4.1b: Research methods

Non-directional

A hypothesis which states that there is a difference but not what the difference will be.

Directional

A hypothesis which states the direction of the difference or relationship.

Null

A hypothesis which states the study will find no difference between the IV and DV.

IV

The variable of the study/experiment which is changed.

DV

The variable which is measured.

Pilot

A small study carried out before the bigger study.

Confederate

A person who participates in a study, but is not measured.

Extraneous

In an experiment, a variable other than the IV that might cause unwanted changes in the DV.

Situational

Variables where environmental factors may cause issues with participants.

Standardised

A set of procedures that are the same for all participants in order to combat situational variables.

Experimenter

Variables where the presence of the experimenter may cause problems with participants, such as their gender, age, race or accent.

Double-blind

A technique where the person conducting the experiment doesn't know the aims, same as the participants.

Participant

A variable where characteristics such as age, gender, and intelligence that vary from one individual to another and may affect the results.

Matched pairs

An experimental design where subjects are first matched by a characteristic and then assigned into different groups.

Repeated

An experimental measure where the same participants are used in all the conditions in an experiment.

Demand characteristics

Cues in an experiment that tell the participant what behavior is expected.

Target population

The total group to be studied or described and from whom samples may be drawn.

Random

A method of sampling that gives each person in a group the same chance of being selected.

Systematic

A method of sampling where every nth item in the target population is selected.

Opportunity

A method of sampling where anyone who can be asked is asked.

Volunteer

A method of sampling where participants are invited to take part.

Stratified

A sampling method that is variation of random sampling; the population is divided into subgroups, stratas, and weighted based on demographic characteristics of the national population.

Snowball

A method of sampling that recruits via referrals from one participant to others.

Reliability

Ability of a test to yield very similar scores for the same individual over repeated testings.

Internal validity

Extent to which the IV effects DV; with no issues from EVs.

External validity

Extent to which we can generalize findings across populations, areas and times.

Mundane realism

The extent to which an experiment is similar to real-life situations.

Test-retest

Method of reliability checking where the test is administered to the same group twice and the findings are compared.

Split-half

Method of reliability checking where the test is split into halves, which are compared.

Inter-rater

Method of reliability checking where different measures are used (if measuring factors such as behaviour) and findings are then compared.

Primary

Data that is collected firsthand by the researcher and are specifically designed. They can be lengthy or expensive, but have total control.

Secondary

Data that is collected from already available research, so struggles to fit the needs of research. Quick and easy.

Quantitative

Data that is numerical, and can be divided into different forms.

Qualitative

Data that is word based, such as interviews and articles.

Nominal

Data is organised into named categories, such as 'yes' and 'no'. Measures frequencies, lowest level of measurement.

Ordinal

Ranked data, such as a happiness scale. Middle level of measurement.

Outlying

Anomalous data, skews the data and increases standard deviation.

Central tendency

Measures that calculate an average; mean, mode and median.

Dispersion

Measures that show how representative the mean is; range and standard deviation.

Standard deviation

Sum that quantifies the amount of dispersion of the values in a data set; shows how accurate the mean is.

Interval

Data that uses numerical measurements with equal distances between them. Highest level of measurement, can go into negatives.

Ratio

Data that uses numerical measurements with equal distances between them. Highest level of measurement, cannot go into negatives.

Correlational

Studies that aim to show a correlation.

Correlation coefficient

A scale from 1- - 1 that rates the strength of the correlation.

Case study

A non-experimental method that entails an in-depth examination of a person, institution or community using many different investigation methods.

Co-variables

Replacement for IV and DV in non-experimental studies.

Longitudinal

Studies that take place over a long period of time, often case studies.

Attrition

People dropping out a study for various reasons.

Cohort

Type of effects where generational experiences make the study un-generalisable across generations.

Self report

Techniques where a person describes their own opinions and emotions.

Questionnaire

A self-report method which uses written questions with written answers.

Interview

A method of self-reports that uses verbal questions with verbal answers.

Closed

Questionnaire questions with set answers to be selected from.

Unstructured

An interview that allows for unplanned questions, which can have a small set list or be completely random.

Validity

Whether the observed effect is genuine, whether it measures what it's supposed to and can be generalised.

Concurrent

Measure of validity where the experimental measure is compared to past measures.

Face

Measure of validity where the experimental measures are looked at to decide whether they measure what they’re supposed to.

Ecological

Measure of validity where the results are checked for whether they can be generalised across environments.

Population

Measure of validity where the results are checked for whether they can be generalised across the population.

Order

Effects where participant fatigue or boredom is caused by to repeating a study, can occur during a repeated measures design.

Counterbalancing

How to overcome order effects, using and AB or BA method, or the ABBA method.

ABBA

Type of counterbalancing where participants complete all conditions twice.

AB

Type of counterconditioning where groups are split into two, with one doing condition A first and then B, and the other doing the reverse.

Social desirability

Bias that causes participants to give a more socially acceptable answer.

Lab

Experiments that occur in a controlled environment.

Field

Experiments that occur in natural environments.

Quasi

Experiments where the IV is not manipulated for ethical or plausibility reasons.

Content

Analysis of already existing content, such as advertisements and books.

Behavioural categories

Breaks down content into quantitative data, done in content analysis, thematic analysis and observations.

Cross-sectional

Non-experimental studies that compare two or more groups of people with a difference, such as young and old people.

Debrief

An interview after a study/experiment to check a person’s welfare and allow a person to withdraw their data.

Presumptive

Consent that is assumed from asking a similar group.

Retrospective

Consent given after a study.

Content

Measure of validity where the measure is checked for how well it measures what it is supposed to.

Construct

Measure of validity where the measure is checked for whether it actually measures all parts of what it intends to test for.

Predictive

Measure of validity where the measure is checked for whether it can be used to make measurable predictions.

Observation

A type of non-experimental research where behaviour is observed.

Event

Type of observational sampling where the frequency of behaviours within one time frame is counted.

Time

Type of observational sampling where the frequency of behaviours across multiple time frames is counted.

Overt

Type of observation where the participants have given informed consent to be observed.

Covert

Type of observation where the participants are observed in secret.

Intervening

The third variable in correlation studies that may have caused the correlations.

Difference

Type of quasi experiment that studies pre-existing differences between participants, such as gender.

Natural

Type of quasi experiment that studies a naturally occurring event, such as a natural disaster.

Confounding

Type of extraneous variable that affects individual participants, such as mood. Usually found after a study

Independent

Measure where different participants are used in each condition.

Artefacts

The items investigated in content analysis.

Inter-observer

Measure of reliability that uses multiple observers.

Normal

Distribution that follows the 68-95-99.7 rule, and peaks in the centre.

68, 95, 99.7

Rule of calculating the percentage of a sample in a normal distribution curve.

Negatively

Data that is skewed towards the right with an elongated left tail. Mean is lower than median and mode.

Positively

Data that is skewed towards the left with an elongated right tail. Mean is higher than median and mode.

Thematic

Content analysis that identifies themes.

Electroencephalogram (ECG)

Brain scan where electrodes are placed on the scalp to measure the electrical activity of different parts of the brain. Not very accurate as doesn’t give specific brain areas, but the least invasive. Used for epilepsy diagnoses and sleep studies.

Computed axial tomography (CAT)

A brain scans where many x-rays are taken of the brain and combined to form 2D or 3D images. It provides high quality images and shows abnormal structures well, but requires higher amounts of radiation and only provides structural information.

Magnetic resonance imaging (MRI)

Magnetic fields are used to change brain atom alignment. Has no radiation, gives details of brain soft tissues and provides detailed 3D imaging, but is expensive, claustrophobia-inducing, extremely loud and takes a long time.

Functional MRI (fMRI)

Magnetic fields are used to change brain atom alignment. Has no radiation, gives details of brain soft tissues and provides detailed 3D imaging alongside images of brain functioning in different areas, but is expensive, claustrophobia-inducing, extremely loud and takes a long time.

Positron emission tomography (PET)

Slightly radioactive glucose is injected into a patient, and goes to the areas of the brain which are the most active. Radiation detectors are used, which take 10-40 minutes. Shows chemical information, such as whether tumours are malignant or benign, and shows the brain in action. However, it is costly, can only occur a few times due to radiation and is less precise than MRI.

Chi-squared

Inferential statistic used when the data is nominal and uses independent measures. The calculated values need to be equal or above the critical value.

Sign test

Inferential statistic used when the data is nominal and uses related measures. The calculated values need to be equal or below the critical value.