Research Methods AS L1 - 4 (experiments, observations, interviews, correlations) Flashcards
Independent Variable and its purpose:
Variable that the researcher manipulates in order to determine its effect on the dependent variable
What is it called when the independent variable is divided into levels?
Experimental conditions eg. 10g, 20g, 30g
Control condition:
IV is not manipulated at all
Dependent variable:
Variable that is being measured
Confounding variable
Any variables, other than IV, that have affected the DV
What does a successful experiment require and what does this mean?
- Operationalised IV and DV
- Operationalisation = Defining the variables and stating how they will be measured
Laboratory experiment:
- An experiment carried out in controlled conditions, allowing high control over IV and elimination of EV
- Pps randomly allocated to a condition using bias-free method
- Conducted in artificial setting
Extraneous variable (give example):
Other variables the experimenter wants to hold constant eg. Pp’s personal experiences
Strengths and weaknesses of lab experiments:
+ Highly controlled IV
+ Cause and effect relationship can be established
+ Reliable results
- Demand characteristics including social desirability bias
- Often lack mundane realism/ecological validity
Social desirability bias:
Participants behave in a more positive light than normal
Mundane realism:
Extent to which an experiment reflects real life
Field experiment:
Experiment carried out in natural setting
Strengths and weaknesses of field experiments:
+ Greater mundane realism/ecological validity
+ Cause and effect relationship can be established
+ Less chance for demand characteristics
- Less control over EV
- Less control over sample (may not be representative)
- May be unreliable and difficult to replicate
Sample:
Whoever is taking part in experiment
Natural experiment:
- An experiment where the researcher takes advantage of a naturally occurring IV (not directly caused by the experimenter) eg. single-sex schools
- Participants already meet the conditions of experiment
Strengths and weaknesses of natural experiment:
+ High level of mundane realism/ecological validity
+ Very useful when it is impossible/unethical to manipulate IV
- Low control over EV
- Difficult to replicate/unreliable results
- Difficult to establish cause and effect relationship
Quasi experiment:
- An experiment where IV is naturally occurring and exists all the time eg. Age, gender
- Usually take place in lab
Strengths and weaknesses of Quasi experiments:
+ High level of control over IV
+ Replication is very likely
- Lack of ecological validity
- Demand characteristics may be shown
Observation:
When a researcher watches/listens to pps engaging in whatever behaviour is being studied
Non-participant observation:
When researcher does not get directly involved with the interactions of the pps or participate in their activities
Strengths and weaknesses of Non-participant observation:
+ Won’t risk missing details as they only have one sole responsibility
+ Demand characteristics may be shown as pps may notice they are being observed
- Lack a deeper understanding of observed behaviour as they are not involved
Participant Observation:
When researcher is directly involved with the interactions of the pps and participates in their activities
Strengths and weaknesses of Participant observation:
+ Psychologist has an understanding of the activities themselves
- May miss details as they have multiple things to do
- Risk of bias as objectivity may be lost
Covert observation:
- Psychologist goes undercover and does not reveal their true identity
- May give themselves a new identity
Strengths and weaknesses of Covert observation:
+ Demand characteristics will not be shown
- Ethical issues of deception
Overt observation:
- Psychologist reveals their true identity
Strengths and weaknesses of Overt observation:
+ No ethical issues as deception is avoided
+ Participants can give fully informed consent
- Demand characteristics (observer effects) may be shown
Naturalistic observation:
- Participants are observed in their natural environment
- No deliberate manipulation of IV
Strengths and weaknesses of Naturalistic observation:
+ Reduced chance of observer effect
+ High mundane realism
+ Useful when it is unethical/impractical to manipulate IV
- No control over EV
- Cause and effect relationship cannot be established
- Risk of observer bias due to lack of control
Controlled observation:
- Researcher observes pps in a controlled environment
- Allows for manipulation of IV
Strengths and weaknesses of Controlled observation:
+ Cause and effect can be determined due to high control
+ EV can be controlled
+ Likely to yield rich and detailed qualitative data
- Lack of mundane realism
- Observer effects can occur
- Risk of observer bias
Two types of sampling procedures:
- Event sampling
- Time-interval sampling
Event sampling:
- All occurrences of the types of behaviour the researcher is interested in is recorded
- Any other behaviour is ignored
Time-interval sampling:
- Observation only takes place during specific time periods
- Occurrences are only measured in this time frame
Self-report techniques:
Research methods where pp gives info about themselves w/out researcher interference
Interviews:
Researcher asks questions in face-to-face situations
3 types of interviews:
1) Structured
2) Unstructured
3) Semi-structured
Structured interview:
- Same questions asked in same order
- Provides quantitative data
- Usually has closed questions
Strengths and weaknesses of structured interview:
+ Question can be repeated to aid understanding
+ Quick
- Risk of interviewer effect
- Lack qualitative data
Unstructured interview:
- Informal in-depth conversational exchange
- Questions are not pre-planned but may consist of themes
- Provides qualitative data
- Usually open questions about opinions
Strengths and weaknesses of unstructured interview:
+ Useful for investigating sensitive/controversial topics
- Risk of social desirability bias
- Interviewers must be well-trained, which can be time-consuming and expensive
Semi-structured interview:
- Mixture of structured + unstructured techniques
- Produces both qualitative + quantitative data
- Closed + open questions
- Some questions are pre-planned, others are based on themes
Strengths and weaknesses of semi-structured interview:
+ Includes best of both methods
+ Quicker than unstructured interviews
- Risk of social desirability bias
- Answers may be irrelevant to interview topic
- Risk of interviewer effect
- Responses may be harder to analyse
Closed questions:
- Pps either answer yes/no or choose from fixed responses
- Quantitative data
Open questions:
- Pp can answer in their own words
- Qualitative data
Interviewer effect:
Where an interviewer may inadvertently affect respondent’s answers due to factors such as manner and appearance
Qualitative data:
Non-numerical data that uses words to give a description of what ppl think and feel
Quantitative data:
Data that represents how much there is of something
Questionnaire:
- Pps are given a written set of questions and instructions about how to record their answers
- Can be self-administered, delivered by post, Internet or face-to-face
What factors should be considered when designing a questionnaire?
- Type of data required –> qualitative/quantitative affects whether open/closed questions are asked
- Ambiguity –> avoid vague questions
- Double barrelled questions –> Avoid 2 in 1 questions
- Leading questions
- Complexity –> avoid jargon
Strengths and weaknesses of questionnaires:
+ Quick, easy + cheap compared to interviews
+ Possible to have large sample
+ Easy to replicate/ Reliable results
- Questions can be ambiguous
- No researcher present so ambiguous questions cannot be clarified + pps cannot be reminded to fill in questionnaires (could be unanswered/ filled in and unreturned)
- Can have low response rate if sent by post
Correlation:
Technique for analysing strength of relationship between 2 quantitative variables (co-variables)
Positive correlation:
As one variable increases, the other variable also increases
Negative correlation:
As one variable increases, the other decreases
No correlation:
No relationship between the 2 variables
What is the range of numbers for the strength of a correlation, which refers to a strong positive, strong negative and no correlation and what is the strength of a correlation known as?
- Range: -1 to 1
- Strong positive is closer to 1
- Strong negative is closer to -1
- No correlation is zero
- Strength of correlation = Correlation coefficient
What are correlations plotted on?
Scatter graph/ scattergram
Strengths and weaknesses of correlations:
+ Strength of relationship can be established + precisely measured
+ Allow researchers to investigate things that would be unethical/impractical to manipulate experimentally
+ Predictions can be made about one of the variables based on info about the other variable
- Cause and effect unclear (correlation or causation?)
- Possible 3rd unknown variable involved rather than a correlation
- Can only measure linear relationships not curvilinear ones
Curvilinear relationships:
Positive relationship up to one point but then the relationship becomes negative
