Experimental Design Flashcards
Understanding indépendent and dependent variables, advantages and disadvantages of a repeated measure design, understand the advantages and disadvantages of an independent design. Understanding the nature of a matched participant design (13 cards)
EXPERIMENTAL DESIGN
- EXPERIMENTAL DESIGN METHOD ALLOWS US TO CONFIDENTLY GATHER EVIDENCE IN ORDER TO SUPPORT THEORIES:
To gather reliable and valid evidence, we need a scientific approach rather than just guessing.
Experiments help us understand cause and effect, meaning they test if one variable (X) affects another variable (Y).
It’s important to control other factors that could influence the results to ensure accuracy.
Careful planning makes experiments more reliable, leading to stronger conclusions.
TO APPLY AN EXPERIMENTAL DESIGN
- WE MUST MANIPULATE AND MEASURE CERTAIN TYPES OF VARIABLES
A variable is something that varies (changes).
In experiments, variables can have different values, like:
Height, weight, IQ
Personality, preferences (e.g., sweet vs. salty food)
Skills, education level, memory ability
We study differences between variables or how they relate to each other.
EXPERIMENTAL DESIGN: MAKING A CASUAL INFERENCE
IF WE FIND OUR MANIPULATION OF THE CASUAL VARIABLE LEADS TO A CHANGE IN THE EFFECT VARIABLE, WE CAN INTER A CAUSE-EFFECT RELATIONSHIP BETWEEN THE TWO VARIABLES
In the example “Does word length affect memory?”, we test this by:
Giving participants both long and short words
Checking how many words they remember
If they recall more short words, we can say shorter words are easier to remember. This helps us find cause and effect.
Experimental Design
Formal terms for variables in experimental design = independent (IV) and dependent (DV) variables
The causal variable that you manipulate is known as the INDEPENDENT VARIABLE (IV)
The effect variable that is measured is known as the DEPENDENT VARIABLE (DV)
So, in our example of ‘the effects of word length on recall’
The IV is word length, and there are 2 values or levels: short words, and long words (these are also known as ‘experimental conditions’ – the thing we manipulate)
The DV is the number of words recalled (the outcome measure)
OVERVIEW OF DV
the DV cannot be controlled
it is the outcome measure that is obtained from the participant’s performance in the course of the experiment
the DV is always reported in terms of the units of measurement, e.g.:
reaction time (RT) in milliseconds
number of words correctly recalled
number of errors
height in centimetres
weight in kilos
heart rate in bpm (beats per minute)
OVERVIEW OF IV
the IV is always controlled/manipulated
the values or levels of the IV are the CONDITIONS in the experiment
we can compare different levels of the IV - e.g., short vs. long words – here there are 2 EXPERIMENTAL CONDITIONS
or we can compare the presence of the IV with its absence (as in the oestrogen study on rats earlier):
The conditions where the causal variable is present (i.e., oestrogen) is known and the EXPERIMENTAL CONDITION
The condition where the causal variable is absent (i.e., the saline solution) is known as the CONTROL CONDITION
EXPERIMENTAL DESIGN: HYPOTHESIS TESTING
TYPES OF EXPERIMENTAL DESIGN
1) REPEATED- MEASURE DESIGN
2) INDEPENDENT DESIGN
3) MATHED PARTICIPANT DESIGN
REPEATED MEASURES DESIGNS
Each participant takes part in all experimental conditions
Also known as WITHIN-PARTICIPANTS design
e.g., for our experiment on word length, each participant will be presented with a list of short words AND a list of long words (both levels of the IV)
Scores on the DV would be from the same participants and RELATED to each other
INDEPENDENT DESIGN
Each participant is assigned to ONE experimental condition
Also known as BETWEEN-PARTICIPANTS design
e.g., for our experiment on word length, each participant will be presented with EITHER short words OR long words (only ONE level of the IV)
Scores on the DV would be from different participants and UNRELATED to each other
MATCHED PARTICIPANT DESIGN
Pairs of participants are matched on key variables that may affect the results
Within each pair, participants are randomly allocated to different experimental conditions
Example -> if Diet A leads to more weight loss than Diet B
Participants matched on several variables, e.g., gender, age, heath issues, previous dieting experience, socio-economic status, stress levels
Very time-consuming and can be expensive to design
