Research Methods Flashcards
Aim.
General purpose of the study.
Hypothesis (H1).
Testable, predictive statement that says something will happen.
Directional hypothesis.
Very precise, tells us exactly what the researcher thinks will happen.
Non-directional hypothesis.
Predicts that there will be some effect or difference seen, but does not specify what that effect or difference will be.
Null hypothesis (H0).
Nothing will happen and anything that does will be due to chance.
Independent variable (IV).
Can be changed or manipulated.
Dependent variable (DV).
Measures the effect of the change made my the IV.
Operationalise.
Define variables in a form that can be easily measured and tested.
What do experiments look for?
Cause and effect relationship.
Extraneous variable (EV).
Any variable, not the IV, that may affect the DV if we don’t control it.
Participant variables.
Things to do with the participant that could affect the DV.
Situational variables.
Things to do with the environment that the research is carried out in that might affect the DV.
Pilot study.
Small scale trail of the investigation.
Standardised procedures.
Control situational variable.
Cofounding variable.
Not the IV, but could become a second, unintended IV for some participants.
Demand characteristics.
Cues that help the participant interpret what is happening and try to second guess the aim and how they should behave - please you effect and screw you affect.
Single blind procedure.
Information will be kept from just the participants.
Investigator effects.
Researcher unintentionally or unconsciously influences the outcome of the research.
Double-blind procedure.
Both the participants and the researcher are kept unaware of certain information.
Types of experiments.
Lab, quasi, field, natural.
Lab experiments.
Manipulating the IV in a controlled environment.
Evaluation of lab experiments
Strengths - high level of control, easy to control extraneous variables, easy to replicate.
Limitations - demand characteristics, experimenter bias, low ecological validity.
Field experiment.
IV is manipulated deliberately in a more natural setting.
Evaluation of field experiments.
Strengths - behaviour is more likely to reflect real life, less likelihood of demand characteristics.
Limitations - More expensive and time consuming, no control over EVs, low reliability.
Natural experiment.
IV is naturally occurring.
Evaluation of natural experiment
Strengths - behaviour is more likely to reflect real life, less likelihood of demand characteristics, high ecological validity.
Limitations - researcher cannot control IV, low internal validity, difficult to replicate.
Quasi experiment.
IV is based on an existing difference between people.
Evaluation of quasi experiment
Strengths - can study variables which cannot be manipulated, practical.
Limitations - harder to establish relationships, ethical issues, absence of control means other factors could have caused the effect.
Generalise.
To apply from one situation or group to another.
Internal Validity.
The extent to which the findings from the study actually measure what they claim to measure.
External validity.
The extent to which the findings from the study can be generalised outside the original context.
Reliability.
Consistency.
Internal reliability.
Was the experiment consistent in itself.
External reliability.
If the experiment was repeated are the results consistent over time.
Experimental design.
The different ways participants are allocated to the different conditions.
Independent group design.
Participants only take part in one condition, they are randomly allocated to either the control or the experimental condition.
Evaluation of independent groups design
Strengths - avoids order effects
Limitations - more people needed - time consuming and expensive. Participant variables (can be minimised with random allocation of participants into conditions)
Repeated measures.
Participants take part in both control and experimental condition.
Evaluation of repeated measures
Strengths - same participants (reduced participant variables). Fewer people needed.
Limitations - order effects (performance may be better in the first/second condition) - can try to minimise this with counterbalancing (using an alternate order)
Matched pairs.
Participants only take part in the control or the experimental condition, but before this, they are matched with another participant based on key variables.
Evaluation for matched pairs
Strengths - reduces participant variables, avoids order effects.
Limitations - very time consuming to find matching participants, impossible to match people exactly unless they are identical twins.
Random sampling.
Names chosen using a random generator.
Evaluation of random sampling and systematic sampling.
Strengths - unbiased, reduced researcher bias.
Limitations - random sample is not guaranteed, time consuming.
Systematic sampling.
Researcher randomly picks the first participant, then selects the Nth participant.
Stratified sampling.
Dividing target population into sub-categories, randomly select participants into each sub-category.
Obtain sampling frame.
Reach proportion.
Evaluation of stratified sampling
Strengths - most representative sample, reduced researcher bias as it is an objective sampling technique.
Limitations - very time consuming as you have to identify sub groups, still not fully representative (some important sub groups might be left out)
Opportunity sampling.
Selecting anyone available at that time.
Evaluation of opportunity sampling
Strengths - convenient - less costly and time consuming
Limitations - biased sample, researcher bias.
Volunteer sampling.
Participants select themselves (advertisement).
Evaluation of volunteer sampling
Strengths - convenient (people come to you), useful way to locate willing participants so people are less likely to drop out, more ethical.
Limitations - biased sample (only specific type of people are likely to volunteer), demand characteristics.
Ethical issues.
Conflict between what the researcher needs to do for the research and the rights of the participants.
Informed consent - decision whether they want to participate or not, for children under 16, parent/guardian.
Deception
Protection from harm
Privacy/confidentiality.
Informed consent.
Ensuring participants know exactly what they are getting into and they have a right to withdraw.
Other forms of consent
Presumptive consent - similar group of people asked if the study is acceptable, if they agree then the consent of the original participants is presumed.
Prior general consent - participants give their consent to studies involving deception.
Retrospective consent - participants asked for consent after the study, during the debrief.
Deception.
Deliberately misleading or withholding information from participants.
Protection from harm.
Participants should be protected from harm psychologically and physically.
Privacy.
The right participants have to control information about themselves.
Confidentiality.
The right to have any personal data protected and kept anonnymous.
Main principals of the British Psychological Society.
Respect, competence, responsibility, integrity.
What must happen before any type of research takes place.
An ethics document must be submitted to an ethics committee.
Naturalistic observation.
Watching and recording spontaneously occurring behaviour in the participants own natural environment.
Evaluation of naturalistic oberservations
Strengths - higher external validity.
Limitations - less replicable.
Controlled observations.
Watching and recording behaviour in a structured environment in which conditions are manipulated.
Evaluation of controlled observations
Strengths - easily repeated
Limitations - low external validity.
Covert observations.
Watching and recording of behaviour without the knowledge of the participant, observer is hidden.
Evaluation of covert observations
Strengths - less likely to produce demand characteristics.
Limitations - more likely to create ethical issues.
Overt observations.
Watching and recording of behaviour with the participants knowledge, observer is clearly visable.
Evaluation of overt observations
Strengths - less likely to create ethical issues.
Limitations - more likely to produce demand characteristics.
Participant observation.
Watching and recording of behaviour with the observer becoming a member of the group they are observing to get more accurate results.
Evaluation of participant observations
Strengths - increase insight and understanding of behaviour.
Limitations - decrease observer objectivity as the person is involved in the group.
Non-participant observation.
Watching and recording of behaviour with the observer remaining outside the group.
Evaluation of non-participant observations
Strengths - more likely to increase observer objectivity.
Limitations - less likely to increase insight and understanding of behaviour.
Observational design.
Decide which behaviours you will observe and which observation it will be.
Unstructured observations.
Recording every behaviour of interest.
Evaluation of unstructured observations
Strengths - good for small scale observations, good depth of detail in data collected (qualitative).
Limitations - may be too much going on in an observation to record it all.
Structured observations.
Focus on recording specific behaviours observing.
Evaluation of structured observations
Strengths - recording data is easier and more systematic (quantitative)
Limitations - lack of richness and depth of data.
Operationalised checklist.
Break the target behaviour up into behavioural categories.
Event sampling.
Counting the number times a particular behaviour occurs across the whole event.
Evaluation of event sampling
Strengths - higher validity because the entire event is observed.
Limitation - more time consuming,
Time sampling.
Counting the number of times a particular behaviour occurs in a pre-established time frame with set intervals.
Evaluation of time sampling
Strengths - less time consuming, easier to manage.
Limitations - may not record key behaviours.
Self report techniques.
Method of gathering data where participants provide information about themselves, non-experimental method.
Evaluation of self report techniques.
Strengths - can study people who are geographically distinct.
Limitations - questionnaires rely on people actually completing them and returning them, require certain level of literacy (may not be accessible to everyone).
Types of interviews
Structured - set list of questions.
Unstructured - more spontaneous.
Semi-structured
Evaluation of a structured interview
Strengths - easy to replicate because it uses a fixed set of closed questions, quick to conduct so large sample size can be gathered in small time space.
Limitations - not flexible, only closed questions are asked which produced quantitative data.
Evaluation of an unstructured interview
Strengths - more flexible and adaptable, generates qualitative data which increases understanding, increased validity (able to ask for clarification of certain points).
Limitations - can be time consuming to conduct and analyse qualitative data, employing and training interviewers is expensive.
Closed questions.
Range of possible answers that responding participants must choose from. Likert scale, rating scale, fixed choice option. Produces quantitative data.
Open questions.
No answers to choose from, questions where participants are unlikely to give the same answers, Questionnaires, interviews. Produces qualitative data.
Cyril Burt affair.
Used twins to show that intelligence is genetic. Repeated with 53 sets of twins and reported an identical correlation. Accused of inventing data, confirmed when reporter tried to find 2 of Burt’s assistants and they didn’t exist.
Peer review.
Assessment of scientific work by others who are experts in the same field to ensure papers that are published are valid and unbiased.
Identify any unscientific research before it is published.
Improves the quality - suggesting amendments.
Quantitative data.
Data gathered in numerical form, can be put into categories or in rank order.
Evaluation of quantitative data
Strengths - easy to analyse, more accurate, reliable.
Limitations - makes it difficult to understand complex behaviour (oversimplified)
Qualitative data.
Data which is descriptive, observed or reported rather than measured.
Evaluation of qualitative data
Strengths - better understanding of complex behaviour, data is rich and detailed, insight of thoughts and feelings.
Limitations - harder to create accurate results, less reliable, cannot be put into categories (more difficult to analyse).
Primary data.
Data collected or observed directly by the researcher by participants which is specifically for the purpose of the research.
Evaluation of primary data
Strengths - more accurate and reliable, private and unique data.
Limitations - expensive, time consuming.
Secondary data.
Data collected by someone other than the researcher of the study and the purpose was for something other than the aims of the study.
Evaluation of secondary data
Strengths - data is already collected (low cost/free), wide variety of sources.
Limitations - lack of accuracy, research may be out of date, no control over quality.
Mode.
Most common number.
Mean.
Add the numbers up and divide by the how many numbers there is.
Median.
Order numbers from smallest to largest and find the middle number.
Evaluation of descriptive statistics
Mode - simple to work out, unaffected by extreme scores.
Least sensitive.
Median - easier to calculate, unaffected by extremes.
Ignores most of the scores, does not work well with small sets of data.
Mean - powerful measure.
Can be affected by extremes, so results may be misleading.
Range.
Difference between the largest and the smallest number.
Standard deviation.
Tells us how spread out the scores are around the mean data.
Measures of central tendancy.
Mean, mode and median.
Measures of dispersion.
Range and standard deviation.
The sign test.
Draw conclusions about whether the result is significant or not. Data is organised into categories and then look for a difference.
S obs.
Observed value of S.
S crit.
Critical value we obtain from a sign test critical value table.
Correlations
Non-experimental method that measures the strength and direction of a relationship or link between 2 co variables. Can be positive or negative.
Types of correlation
Positive - as one co-variable increases so does the other.
Negative - as one co-variable increases the other decreases.
Zero correlation.
Curvilinear relationships (Yerkes-Dodson law of arousal)
Correlation coefficients
Numerical value which is somewhere between -1 and +1, tells us about the strength and direction of the relationship between the co-variables.
Perfect +/- 1
Strong +/- 0.9-0.7
Moderate +/- 0.6-0.4
Weak +/- 0.3-0.1
Zero
Case studies
In depth study over time of a case which is usually of an individual or small group. Often used to study a unique set of circumstances.
Can collect qualitative or quantitative data
David Reimer
Evaluation of case studies
Strength - may generate hypothesis for future studies and could lead to development of theories.
Limitations - information which makes it into to final report is based on subjective interpretation from the researcher (bias), difficult to generalise as circumstance is so unique.
Content analysis
Indirect observation of behaviour by examining an artefact such as an interview or diary and identifying themes in it.
Create a coding system which creates behavioural categories. Then go through the sample and tall each time the category appears.
Thematic analysis
Type of content analysis which produced qualitative data.
Involves identification of themes
Evaluation of content analysis
Strengths - inobtrusive - doesn’t involve the researcher interacting with the people being studied, high external validity.
Limitations - time consuming, subjective (researcher selects and records data).
Reliability
Test retest - same test on same people but different occasion with significant time between.
Results should be similar - correlation coefficient should exceed +0.8 (strong positive)
Inter-observer - 2 different people carrying out the same study. Must establish inter-rater reliability.
Correlation coefficient should exceed +0.8
Improving internal validity
Control EVs and CVs - run a pilot study.
Single blind procedure to reduce demand characteristics.
Use double blind study to reduce investigator effects.
Ensure confidentiality to prevent social desirability bias.
Ensure behavioural categories are clear to avoid issues in observations.
Improving external validity
Use stratified sampling to get a representative sample to increase population validity.
Conduct study in a natural setting (field) to increase ecological validity.
Use a natural task to ensure mundane realism.
Test-retest over time to increase temporal validity - discard outdated theories.
Check researchers interpretation in qualitative research methods is correct.
Assessing validity after a study
Face validity - subjective assessment to see id the measurement tool measures the behaviour it claims to.
However this is subjective
Content validity - objective assessment of measurement tool to see if it relates to and measures the behaviour. Asking an independent expert - no bias.
Concurrent validity - comparing results to those of research using well established measuring tool.
Features of a psychological report
Abstract - overview of study, idea of results and findings, includes short summary of aims, hypothesis, method, results and conclusion.
Introduction - looking at past research so the reader understands the reason for the study. State aims and predictions.
Method - description of what the researcher did: materials, ethics, design, sample, apparatus. Reasoning behind choices.
Results - clear summary of findings, descriptive statistics used and inferential statistics given.
Discussion - consider what the results tell you. Summary of results and relationship to previous research. Real world applications and suggestions for future research.
Referencing - list sources (author, date, title)
Features of a science - Thomas Kuhn and paradigms
Set of shared assumptions and methods.
Psychology is a pre-science as it has too much disagreement at its core.
Model revolution: some scientists produce theories and evidence to challenge an existing accepted paradigm, more and more challenge it and then a new paradigm is acceptable. This is a paradigm shift.
Kuhn cycle - pre-science -> normal science -> model drift (questioning accepted paradigms) -> model crisis (increasing questioning) -> model revolution (paradigm shift) -> paradigm change -> normal science.
Features of a science - hypothesis testing
Testing theories (set of general laws that have the ability to explain particular behaviours)
Example : Bobo doll study for SLT.
Features of a science - empirical methods
Body of research evidence for theories.
Example : limited capacity of STM (7+-2) and digit span research
Features of a science - falsifiability
Test hypotheses to falsify them, so we can rule them out as explanations, getting to the truth through the process of elimination.
Any hypothesis we cannot falsify is not scientific.
Initial dopamine hypothesis for schizophrenia (too much dopamine) was falsified as when patients were given drugs to reduce levels of dopamine not much changed.
Not everything can be falsified - id, ego and superego.
Features of a science - objectivity
Personal bias must be minimised
Examples : lab experiments, brain scans, controlled observations (strange situation), case studies.
Features of a science - replicability
Must be reliable.
Examples : digit span, research into obedience.
Features of a science - controlled experiments
Lab experiments
Examples : Asch, Bobo doll.
