Research methods Flashcards
Deductive scientific cycle
Owls Take Hugs Then Read Meaningful Stories
- Observation of a behaviour
- Formulation of theory
- Creation of testable hypothesis
- Test and refine hypothesis with a study
- Modify and repeat (if necessary)
- Support/ reject a theory
External validity examples
- Population
- Ecological
- Temporal
What makes research scientific?
- Controlled (extraneous) variables
- High external/ internal validity
- High test reliability(consistency of the results)
Theory is
Collection of general principles that explain observations- help understand a phenomena
Inductive scientific cycle
Otthers Playfuly Hide Tiny Hats
1.Observation
2. Pattern
3. Hypothesis
4. Test and refine
5. Formulate hypothesis
Features of science: Empirical method
- Observation/testing
The only source of knowledge comes through our senses- sight etc - Opposite to rationalism(reason and logic)
- All knowledge is based on/ may come from personal experience
Features of science: Objectivity
- Measurement must not be affected by the researcher
- Researchers should try to remain completely unbiased
Features of science:
Replicability
- The results of a study should be consistent if all variables are kept constant by the researcher
2.A study should be written with enough detail to replicate (standard timings etc)
Features of science:
Fallibility
- Possibility that hypothesis could be false- via testing
- Falsifiable- possible to conceive an observation or an argument which negates the statement
- Karl Popper
Independent Variable is
Variable the experimenter manipulates or changes, it is assumed to have a direct effect on the Dependent Variable
Dependent Variable
V. being tested and measured in an experiment, it is “dependent” on the Independent V.
Extraneous Variable
Anything (other than Independent V.) which might have effect on the Dependent V. (participant, situational, experimenter variables)
Confounding Variable
V. that aren’t controlled in an experiment- they do affect the results (may ruin them!)
Operationalised variable
Clearly defining behaviour to be recorded in a way that makes a broad construct specific and measurable- to measure the behaviour objectively
Research aim is
Stated intentions of what question is planned to be answered (E.g to investigate effect of caffeine on memory)
Directional hypothesis( one-tailed) is
Based on prior research, you can predict the outcome
E.g caffein will increase memory score
Non-directional hypothesis is (two-tailed)
No prior or conflicting research
E.g caffeine will affect memory score
Null hypothesis is
Statement of no difference between the population being studied
“The IV will not affect the DV”
Extraneous Variables examples
- Participants: age; gender; intelligence
- Situational: controlled environment (same temp, noise etc)
- Experimenter: personality; appearance+ conduct must be the same
Independent Groups Design is
- Different participants are used in each condition of the Independent V.
Example: Harlow 1959
Evaluation:
1. Avoids order effects (boredom; fatigue)
2. As participants experience only 1 condition-demand characteristics are less likely
3. Less control of the participants- weakness
Dealing with limitations: should be done by random allocation (names in a hat)
Independent group design evaluation
- Avoids order effects (boredom; fatigue)
- As participants experience only 1 condition-demand characteristics are less likely
- Less control of the participants- weakness
Dealing with limitations: should be done by random allocation (names in a hat)
Repeated Measures design
- The same participants take part in each condition of the Independent V.
Example: in a candy taste test, the researcher would want every participant to taste and rate each type of candy.
Evaluation:
1. Controlled participant variable as they do both conditions
2. Order effect is likely- weakness
3.Demand characteristics are likely as ppts are exposed to both conditions- weakness
Dealing with limitations: Counterbalancing- split into control and experiment group- make them do each condition oppositely
Repeated Measures Evaluation
Evaluation:
1. Controlled participant variable as they do both conditions
2. Order effect is likely- weakness
3.Demand characteristics are likely as ppts are exposed to both conditions- weakness
Dealing with limitations: Counterbalancing- split into control and experiment group- make them do each condition oppositely
Matched Pairs design
- Pairs of participants are matched in terms of key variables that may affect the Dependent V.( age etc)
- 1 member of each pair must be randomly assigned to the experimental group, the other to the control group
Example: Bandura, Ross n Ross (1961)- how aggression is learned in boys and girls
Evaluation:
1. Avoids order effects
2. Helps control participant variables
3. Smaller sample size if not everyone has a match- weakness
Matched Pairs Evaluation
Evaluation:
1. Avoids order effects
2. Helps control participant variables
3. Smaller sample size if not everyone has a match- weakness
Factors to consider in an experiment
- Control of the ppts variables between conditions
- Time, ease, money
- Control of other effects
- Likelihood of demand characteristics
Lab experiment is
- Controlled setting, unnatural for the ppts
- Attempt to control all variables, but the Independent V.
Strength:
High internal validity: can establish cause and effect as extraneous variables are kept constant
Weakness:
1.Lacks mundane realism (unrealistic task leads to lack of emotional connection with the activity)
2. Low ecological validity (unrealistic tasks)
3. Demand characteristics- aim of the study is easier to guess if you know you are taking part in a experiment
Field experiment is
- Natural setting
- Independent V. are still deliberately manipulated (Bickman- people wee asked to pick up litter by a confederate dressed as security, milkman, jacket n tie)
Strength:
High ecological validity- more likely to generalise in real life
Weakness:
1. Ethical issues- deception and lack of informed consent
2. Low control of extraneous variables- can lead to low internal validity
Natural experiment
- The Independent V. varies naturally as it may be unethical or impractical manipulate
- The Dependent V. may be tested in a laboratory
Example: Rutter et al(1998)- Romanian orphans
Quasi experiment
- Independent variable occurs naturally, but is only a difference between people. Not been made to vary by anyone
Example: Taylor et al- proposed that women are more likely to seek social support from friends or family, unlike men
Evaluation Natural experiments
Strength:
1. Less likelihood on demand characteristics- as participants may not know they are being studied
2. Can be used in ethically unacceptable to manipulate the Independent V. situations
3. High ecological validity
Weaknesses:
1. More expensive and time consuming
2. No control over extraneous variables that might bias the results- difficult to replicate
3. Not true experiment as you can’t randomly allocate conditions- can’t establish cause and effect
Validity is
The study measures what it is supposed to measure. The results( Dependent V) are due to the Independent V. of extraneous variables are controlled
Internal validity is
- A causal relationship an be found
- The results(DV) are due to the IV if extraneous variables are controlled.
For this to happen, need:
1. No investigator effects
2.No demand characteristics
3.Standardised instructions
4. Large, random samples
External validity is
Measures of generalisability- generalising beyond the research setting to real life/ target population/ different periods of time
Concurrent validity is
A measure of how well a particular test correlates with a previously validated measure (E.g do the results of a new IQ test correlate with Wechler’s when given to a representative sample of the population)
Ecological validity is
How well the results of a study reflect real life
Population validity is
How representative the sample of it’s target population
Face validity is
The degree to which a procedure appears effective in terms of it’s stated aims ( does it measure what it aims to?)
Temporal validity is
Whether findings from research at 1 point of time actually reflect the way behaviour would occur at different point in time
Demand characteristics are
In an experiment, participants are often unsure what to do and look for the cues as how to behave. May alter their behaviour as a consequence of cues in the research situation
To deal with Demand Characteristics:
- Single-blind design: the participant is not aware of the research aims and/ of which condition of the experiment they are receiving. This prevents the participant from seeking cues about the aim and reacting to them.
- Experimental realism: If the experimenter makes a task very engaging the participants pay more attention to the task and not the fact that they are being observed
To deal with Investigator effects:
- Double blind design: both the participant and the person conducting the experimenter are “blind” to the aims and hypothesis. Thus the person conducting the investigation is less likely to produce cues
Investigator effects are
how can they influence an experiment
- Any cues(other than the Independent V.) from an investigator that encourage certain behaviours in the ppts and might lead to fulfilment of expectations
- Indirect effects: investigator may operationalise the measurement variables in a way that desired result is more likely
Participant variables are
- Any characteristics of individuals
- Act as extraneous variables only if independent group design is used
- Repeated measures- control ppts effects
- Matched pairs- hopefully controlled
5.Need to ensure that they reflect the general population with a large, random sample
To deal with Participant variables:
- Randomly allocate ppts into independent groups and use matched pairs design
Situational variables are:
- Features of a research situation that may influence ppts behaviour and act as confounding variables
- They are only confounding if vary systematically with the Independent V.
- E.g: noise, temp, presence of others
To deal with Situational variables:
- Control the environmental factors, use random allocation method
- Test ppts beforehand- ask if there were any situational problems
- Use Double Blind design- both the ppt and the persona condo ting the experiment are “blind” to the aims of the research. The person than is less likely to provide cues
Reliability is:
Consistency of a measure. Two subtypes: external and internal
External reliability is
extent
The extent to which a measure varies from one use to another
External reliability test
- Test the same ppts using the same design and controls after a waiting time.
E.g: give questionnaire twice after a given time period - Correlations of results should exceed +0.8
Test-retest (external) reliability, how to improve:
- Control of variables:
= Participant (use the same ppts); =Experimenter(standardised instructions and same experimenter); =Situational(same location) - Consider timings between tests: if duration is too brief then ppts may recall information from the first test.
If too long- ppts could have changed in some important way which could lead to bias in the results
Inter- observer (external) reliability is
- The extent to which 2 or more observer s are observing and recording data consistently
- A concordance rate of +0.8 or above between the observers would indicate high inter-observer reliability
- If only 1 observer- should film the behaviour and correlate their ratings recorded at the time when they carried out a second observation of the recording for the recording later
Inter- observer (external) reliability, how to improve:
- Categorise behaviour categories appropriately
- Categories should not overlap
- Train the observers
- Film the behaviours to rate later
Internal reliability is:
Wether a test is consistent within itself
E.g all the questions on an IQ inventory should be measuring IQ
Split- half method (Internal reliability test)
- Data collected is split randomly in half and compared, to see if results taken from each part of the measure are similar
E.g odd and even responses on an IQ test are analysed and correlated
Split- half method (Internal reliability test), how to improve:
size of the group
- Can only be effective with large questionnaires in which all questions measure the same construct
- This means it would not be appropriate for tests to measure different constructs
Sampling is
1.Process of selecting a representative group from the population under study
2. Group of people who take part in investigation
Target population is
Total group of individuals fro which the sample might be drawn
Opportunity sample, how it’s done
- Selecting individuals who are available and willing to partake
E.g university students
Opportunity sample, pros and cons
Strength:
1. Easiest- uses the best suitable participants
2. Takes less time to locate the sample
Weakness:
1. Biased- sample is from a small part of population
Random sample (Lottery method), how it’s done:
Every member of the population has an equal chance of being selected
E.g by placing names in a hat and drawing the required sample size
Random sample (Lottery method), pros n cons
Strength:
1. Sample should represent the target population and eliminate sampling bias
2. Unbiased selection due to the hat method
Weakness:
1 Need a list of all members of the population and contact all of the selected
2. Time consuming
Stratified sample, how it’s done:
- Work out how many peopler in target population
- Work out what % of these are in subgroup/ strata
- Calculate the % of the subgroup that should be in the sample
- State how you would choose these ppts randomly from the population of your sample (names in the hat)
Stratified sample, pros n cons:
Strength:
1. Ensures the key characteristics of the population are represented in the same proportion in the sample
2. Proportional and randomly selected representation of subgroups
Weakness:
1. Extremely time consuming
2. Difficult to do
3. Rarely used in Psychology
Systematic sample, how it’s done:
A mathematical technique, when the population is divided by the sample size to dictate which participants are selected for the sample
E.g: select every 10th person in the phone book
Stratified sample, pros n cons
Strength:
1. Unbiased- participants are selected using an objective system
Weaknesses:
1. Not truly unbiased/ random, unless a number is selected using a random method and start with this person, then every nth person
Volunteer sample, how it’s done:
Self- selected sampling- people volunteer to participate
E.g: by responding to an advert in a newspaper
