is data that is rank ordered or non-scientific/subjective in nature.
E.g participant 1 was 5th
Or
Rating or ranking scales
Q. What is an OV and CV in stats testing?
Observed Value – the calculated value in a statistical test.
Critical Value – The value that is obtained from the stats table
Q. What are the 3 levels of measurement/data?
Nominal
Ordinal
Interval
Q. Explain what is meant by Interval level data?
This is precise scientific data where there are equal distances between data points on a scale.
E.g distance (cms), time (seconds) etc
Q. What is the significance level accepted in Psychology?
P<0.05
Q. What is the difference between a covert and overt observation?
Covert – means participants don’t know they are being studied as the researcher is hidden.
Overt - means they can see they are being researched.
Q. What is the difference between a structured and unstructured observation?
A structured observation is where there is a coding sheet and a system to record observations.
Unstructured means all behaviour is recorded without a system.
Q. What are the different measures of central tendency?
Mean
Mode
Median
Q. What is the difference between quantitative and qualitative data?
Quantitative – numerical data that expresses the quantity of a measure.
Qualitative – detailed and descriptive data.
Q. What is meant by “results were significant to p<0.01”
It means that the results probability of being due to chance is less than 1% which is lower than the standard accepted 5% significance level. Chance of a type 1 error is reduced.
Q. What are the 3 D’s that are used to determine what stats test to use?
Data – what level of data is the DV?
Design – What experimental design is the study?
Difference – is the study looking for a difference or relationship?
Q. Explain what ecological validity is.
Ecological Validity – whether the study setting reflects normal real life and so whether results can be generalised to a real life setting.
Q. Describe ways of dealing with informed consent.
- Get p’s to sign consent form before study.
- Get retrospective consent
- Get presumptive consent
Q: What is a pilot study?
It is a small scale trial study to test the procedures etc of a study with the view to make improvements before the real study.
Q. Explain what population validity is.
Population Validity – Whether the sample used is representative of the target population and so whether results can be generalised to that target population.
Q. Explain how protection from harm can be dealt with.
Do a pilot study to identify risks
Do a debrief after the study if any harm was caused during the study.
Q. Name ways of dealing with confidentiality?
Don’t record identifiable data (e.g names).
Create a participant coding system.
Q. Explain 2 advantages and 2 disadvantages of a laboratory experiment.
(+) High control of EVs = means a cause an effect relationship can be established
(+) Easy to replicate to ensure reliability.
(-) Low ecological validity
(-) Demand characteristics likely
Q. Explain what is mean by “privacy”.
Privacy is about making sure not to observe people, unless in public places or with their consent, and to respect the fact they may want to keep some information about themselves private.
Q. Describe ways of dealing with deception.
Don’t deceive unless necessary.
Do a debrief after the study.
Q. Explain 2 advantages and 2 disadvantages of a field experiment.
(+) High ecological validity
(+) Demand characteristics less likely
(-) Low control of EV’s – difficult to get a cause and effect relationship.
(-) More difficult to replicate
Q. Explain what is meant by a “lack of informed consent”?
This means that the researcher didn’t obtain appropriate consent from participants or their parents (if under 16) to take part.
Q: What is an advantage and a disadvantage of a Quasi Experiment?
(+) Less experimenter bias in allocation of p’s because they are automatically assigned.
(-) might be participant variable EV’s as participants cannot be randomly assigned.
Q. Explain 2 advantages and 2 disadvantages of a natural experiment.
(+) High ecological validity
(+) Demand characteristics less likely
(-) No real control of EV’s – difficult to get a cause and effect relationship.
(-) More difficult to replicate
Q. What is needed if you are drawing a conclusion?
Conclusion: i.e what do findings suggest
Justification: This is because the mean was higher in condition 1 than 2 (state how much higher in numbers)
Q. What could be done to minimise EV’s?
Pilot study – to identify them prior to the main study
Standardise Procedures – so no situational variables
Matched pairs/repeated measures – to avoid participant variables
Q. Explain 2 advantages and 2 disadvantages of using a questionnaire.
(+) More likely to get honest answers than an interview
(+) Quick and easy to replicate for large samples.
(-) Could have social desirability bias.
(-) questions could be misinterpreted by p’s.
Q. What is the difference between a participant and non-participant observation?
A participant observation is one where the researcher is acting like a participant and observes secretly.
A non-participant observation is where the researcher is observing and not taking part.
Q. Why might a pilot study be used? x3
- To identify EV’s
- To see what ethical issues might arise e.g what risks might need minimising
- To check the procedures work
Q. What is the name of the group of people that the study is trying to generalise to?
Target population
Q. What are the different types of extraneous and confounding variables?
Participant Variables (differences in people – only in independent measures)
Situational Variables – anything in testing environment.
When would a sign test be carried out?
• In experiment/difference studies
• With Repeated Measures Design (related data)
• With Nominal Data
Q: What is the difference between primary and secondary data?
Primary data – Data which is collected for the purpose of the current study.
Secondary Data – Data which was collected for another purpose/different study and then used in the study.
Q: What is a Meta-Analysis?
It is a large scale review study whereby a researcher analyses lots of studies data on a topic to find an overall conclusion.
Q: Explain two differences between correlations and experiments?
- Experiments examine differences between conditions of an IV whereas correlations examine relationships between covariables.
- Experiments aim to get a cause and effect relationship whereas correlations cannot establish cause and effect.
Q: What do the following symbols mean? ∝ ~
∝ directionally proportional
~ approximately equal to
Q: Explain what the following signs mean: ≥ > »_space; «_space; < ≤
≥ - more than or equal to
> - more than
» - a lot more than
«_space;- a lot less than
< - less than
≤ - less than or equal to
Q: What is meant by standard deviation?
This is a measure of dispersion. It assesses the average distance each data point is from the mean.
Q: What is an advantage and a disadvantage of using the standard deviation?
Q: What type of graph would be usually used with an experiment and for a correlation?
Experiment: bar chart
Correlation: scatter gram
Q: What is meant by random allocation and what is its purpose?
It is assigning participants at random to each condition of an IV.
It is done to reduce participant variables (EV’s).
Q: How is a sign test conducted?
- State the signs – if p’s scored the same, more or less in second condition
- Take lowest number of signs
- Compare to critical value table (reduce n if p’s scored same.
- State significance of result
Q: How to calculate a percentage?
Part of a whole (small number) divided by the whole (big number) and then times by 100.
Q: What is meant by Peer Review?
It is the assessment of a researchers work by other experts in the same field of study (i.e their peers!). Their task is to report on the quality of the research before publication in a journal.
Q: What is meant by normal distribution?
It is where the data forms a “bell curve”.
Q: What is a case study?
It is a longitudinal, in depth investigation of a person or small group involving interviewing, observing and testing them.
Q: What are behavioural categories (or coding systems) in observations?
Behavioural Categories are using coding systems to break behaviours up I.e operationalisating it. E.g aggression can be categorised into verbal, kick, hit etc
Q: In ethical issues, what is meant by a cost-benefit analysis?
It is whereby an ethics board will weigh up what the scientific benefits/gains of a piece of research will be versus what ethical costs there will be to participants.
Q: What techniques can be done to create a random sample?
The lottery method – draw names out a hat
Random Number Generators – using a computer to randomly generate numbers assigned to each participant.
Q: What is a double blind method?
It is where the participants and experimenter collecting data are unaware of the aims of the investigation.
Q: What two hypotheses are needed for every investigation?
Null & Alternative
Q: When would a non-directional hypothesis be preferred to a directional hypothesis?
When the study is examining something that there is little previous research on, as the researcher is unsure which way the results will go.
Q: What is meant by temporal validity?
This is whereby results of a study done a long time ago can be generalised to modern day behaviour. It is a form of external validity.
Q: What is a case study?
An in-depth (usually longitudinal) study that gathers a lot of detail about one person or a small group.
It involves a variety of techniques (e.g observations/interviews) and a info comes from a range of sources (the individual & family, friends)
Q: What is a content analysis?
A form of observation but instead of people, their communications (newspapers, tv ads, magazines) are studied instead.
The things are studied for patterns and trends within them.
Q: What is a content analysis?
A form of observation but instead of people, their communications (newspapers, tv ads, magazines) are studied instead.
The things are studied for patterns and trends within them
.
Q: What are the strengths and limitations of a case study?
- difficulty in generalising
- subjective & may have bias
- difficulty in replicating
+in depth data gathered
+ Complex interaction of factors can be studied at once
+ good for rare cases
Q: What are the strengths and limitations of a content analysis?
+ no social desirability or demand characteristics
+ high ecological validity
- time consuming to carry out
-potential issues with observer bias
Q: What things lower internal validity?
• EV’s and CV’s
• Investigator Bias
• Demand Characteristics
• Social Desirability Bias
Q: How could you improve the validity of a questionnaire?
❖ Including a lie scale
❖ Making them anonymous
Both increase the accuracy of answers
Q: How could you improve the validity in experiments?
❖ Use a control group to ensure results are due to manipulation
❖ Use double blind to reduce experimenter bias
Q: How could you improve the validity of observations?
❖ Use a covert design o p’s less aware of the study
❖ Use behavioural categories to reduce opportunities for observer bias
❖ Double blind design
Q: What ways can you assess/check the validity?
Concurrent Validity
Face Validity
Q: What is meant by concurrent validity?
A check that the measuring tool you are using is equal to an existing validated measuring tool.
Checked by comparing with a correlation a participant’s scores on your measure with their scores on an existing measure.
Q: What is meant by Face Validity?
This is whether a test, scale or measure appears “on the face of it” to measure the thing it is supposed to measure.
Done by examining the measuring tool closely or having the tool examined by an expert.
Q: What 3 aspects of a study could have an issue of reliability?
The Procedure
The Researcher
The Measuring Tool
Q: How is a test-retest carried out?
It is done by administering the same test to the same/similar people on different occasions. The scores on the two tests are correlated to ensure a positive 0.8 correlation or higher.
Q: What is a test-retest used for?
A test-retest assesses whether a measuring tool is reliably measuring the same thing each time it is completed by a participant. Usually used with questionnaires.
Q: How could you assess and improve reliability of a questionnaire?
❖ These can be assessed with test – retest.
❖ Anything that is unreliable will need adapting (remove questions, make them clearer)
Q: How could you improve the reliability of an interview?
❖ Reliability can be improved through having the same interviewer for each participant.
❖ Using a structured interview can also improve reliability.
Q: How could you improve the reliability of observations?
❖ Reliability of these can be ensured by making appropriate behavioural categories
❖ Have more than one observer and do an inter-observer reliability check.
Q: How could you improve the reliability of an experiment?
❖ Lab based ones have good control over the procedures to ensure reliability (e.g same testing room/researcher etc).
❖ This is more difficult in real life studies.
Q: What stats tests do you need to know?
Sign test
Chi squared
Spearman’s rho
Mann-Whitney
Wilcoxon
Pearsons
Unrelated T test
Related T test
Q: What is the difference in a type 1 and type 2 error?
Type 1 – false positive – when the experimental hypothesis may have been wrongfully accepted when null should have been.
Type 2 – false negative – when the null hypothesis was potentially wrongfully accepted when the experimental may have been true.
Q. What is a thematic analysis?
This is analysing data to identify themes (patterns) in the data.
It is examining the data to see if something keeps cropping up.
It is more descriptive than coding in quantitative content analysis.
Q: Explain when a Chi Squared test would be used.
For correlations and independent group experiments
both with nominal data
Q: Explain when the two t-tests would be used.
Unrelated: Independent groups, nominal data
Related: Repeated measures/Matched pairs, interval data
Q: Explain step by step what you do to find out if your results are significant after you have collected the data.
- Decide which statistics test is appropriate
- Calculate your observed/calculated value using the statistics test on your data
- Compare the calculated value to a table of critical values.
- Accept or reject the null & experimental hypothesis
Q: Explain when a Spearman’s Rho & Pearson’s R test would be used.
Spearman’s rho: Correlation with ordinal data
Pearson’s R: Correlation with interval data
Q: Explain what is meant by a significant and non-significant result.
Significant means there was less than a 5% chance results occurred due to chance so experimental hypothesis accepted.
Non-significant means a greater than 5% chance so null hypothesis is accepted.
Q: Explain when a Mann Whitney test would be used.
Independent group experiment with ordinal data
Q: Explain when a Wilcoxon test would be used.
Repeated measures/ matched pairs experiment with ordinal data
Q: Explain when a Sign Test would be used.
Repeated measures/Matched pairs experiment with nominal data
