Research Methods

Null Hypothesis

  • a hypothesis where any observed differences or effects are due to random chance or sampling variability

Operationalise

  • to define the hypothesis in a way that makes it testable through empirical research

One tailed vs. Two tailed hypothesis

  • one tailed > specifies that direction of the expected effect of difference between groups or conditions i.e “the new drug is more effective than the placebo”
  • Two tailed > does not specify a direction for effect i.e “there is a difference in the effects of the new drug and the placebo”

Categorical variables

  • variables that represent groups and divide data in distinct, non-overlapping groups/levels

Extraneous variables

  • The secondary variables that go uncontrolled that can impact the outcome of the experiment

Confounding variables

  • specific type of extraneous variable that can distort the relationship between the IV and the DV as its an improperly controlled variable that can lead to unreliable/incorrect interpretations of data

Standardised procedure

  • the idea that directions given to participants during an experiment are exactly the same. This is the most basic form of "control" for a study.

TEACUP

  • Testable
  • Empirical
  • Application
  • Clearly defined variables
  • Unbiased
  • Predicts behaviour

Random Sampling

  • All the people in the population have an equal chance of being selected for research

Pros and cons of random sampling

  • pros - reduces sampling bias, statistical validity, allows generalisation of findings to entire population
  • Cons - randomness has potential for under representation, selected people may nit want to participate

Stratified sampling

  • The people in the population are separated into subgroups and then in the sample are made sure to be in same proportion to the population.

Pros and cons of stratified sampling

  • pros - ensures representation, controls variables effectively
  • Cons - still cannot account for representation holistically, must be combined with another method to work

Opportunity sampling

  • convenience sampling, whoever is there and is ready to take part makes up the sample

Pros and cons of opportunity sampling

  • pros - easy, relatively homogenous
  • Cons - biased, under representative, sampler bias

Self-selected sampling

  • Volunteer sampling, people who want to be part of sample sign up

Pros and cons of self selected sampling

  • pros - simple way to get a large number of participants that are usually motivated and unlikely to leave the research
  • Cons - under representative, around 95% of the population will not sign up

Purposive sampling

  • looks for people with specific traits and is usually accomplished with self-selected sampling

Pros and cons of purposive sampling

  • pros - simple way to get a large number of participants that are usually motivated and unlikely to leave the research
  • Cons - under representative, around 95% of the population will not sign up

Snowball sampling

  • pyramid scheme sampling

Pros and cons of snowball sampling

  • pros - builds trust through mutual connection
  • Cons - not representative

Sampling bias

  • when the sample is not representative of a larger group and can lead to skewed or inaccurate conclusions and undermine its ecological validity

Ecological validity

  • refers to the extent to which the findings of a research study are able to be generalized to real-life settings.

BPS ethical guidelines

  • Informed consent
  • Right to withdraw
  • Deception
  • Protection from Harm
  • Confidentiality

Quantitative vs Qualitative data

  • Quantitative data = numerical information and analyses data using statistical techniques
  • Qualitative data = non-numerical information that is descriptive in nature and analyses data by identifying themes or patterns

Quantitative research methods

  • case studies
  • Naturalistic observations
  • interviews

Qualitative research methods

  • lab experiments
  • Field experiments
  • Quasi experiments
  • Natural experiments
  • Correlation experiments

Lab experiments

  • an experiment done under highly controlled settings

Pros and cons of lab experiments

  • Pros - Controlled environment, Replicability, Causality
  • Cons - ecological validity, ethics, demands characteristics

Field experiments

  • Experiment done in a natural setting, less control over variables and cannot be easily replicated

Pros and cons of field experiments

  • Pros - high ecological validity, Reduced demand characteristics, Opportunity for unexpected discoveries
  • Cons - Lack of control, Ethical concerns

Quasi experiments

  • No IV is manipulated and participants are not randomly allocated to conditions. Instead, it is their traits that set them apart - a fish seller, a hot dog vendor and a jeweler.

Pros and cons of quasi experiments

  • pros - high ecological validity, little or no sampling bias, demand characteristics minimised
  • Cons - no control over IV (may differ from participant to participant), no control over extraneous variables

Natural experiments

  • An experiment that is the result of a naturally occurring event

Correlation experiments

  • An experiment where the principle is if one variable changes, the other one does as well. No IV is manipulated, only a relationship is identified
  • Positive correlation 📈 negative correlation 📉

Pros and cons of correlation all experiments

  • Pros - can naturally investigate variables that could be unnatural or impractical to test, clearly allows investigator to see a relationship (if any) in graphical form
  • Cons - cannot imply causation, cannot go beyond given data, influence by demand characteristics, expectancy effect, social desirability effect and screw you effect

Demand characteristics

  • participants may influence the experiment because they believe that they know what the researcher is looking for or what the researcher is trying to do

Expectancy Effect

  • participant attempts to discern the experimenter's hypotheses with the goal of "helping" the researcher

Social desirability effect

  • participant answers in a way that makes him/her look good to the researcher

Screw you effect

  • participant attempts to discern the experimenter's hypotheses, but only in order to destroy the credibility of the study

Interview types

  • Structured - planned questions
  • Semi-structure - open-ended questions + additional questions
  • Unstructured - no planned questions, only time and topic specified

\