GROOT Flashcards
(65 cards)
What is a construct?
An intangible, abstract attribute that is theorised to underlie observable behaviour. Not directly observable or measureable
What is an operational definition?
The process of defining and measuring an unobservable construct indirectly
What are 3common type of research question?
Association (is happiness related to sunshine?)
Difference (are arts students happier than science students?)
Prediction (does happiness influence academic performance?)
Where do research questions come from?
Interest, practical problems, theories, observation
Why is a literature search and review important?
Helps to inform what study has been done in area of interest, when work has been done, helps inform what next logical step will be
Qualities of a good hypothesis
Should follow on from literature review
Very specific in contrast to research question
Testable - observable and measureable qualities
Should be refutable - should be able to be proved or disproved
Should be predictive of an effect
Should not ever predict nothing will happen
Must be positive statement of effect
What is a variable?
A characteristic that can change or vary between people (could be intangible - like IQ, or tangible - height, weight, etc)
When do you use an IV and DV in an experiment?
When doing a causal experiment
Define IV
Variables that are causal or influential to focus of research - the thing you change
Define DV
Affected by IV - what you measure
What is a continuous variable?
Allow for decimal or fractional values between points on a scale to be obtained (time, distance, etc)
What is a discrete variable?
Seperate, indivisible categories, and values can’t meaningfully exist between points on a scale - no decimals or fractions (no. of people in a family)
What is a nominal scale?
Measurement based on a set of categories
Values assigned to indicate different category, and no intrinsic ordering (different values don’t = differences in magnitude, just indicate membership to different categories)
What is an ordinal scale?
Values assigned to indicate order, but does not describe magnitude of differences between points on a measurement scale (e.g. running race, 1, 2, 3, indicates order of arrival at finish line, but does not describe how far ahead each runner was of the other, in either time or distance)
What is an interval scale?
Numbers indicate an actual amount of something, e.g. temperature, has equal units of measurements seperating scores, but there is not a real 0, in that 0 does not mean complete absence of something,
What is a ratio scale?
Numbers equal an actual amount of something, with even amounts seperating scores, has a real 0, could measure distance with this scale
What is research design?
Refers to structure and format of an experiment
What are the 5 main types of research design?
Descriptive, correlational, experimental, quasi-experimental, non-experimental
What is descriptive research design?
Describing natural state of individual variables experienced by a certain group of people.
No experimentation, nothing is manipulated, focus is on observation only (height of people in class, then could describe what is typical)
What is correlational research design?
Concerned with relationships between variables, with no experimental manipulations, instead looking for relationship between 2 variables (e.g. relationship between happiness and exercise).
Cannot make claims about cause and effect or which variable influences the other.
Could also be a 3rd variable that explains relationahip between the other 2
What is experimental research design
Determining cause and effect in a relationship between variables
2 core features: manipluation of hypothesised independent variable and uses carefulyl controlled experimental conditions to increase internal validity or degree of confidence in direct relationship between 2 variables
What is random assignment of participants to different groups and how does it increase internal validity?
Randomly assigning people from sample to groups, which manipulates IV
Rules out biases that may effect group allocations
What is quasi-experimental research design?
Investiage cause and effect, and some effort made to include controls for extraneous variables, but less stringent controls possible, preventing thorough evidence of cause and effect, mainly in that don’t include random assignment of people to groups
What is non-experimental research design?
Similar to correlational designs, looking for relationships between variables, but do not attempt to alter variables or explain cause and effect.
Invoves 2 or more groups of people on 1 variable and compare groups
(are arts or science students smarter?)