Chapter 1: Science Skills and Research Methods Flashcards
Aim
- A statement outlining the purpose of the investigation
- E.g. the aim of this investigation is to compare differences in the amount of sleep obtained by adolescents and elderly people
Hypothesis
- Testable prediction of relationship between variables
- Must include:
- IV and DV
- Direction (e.g. strengthen, weaken, increase, decrease)
- Comparison groups (generally the control group/s)
NOTE: The IV and DV do not need to be operationalised.
Theories and models
Used interchangeably
- Interrelated concepts that attempt to explain observations and make predictions about future events
Variables
- A variable is something that can change
- Independent – changed / manipulated
- Dependent – influenced by IV and is measured
- Controlled – variables other than the IV that are kept constant to ensure that changes in the DV are solely due to changes in the IV
Extraneous vs confounding variables
- Extraneous – variable other than the IV that are not controlled and can have unwanted effects on the DV
-
Confounding – variable other than the IV that has directly and systematically affected the DV
- Can only be identified at the end of an experiment, as they
must have consistently and predictably affected the results
- Can only be identified at the end of an experiment, as they
Participant-related variables
Individual participant differences / subject variables
- Characteristics of a study’s participants that may affect the results
- E.g. age, intelligence, and SES
- Extraneous / confounding when not a feature of the experiment
Order effects
- Tendency for the order in which participants complete experimental conditions to affect their behaviour
- Practise – better performance in later conditions due to
having done it before - Fatigue – worse performance worse in later conditions due to being tired / bored from completing a prior task
Placebo effects
- Response to inactive substances or treatments due to expectations or beliefs
Experimenter effects
Experimenter bias
- Researcher’s expectations affect results of an experiment
- If they expect / wish to see a certain result, they can hold a confirmation bias when collecting data
Situational variables
- Environmental factors that may affect the DV
- E.g. temperature, lighting, weather, time of day
Population vs sample
-
Population – entire group of interest from which the researcher draws a sample and seeks to generalise the results of their investigation
- Members typically share characteristics (e.g. students)
-
Sample – part of the population selected for research
- Always smaller than a population
- Should ideally accurately reflect entire population of interest
NOTE: The population is often referred to as the target population.
Representative vs biased sample
-
Representative – closely resembles the population from which it is drawn from
- Sample has minimal errors in representing the population
- Equal distribution of key participant characteristics
- Biased – does not adequately represent the key characteristics of its population
NOTE: Generally, the larger the sample, the more likely it is to be a representative.
Sampling techniques
-
Random – every member of the population is equally subject to being selected to be part of the sample
- Poses minimal error in representing the population
-
Stratified – sample consists of subgroups in the same proportion as they occur in the population of interest
- Used to study behaviour / mental processes that tend to vary among different subgroups within a population
-
Convenience – selecting members of the population that are easy to involve in the study
- E.g. asking acquaintances or surveying people on the street
Experimental and control groups
- Experimental – exposed to the independent variable
- Control – not exposed to the independent variable
Random allocation
- Participants are equally as likely to be in one group as the other
- E.g. coin tossing or drawing names from a jar
Between subjects (independent groups)
Experimental design
- Each participant is assigned to one group and provides only one piece of data
- Advantage – cost and time efficient
- Disadvantage – less control over participant differences
Within subjects (repeated measures)
- Each participant is assigned to all groups and provides multiple pieces of data
- Advantage – individual participant differences can be controlled
- Disadvantages – performance can be influenced by fatigue or boredom (order effects)
Mixed design
- Combines features of the between and within subjects design
- Advantage – takes advantage of the strengths of each design
- Disadvantage – limitations of both designs still exist
Correlational study
- Researchers observe and measure the relationship between variables without any active control or manipulation of them
- Used when not ethical / possible to experimentally manipulate IV
- E.g. trauma, SA, drug misuse, age, gender
Direction of correlation
- Positive – as one variable ↑, the other variable also ↑
- Negative – as one variable ↑, the other V ↓ (and vice versa)
- Zero – no relationship between the variables
Interpreting deviations
-
Higher standard deviation – more varied data
- More outliers
-
Lower standard deviation – less varied data
- Less outliers
NOTE: Standard deviation is a method that best reveals the effect of outliers.
Types of data
- Primary – collected first-hand by current researchers
- Secondary – collected earlier by different researchers
- Qualitative – non-numerical information (usually descriptions)
- Quantitative – numerical information
- Subjective – based on personal opinion
- Objective – measurable, verifiable and free from personal bias
Repeatability and reproducibility
-
Repeatability – how close measurements are to each other in identical conditions
- Observer repeats experiment and produces same results
-
Reproducibility – how close results are when the same variable is being measured under different conditions
- Other people repeat experiment and produce same results
NOTE: The term ‘reliable’ no longer appears on the SD.
Types of errors
- Random error – unpredictable (chance) variations in measurements that affect the precision of an experiment
- Systematic error – errors that are consistent and are due to fault in the method/equipment (affect the accuracy of a measurement and cannot be improved by repeating an experiment)
- Personal errors – mistakes made by the experimenter or researcher
Research methods
- Experimental methods investigate what causes an outcome
- Correlational methods measure the relationship between variables
-
Descriptive methods describe what is occuring
- Observational studies (e.g. participant observations)
- Self-reports (e.g. interviews, questionnaires)
- Case studies (e.g. specific activities, behaviours, events or problems)
Validity
- Internal – how successfully an experiment measures what it was intended to measure
- External – the extent to which research findings can be generalised to a greater population
Robust findings
- Findings that will be produced again when the data is collected from another sample
- Can be linked to repeatability and reproducibility