Basics Overview Flashcards
Different Types of Research Design
There are two basic types of design.
The design you use depends upon your research question.
Are you looking for a relationship between or amongst variables?
relational design
Are you looking for a difference between groups or conditions?
experimental design
( quasi-experimental)
Relational Design Examples
Is there a relationship between height and age in children?
This is an example of CORRELATION.
Is there a relationship between height and age in children?
This is an example of CORRELATION
Does age predict height in children?
This is an example of REGRESSION.
Experimental Designs 1
BETWEEN-SUBJECTS design:
Differences between groups
Different people in each group
Each person takes part in only one condition and contribute only one score
aka Independent Groups design
Example of Between Subjects Design
Are older adults’ reaction times longer than younger adults?
Is there a difference between older and younger age groups in reaction times?
Quasi-experimental: random allocation to groups isn’t possible
Big issue is that participants may vary in other ways than the IV: individual differences
Experimental Designs 2
WITHIN-PARTICIPANTS / WITHIN-SUBJECTS design
Differences between conditions
The same people in each condition
Each person takes part all conditions and contributes a score in each condition
Issues with order effects, but easily controlled by counterbalancing
Controls for individual differences: each participant is their own control
Example of Within Subjects Design
Are Alpha wave amplitudes in individuals greater with eyes open or eyes closed?
Is there a difference in alpha amplitudes when people’s eyes are open or closed?
This is a within-subjects design
Different Types of Variables
Independent Variables (IV) They can be changed/manipulated by the experimenter (e.g. learning methods, dual or single tasks, congruent or incongruent conditions)
Dependent Variables (DV) They will be affected by the manipulation (changed by the influence of the manipulated independent variable (e.g. exam results, response accuracy, reaction times))
Confounding Variables
The IV is manipulated (cause).
The DV is measured (effect).
Confounding variable: anything that could influence your outcome/DV (e.g. time of day, noise level, previous knowledge) that isn’t your independent variable
Levels of Measurement
There are four basic ones (NOIR) NOMINAL ORDINAL INTERVAL RATIO
Nominal Data
Not numerically related
Categories/categorial
Mutually exclusive
Remember: Nominal = name
Ordinal Data
Not numerically related Ranked or placed in order Differences between scores don’t represent REAL differences. They are discrete. Remember: ordinal = order
Interval Data
Numerically related
Order
Differences between individual scores are equal
Don’t have a ‘true zero’ (can be negative, e.g. temperature in celcius, difference in reaction times between two conditions)
They can be discrete (whole numbers) or continuous (with decimal places).
Remember: interval = space in between
Ratio
Numerically related Order Equal differences between values Absolute zero, scores cannot be negative They can be discrete or continuous.