Slides in this set

Slide 1

Preview of page 1

Volunteer (self-selected sample):
- PPs volunteer themselves after seeing an advertisement for PPs.
- Access to variety of PPs ­ more likely to get a representative sample
- Relatively cheap
- Fast to find PPs
- PPs more likely to be co-operative
- May be the only way to get a large enough sample
- Good for when you do not have the information about who the relevant groups are
- Likely to be a biased sample (volunteer bias) as there may be a certain type of person who
volunteers e.g. students
- Sample may be different to target population due to volunteering (may have keen interest in
topic/ may have less to do than others)
- May try to be seen as social acceptable/ show demand characteristics.
- Sample collected by asking individuals who are available at the time and fit the criteria you're
looking for. PPs chosen as they are convenient; may have volunteered or may be known to
- Quick- no time spent planning or using sophisticated systems for selection
- Cheap
- Non-representative sample likely, so can be hard to generalise findings
- May try to be seen socially acceptable/ show demand characteristics
- Researcher bias when selecting
- Every member of the population has an equal chance of being chosen. PPs are chosen
mathematically using chance e.g. electoral register
- Avoids bias as the researcher has no control who's chosen
- The law of probability says that the researcher will normally get a representative sample
- Time consuming- all potential PPs need to be identified before a sample can be drawn
- Small chance that a `freak' sample may be drawn- unrepresentative
Deception Protection from harm
Right to withdraw Privacy e.g. confidentiality and invasion of privacy
Informed consent…read more

Slide 2

Preview of page 2

Experiments: a situation looking at cause & effect and involves manipulation of a variable
Hypothesis: a statement that can be tested to see if it is true or not
· Research: general prediction, not enough information to base an investigation on
· Alternate: enough detail for the experiment to be carried out with components having been operationalized (make the
variables measurable)
(One-tailed): There will be a significant increase in the number of nightmares participants have after eating 100 grams of
cheese an hour before bed.
(Two-tailed): There will be a significant difference in the number of nightmares participants have after eating 100 grams of
cheese an hour before bed.
· Null: Predicts no difference e.g. There will be no significant difference of the number of nightmares suffered dependent
on the amount of cheese eaten before bed.
Variables: anything in an experiment which can come in different forms or in different values
· Extraneous: Anything other than the Independent Variable that could affect the results (Dependent Variable). As far as
possible these are controlled.
· Confounding: Extraneous variables that cannot be controlled
Confounding variables exist because: may be unethical to control them; may not be able to control them; may not be
known what is affecting the results.
· Independent: the variables you manipulate Dependent: the variables you measure
Bias (affects validity)
· Demand Characteristics: PPs try to make sense of the study they are in and adjust their behave accordingly
· Social Desirability: where a respondent gives an answer that is not necessarily true to look good in front of others.
· Experimenter Bias: the experimenters' expectations or study influences the study. The experimenter may subtly
communicate their expectations to PPs.
· Observer Bias: The presence of an observer may change the behaviour of those being observed.
Validity: Accurate or true, measures what it claims
· Internal validity: whether a study's results were really due to variables suggested by researchers were tested by their
Face validity: whether the measuring tool appears to be doing what it should
Concurrent validity: New measure test scores are correlated with those from an established test
· External validity: whether the results can be generalised if conducted in different environment or using different
Ecological validity: whether the behaviour measured in a test or method is representative of behaviour that naturally
Population validity: how well the sample can be used to extrapolate the results to the population as a whole.
Reliability: the consistency of findings; how much findings can be trusted
· Internal reliability: how consistently a method measures within itself
Inter-rater reliability: consistency between different researchers working on the same study in their findings/observations.
There should be a high positive correlation between the scores of the different researchers.
Intra-rater rater reliability: consistency of a researchers' behaviour. A researcher should produce similar test results, or
make similar observations or carry out interviews in the same way on more than 1 occasion
· External reliability: how consistently a method measures over time when repeated
Test re-test method: Participants take the same test on different occasions- a high correlation between test scores
indicates the test has good reliability.…read more

Slide 3

Preview of page 3

Laboratory experiments: An experiment is carried out in a controlled environment
Easily replicated
High levels of control over extraneous variables
Reliably establishes cause and effect
Lacks ecological validity (artificial setting)
Participants likely to know they are taking part resulting in demand characteristics
Field experiments: An experiment carried out in the natural environment
Easier to generalise findings
Participants in natural environment- high ecological validity
IV can be manipulated
Time consuming
Confounding environmental variables- less reliably establishes cause and effect
Quasi environment: An experiment carried out in a lab or in the field; the experimenter doesn't directly manipulate the IV, the IV is
naturally occurring.
Participants in natural environment- high ecological validity
More ethical as Participants aren't manipulated as much
Confounding environmental variables- less reliable cause and effect
Have to wait for IV to occur naturally or for PP with characteristics of IV to be available
Extraneous variables
PP variables: individual differences between PPs e.g. intelligence, age, gender, fitness CONTROLLED BY:
Repeated measures design: uses same PP in each condition
Removes PPs variables as each PP is tested in both conditions and fewer PP needed
Greater chance of demand characteristics as they go through procedure more than once
Order effects need to be controlled
Matched pairs design: uses different but similar PPs e.g. twins are tried to be matched on any important characteristics that might
affect their performance; 1 PP of each pair enters one condition, the other partner of the pairs enters the other condition
Extraneous variables well controlled e.g.. PP variables as characteristics are matched
Time consuming and expensive
Situational variables: outside influences on the experiment e.g. weather, time of day, noise, order effects (fatigue, boredom
and fatigue). CONTROLLED BY:
Standardisation: all instructions given to PP, procedures followed must be identical for each PP.
Counterbalancing: Researcher changes order of tasks for each PP or uses ABBA technique (half tested with alcohol first then no
alcohol, other half no alcohol first then alcohol) Controls for order effects
Randomisation: Order of tasks is decided on the toss of a coin or another method of selection (controls for order effect)
Experimenter variables: influence on PPs. CONTROLLED BY:
Double Blind: Neither individuals nor researchers know who belongs to which condition preventing researcher influence.
Other experimental designs: Independent measures: Different PP used for each condition of experiment
No order effects as each PP only participates in experiment once
Demand characteristics reduced as each PP only participates in experiment once
Individual Differences may influence results
Fatigue effect, boredom and practise effects: Carrying out a task repeatedly leads to changes in performance
Fatigue effect: deterioration of performance across conditions as PPs become tired and bored
Practise effect: improvement across conditions through familiarity of the task/ environment…read more

Slide 4

Preview of page 4

Observations: Simply observe and record natural behaviour- No IV manipulated
· Covert: PP Unaware they are being observed
· No demand characteristics
· More natural arrangement; high ecological validity
· Unethical
· Overt: PP aware they are being observed
· Ethical- informed consent
· Demand Characteristics
· Structured: Observers decide what behaviours they are looking for and devise a fixed checklist of predetermined
behaviours in which they are looking for. Once one of the behaviours have been seen, a tally is made, creating
quantitative data
· Easy to analyse and assess what's going on
· Reliable
· Researchers may miss interesting behaviours as not on checklist
· Open to researcher bias
· Unstructured: Observer simply records all the actions of the PP's, generating qualitative data
· Detailed notes on PPs behaviour
· Observer may miss things PPs do while making notes
· Difficult to compare qualitative data
· Event sampling: Observers have a predetermined fixed checklist of behaviours to look out for and observe the whole
event, checking off the behaviours as they see them.
· Catches all behaviours you are looking for as looking continuously
· Useful when behaviours recorded only happens occasionally
· Difficult to concentrate for long periods of time
· Time sampling: Observers watch for set periods of time between set intervals of time
· More focused as only watching for short periods of time
· Observations may not be representative
· May miss interesting behaviours
· Participant observation: Observer takes part in the action they're observing
· More Natural arrangement- high ecological validity
· Insight into PPs behaviour
· Difficult to take notes whilst taking part
· Researcher may become too involved
· Researcher bias- observer may influence PPs behaviour
· Non-participant observation: Observer does not take part in THE action they're observing.
· Researchers can take notes easily
· Less influence on PPs behaviour
· Observer Bias- Low ecological validity
· Less natural environment- more demand characteristics
· Naturalistic observation: behaviour studied in natural situation
· Controlled observation: Some variables controlled by researcher, possibly reducing naturalness of behaviour studied…read more

Slide 5

Preview of page 5

Self-reports: Used to gather peoples' opinion and ideas on a topic and to gain an insiders perspective.
Administered: Face to face, by post, en masse to a group in public setting, via phone or internet
Process of self report
Decide attitudes you want to measure -> qualitative or quantitative data -> open or closed questions -> pilot self-report -> modify self report -> administer
self report -> assess data
Questionnaire: asking a large sample of people for information on a specific topic at a particular moment in time
Large sample questioned quickly
Large amounts of data collected about what people think and say they do
Efficient, researcher doesn't have to be there while completed
Social desirability
Untruthful responses to personal questions
Distortion to sampling frame; only those who can read or write can take part- affects generalizability
Postal surveys= low response rate- reduces validity as less representative sample
May receive subjective interpretations to questions
Cannot be sure who completed it.
Open questions: Allows respondent to freely express their opinion and view in their own words
Rich qualitative data- high validity
Respondent free to express what they really think
Qualitative data
Low reliability
Closed questions: Respondents must choose their answer from a fixed list of predetermined answers
Quantitative data, easily statistically analysed and compared
Reliable as can be easily repeated to obtain same results
The respondents answer may not be represented in the choice- low validity
Loses richness of qualitative data
Unclear as to how respondent understood question
Structured interview: Interviewer sticks to strict list of questions and conducts the interviews in exactly same way (standardised procedure)
High intra-rater reliability
Standardised questions= high reliability
Specific info gathered
Easy to compare and analyse responses
Low validity as interviewees not free to expand on answers
Not a real insight into PP
Unstructured interview: Interviewer has freedom to vary questions and follow any line of enquiry
High validity- rich detailed responses
Can build a rapport- more likely to answer personal questions
Time consuming
May not find all info needed
Difficult to analyse answers, researcher must interpret answers
Likert scale: PPs given a list of statements about attitudes and asked to indicate on a scale how strongly they agree or disagree
Rating scale: PP asked to rate their view on scale
Extent of opinion can be shown
Quantitative data
Direction and strength of attitude measured
May misinterpret scale
Respondent cannot expand their answers or justify feelings
May not express true opinions
Social desirability…read more

Slide 6

Preview of page 6

Correlations: A relationship between 2 variables, PP provides data for both variables and
a correlation shows strength and direction of relationship
Positive correlation: as the values of one co-variable increase, the values of the other co-variable
increase (+1 coefficient)
Negative correlation: as the values of one co-variable increase, the values of the other co-variable
decrease (-1 coefficient)
No correlation: There is no relationship between the 2 co-variables (0)
Correlation coefficient measures strength and direction of a relationship
Can suggest ideas for experimental studies
Makes use of existing data, quick and easy to do
Little manipulation of behaviour- measures existing variables so high ecological validity
May be used when impractical or unethical to manipulate variables
Does not show cause and effect- possible 3rd variable influencing it
May lack internal or external validity
Can only be relied on if way of collecting data for 2 variables was accurate
Alternate 1 tailed: There will be a significant relationship between..
Alternate 2 tailed: There will be a significant positive/ negative correlation between...
Null: There will be no correlation between.... Any correlation will be due to chance
Makes use of all values- most representative measure
Unrepresentative if contains extreme values
Unaffected by extreme values
Not all values represented
Used with any data type
Not useful for small data sets
Not useful if there are various values which meet criteria
Easy to calculate
Unaffected by extreme values
Doesn't indicate how widely or tightly spread of data is
Standard deviation: calculates distance of a score from its group mean
More precise than range as all data accounted for
Allows researcher to know how much scores vary amongst themselves
Difficult to calculate
Normal distribution: most values in set of data close to mean while very few examples tend to 1
extreme or the other
Scattergraph: points plotted between 2 co-variables to see if there's a relationship between them…read more



colour coded and really clear. Everything you need, thanks 

Similar Psychology resources:

See all Psychology resources »See all resources »