UNIT 0: Research methods Flashcards
confirmation bias
The tendency to search for and use information that supports our preconceptions and ignore information that refutes our ideas.
Hindsight bias
A tendency to falsely report, after an event, that we correctly predicted the outcome of the event.
Overconfidence
The tendency to overestimate the accuracy of our beliefs and judgements which can hinder problem solving.
Sample
The subgroup of the population that participated in the study e.g. a sample of 20, 15 year old girls took part
Population
All of the individuals in the group to which the study applies e.g. the findings from this study is used to generalise to the population of teenagers globally.
Hypothesis
Prediction of how two or more variables are likely related
Falsifiable
The idea that a hypothesis or theory can be disproven by observation or experiment.
Peer review
Scientific experts who review a research article’s theory, originality and accuracy.
Independent Variable (IV)
The factor the researcher manipulates in a controlled experiment (the cause).
Dependent variable (DV)
The behavior or mental process that is measured in an experiment (the effect).
Operationalised
Describing the specific procedure used to determine the presence of a variable (e.g ‘happiness’ measured by a Likert scale 1-5). It is important for replication of research.
Experimental group
The conditional group where the IV is being used
Control group
The baseline or comparison group that has no IV
Random assignment/Allocation
The division of the sample into groups in a way that every individual has an equal chance of being allocated e.g. assigning participants to conditions using a random wheel generating AI tool.
Confounding variables
In a controlled experiment, confounding variables are factors that cause differences between the experimental group and the control group, other than the chosen independent variable.
For example, when exploring if alcohol (IV) affects driving ability, the researchers found a CV that not all participants had the same driving ability which caused significant differences in the DV.
Generalizable
Generalizability is the degree to which you can apply the results of a study to a broader context of most people, most context and most of the time.
Internal validity
Internal validity refers to whether the design and conduct of a study are able to support the proposal that a cause and effect relationship exists between the independent and dependent variables.
It ensures that no other variables except the independent variable caused the observed effect on the dependent variable. In essence, it is a marker that you are truly testing what you sought to test.
External validity
The extent to which the results from a study apply to all other contexts beyond the original study e.g. if it is a laboratory experiment, can the results be found in normal, everyday observations?
Demand charecteristics
Clues participants pick up on that help them discover the purpose of the study and suggest to them how researchers would like them to respond/behave
Social Desirability Bias
(a very specific type of Self-Report Bias)
-People answering in a way that they think is socially acceptable, rather than their own beliefs or experiences
-threat of SDB is strongest when individuals are being asked questions about behaviours that are commonly thought of as bad e.g. smoking, drinking in excess, taking drugs etc
Experimenter Bias
The unintentional influence of the experimenter’s expectations, beliefs, or preconceived notions on the outcome of a study or research experiment
-Experimenter bias is not a conscious act to invalidate the results, often researchers do not know they are
doing it
Random sample
Every person in the population has an equal chance of being included in the sample group
Representative Sample
A sample that accurately represents the characteristics of a larger population. Mostly assured by the use of Random sampling
Convenience Sampling
Gathering participants who are readily accessible to the researcher
Sampling Bias
When the method used to obtain a sample tends to favour one part of the population over another. In theory, it cannot happen with Random sampling.
Single Blind Procedure
Participants are not told the true aim of the experiment at the beginning of a study. Other details may also be kept from them such as which condition they are in and whether there is another condition at all. It is used to try and prevent demand characteristics.
Double blind procedure
Neither the participants nor the researcher who conduct the study, is aware of the aims or hypotheses of the investigation
Placebo
A physical or psychological treatment given to the control group that resembles the treatment given to the experimental group, but contains NO active ingredient
Placebo effect
A physical or psychological response to the belief that the treatment will have an effect - whilst the drug has no active ingredient, the mere idea of taking a treatment can sometimes elicit a response from participants
Quantitative research
Data in numerical form that is analysed statistically and can be easily converted into graphs, charts etc.
Qualitative
Words e.g. pages of written material, answers to lots of questions. A transcript from an interview, an extract in a diary or notes recorded in a counselling session would all be qualitative data
Measure of Central Tendency
A single summary score that represents a whole set of scores. Measures of Central Tendency are:
Mode, Median or Mean.
Bimodal Distribution
When there are two frequency occurring scores.
Regression toward the mean
The tendency for unusual scores to fall back to the mean over time.
Percentile Rank
Is the percentage of scores that are less than a given score
Range
A simplified measure of data variation and is calculated by the largest score minus the smallest score
Standard deviation
The best measure of variation. It is calculated by measuring how much scores deviate from the mean. The higher the standard deviation, the more spread out (distributed) the scores are from the mean.
Distributions
Normal Curve
Positively Skewed
Negatively Skewed
Correlations
Positive Correlation
Negative Correlation
Correlation Coefficient (Effect Size)
Scatterplot
Third Variable Problem
Directionality Problem
Statistical Significance
The likelihood that the difference found is being caused by chance. The threshold for this is 0.05 - anything less than this shows that the results are caused by the IV and not by chance.
Ethics
Institutional Review Boards
Right to withdraw
Deception
Protection of participants from harm
Informed Consent
Informed Assent
Confidentiality
Anonymity
Debrief
