Week 3: Experimental Designs & Controls Flashcards
Experimental research designs
Aims to establish cause & effect relationships between IV and DV via manipulation & control.
Relies on randomisation of participants into test conditions.
Quasi-experimental research design
The same as an experimental research designs by aiming to establish cause and effect relationship between IV and DV.
Key difference: participants are assigned into groups based on non-random criteria.
Non-experimental research design
Aims to establish a relationship without inferring cause & effect.
Naturalistic observation
A non-experimental research method where behaviour is observed and recorded in real-world settings.
Case Studies
And intensive study of an individual subject.
Survey research
A non-experimental research design which primarily uses surveys as its method for collecting data.
Evaluation research
A non-experimental research design which examines the effects of intervention (policy or practice)
Internal validity
how strongly can we assert that changes in our DV are down to our IV and not extraneous variables
External Validity
how generalisable are our findings - can we relate findings in sample back to population
Experimental manipulation of the IV
experimenter determines which level of the IV a participant is tested at
Individual difference manipulation
a characteristic of the participant determines the level of the IV at which they are tested
Repeated measures experimental design
each participant is tested at each level of the IV
Between groups experimental research design
each participant is only tested at one level of the IV
Mixed experimental research design
more than one IV with at least one IV manipulated between groups and at least one within groups
Factorial research design
Used to manipulate more than one IV at the same time.
Measures the interaction effects between IVs.
Split into main effects and interaction effects.
May include both repeated measures & between groups designs.
Separation
maximise the variation between groups/levels of the IV
Compression
minimise the variation within groups/levels of the IV
Self-assignment
Participants assign themselves
Experimenter assignment
Experimenter assigns participants to conditions
Arbitrary assignment
Participant selection based on seemingly non-relevant criteria.
Potential for bias & confounding variables.
Random assignment
No criteria for selection
Every member equal chance of being assigned to any group
Ensures groups are equal
Eliminates systematic differences between groups
Randomly distributes extraneous variables over groups
Matching assignment
matches participants to treatment groups on specific variables
Individual/Precision Matching assignment
pair participants who have identical scores on matching variables
Practice effect
An order effect problem where participants becomes better at task due to prior exposure.
Overcome by counterbalancing.
Fatigue effect
An order effect problem where participants may become worse as task due to prior exposure.
Overcome by counterbalancing.
Counterbalancing
Used to overcome order effects when using a repeated measures design.
Participant sample is divided in half. One half completes two conditions in order and the other half completing the conditions in reverse order.
Intra-subject counterbalancing
All participant to all conditions multiple times and in different orders.
Simple Carry Over Effect
when performance on the DV in one condition is contaminated by the effects of the previous condition
Differential carryover effects
where carryover effects of one condition of the IV differ depending on the order in which the conditions are completed
Maturation (internal development)
A repeated measure design issue.
Changes due to natural development and expected improvement over time.
History (external events)
A repeated measure design issue.
External events which affect the participants during the study.
Statistical regression
An issue with repeated measure designs.
Refers to the tendency to move up or down towards the mean over time.
E.g. someone scoring below or above par on a measure - potentially likely to move towards the mean of that variables over repeated measurements
Mortality
An issue with repeated measure designs.
Refers to participants leaving the study before it has run for its intended duration.
Measurement issues & how to control for them
Refers to problems with equipment or errors in manual recording of data.
Controlled by training researchers, multiple data recorders, or observed participant data entry.
Attribute effect and how to control for it
Refers to participants responding differently to different experimenters.
Controlled by using the same experimenters in all treatment conditions or standardising how experimenter interacts (e.g. script).
Rosenthal effect vs Golem effect
Rosenthal effect: participants expected to respond MOST favourably to study are treated differently
Golem effect: participants expected to respond LEAST favourably to study are treated differently
Controlled for by: double blind, partial blind, or automation
Demand Characteristics
Refers to how participants knowledge of the aims of study will cause them to perform in ways that conform to the study’s expectations.
Controlled for with: single-blind, double-blind, & deception
Social desirability
Refers to participant performing in ways that paint them in the best possible light, or the experimenter will find most pleasing.
controlled by: single-blind, double-blind, & deception
Hawthorn effect
Refers to participant performing better for the attention received for being in a study and not as a function of the nature of the manipulations of the IV.
Controlled by: single-blind, double-blind, & deception
single-blind
participant is not made aware of true purpose of study
double-blind
neither experimenter and participant aware of treatment condition
Deception
omission of or altering the truth of information given to the participant during a research study
Situational effects
refers to environmental effects on study e.g time of day, weather conditions
