1/36
Experiments, sampling techniques, limitations, ethics
Name | Mastery | Learn | Test | Matching | Spaced |
---|
No study sessions yet.
Comfirmation Bias
Existing belief is confirmed - for example, I won a game wearing red socks, so I must wear red socks again.
Stereotype effect
Believe something that affects out performance - for example, girls are bad at math, so they don’t try as hard
Cause and effect
Relationship between 2 variable, where 1 variable causes an effect on another variable
Hypothesis
Specific and testable statement that proposes a relationship between variables or predicts an outcome in a research study.
Null hypothesis (Ho)
Argues that IV had no effect on DV, we often want to prove this hypothesis wrong using experimental data.
Alternative hypothesis (Ha or H1)
Argues that there is a relationship between IV and DV, or a statistically significant effect between variables. Wants to prove the Null Hypothesis wrong.
Dual coding theory
Processing and recall is better if information is presented in two forms (for ex- image and words)
True experiments
Establishes cause and effect relationship
Participants randomly allocated
Investigates the relationship between IV and DV by measuring
Highly standardized procedures so that other researchers can replicate
To make study of human behavior specific
Falsifiable
Replicable for others
Empirical evidence
Experimental hypothesis
Precise and testable prediction about a predicted outcome in a study
One tailed hypothesis
A specific direction for effect of IV on DV, and effect is significantly greater or less than what is predicted by Null hypothesis (Coffee causes positive performance)
Two tailed hypothesis
Hypothesis predicitng statistical significance or relationship between two variables, but no specification of the direction of effect. (Coffee changes performance)
Confounding variables
Variables that disort relationship between IV and DV (must be controlled or it will contribute to bias)
Quasi-experiment
IV is usually something hte patient already has (gender)
IV can be manipulated
Not randomized because groups are pre-existing instead of random allocations, no random allocations
Casuality can’t be determined
Operationalization
Define variables, and detailed explanation on how IV or DV is measured in the study.
Natural experiment
Takes advantage of naturally occuring event to obersve DV (natural disasters for example)
Subjects cannot be randomly allocated to experrimental groups
Higher external validity because it reflects real world behavior
Can’t manipulate IV because its naturally existing (For example, Laws)
Target population
the group being studied
ex: all high school students worldwide
Sample population
group representing target population (likely be in experiment)
ex: kanto plain high school students
Participants
group who will actually take part in the research (smaller population taken from sample)
Sampling technique
Is used to find a sample group that can be generalized to the target population (external validity)
Population validity
The extent which findings can be generalized from the sample to the target population. It is high when appropriate sampling technique used as it represents target population.
External validity types
Population validity
Ecological validity
Ecological validity
The extent which findings can be generalized from the experiments to other settings or situations. If the procedures are more artificial and standardized (avoids bias and confounding variables), the validity is higher
Selection Bias
When participants are not representitive of the larger population being studied.
Causes of selection bias
Participants self-select/volunteer
Researcher intentionally/unintentionally chooses certain type of person
Certain groups (eg: disabled) excluded from sample because of how participants are recruited (eg: ads)
Self-selected (sampling technique)
By posters, advertisements, newspapers, online posts, public notices
Advantage: Convenient
Disadvantage: Biased, low external validity
Oppurtunity (sampling technique)
Participants chosen because they are available. For example, you ae walking by a metro and asked for an interview. You are already there at the place the research is being conducted.
Advantage: Convenient and quick
Disadvantage: Biased, low external validity
Random Samples (sampling technique)
Every individual in target population has equal chance of being selected to participate
Advantage: High external validity, controls selection bias
Disadvantage: Requires full list of population so its time consuming and not practical or possible
Field experiment
Conducted in real world/natural settings rather than controlled laboratory experiment
Cause and effect relationship
Reduced reactivity
Can manipulate IV and take place in a natural setting
Experimental research design
Repeated measures design
Independent samples design
Matched pairs design
Stratified sampling
A probability sample where population is divided into subgroups based on shared characteristics (ex: gender) then randomly selected in proportion to their presence in overall population
Advantage: High external validity
Disadvantage: Complex and time consuming
Matched pairs design
Participants are grouped based on similar characteristucs then randomly allocated.
Reduces variability
They could also be pre-tested based on a variable (memory) then allocated to conditions.
Independent measures design
2 or more seperate groups of participants where each group take part in only 1 condition of the experiment.
Repeated measures design
The same participants take part in all conditions of an experiment (compares through performance across)
Representational Generalization
Order of population (from participant to population)
Population > Target population > Accessible population > Sample > Participant
Accessible population is when it’s available to the researcher
Sample size is selected by sampling methods.
Why students are not easily generalized to wider population
Different experience
Age
Socioeconomic status
Education level
Life experience
Tend to be more verbal and social