Experimental design Flashcards
The scientific method
is a systematic approach for planning, conducting and reporting research which involves collecting empirical evidence
Empirical evidence
data collected directly from observation or experiment
Experiment
a controlled research study testing the effect of changing one variable on another. An example of quantitative research that generates numerical data
Types of experiments
True experiment
Field experiment
Quasi experiment
Natural experiment
True experiment
when the IV is manipulated and DV is measured under controlled conditions. Participants are randomly allocated to conditions
Field experiment
an experiment done in a natural setting. There is less control over variables
quasi experiment
no IV is manipulated and participants cannot be randomly allocated to conditions. Instead their traits allocate them to groups e.g. age, gender
Natural experiment
an experiment that is a result of a naturally occurring event e.g. changes in behavior after a change in law or changes in behavior after the introduction of television to isolated communities
Extraneous variables
variables other than the independent variable that have the potential to cause unwanted effects on the dependent variable
confounding variable
variables other than the IV that directly and systematically had an effect on the DV (often related to the design of the experiment)
Experimental designs
Repeated measures
matched pair
Independent measures
repeated measures
- Each participant completes both conditions (the control and experimental)
Participant differences are eliminated (e.g. the same IQ etc.
Matched pair
- Participants with similar qualities that may effect the study are matched up together, you then randomly split them into the two groups so that both groups have similar characteristics
E.g. if you are looking at music’s effect on school grades you might match people with similar historic academic performances and similar subjects
Independent measures
- Participants are randomly (have an equal chance of being assigned to wither group) allocated to the control and groups
- Participants complete only one of the conditions
What are the ethical considerations
- deception
- protection from undue stress or harm
- debriefing
- informed consent
- anonymity
- right to withdrawl
ethical consideration
what is deception
- Knowing things about the experiment may change how people perform in the study
- You can use deception if its justified, it doesn’t cause harm or distress and the participant is debriefed after the experiment so that they understand what actually happened and the true purpose of the study
ethical consideration
protection from undue stress or harm
- Throughout the experiment participants are protected form all forms of physical and mental harm
- There are no negative consequences suffered from the study
- Care must be taken with vulnerable groups
ethical considerations
debriefing
- Participants are made aware of the true nature/aim of the study once it has concluded
- If deception was used you must tell them how they were deceived and the reasons why
- Researchers must allow participants to ask for a copy of the study including procedures, results and conclusion
- Should refer participants to relevant healthcare professionals if they are experiencing issues post the experiment
ethical consideration
informed consent
- Participation is voluntary and fully informed
- They must agree in writing to;
The nature and purpose of the research
Method (what they must do)
Withdrawal and confidentiality wrights
Any risks/inconveniences/discomforts they may encounter - Legal guardians must provide consent for child participants or people that cannot provide consent
- No psychologically or physically vulnerable individuals should be allowed to participate
- They must agree in writing to;
ethical consideration
anonymity
- Participants are not identified or identifiable without their consent
- Researchers must explain to participants prior to the study that the details of their involvement in a study cannot be revealed or connect to them personally
- Confidentiality is also pert of this, means that data and results are stored and disposed of in an appropriate way
ethical consideration
right to withdraw
- The participant can leave at any time
- They do not have to provide a reason to withdraw
- The must be no negative consequences around their choice to withdraw
They also have the right to remove their results form the study at any time
Extraneous variable
any variable other than the IV the could cause change in the dependent variable, and effect the results in an unwanted way. You cannot always control them.
Confounding variable
a type of extraneous variable that has not been controlled and HAS HAD an unwanted effect on the dependent variable
type of confounding/extraneous variable
participant variables
differing individual background, characteristics and abilities that may impact their responses in an experiment
E.g. age, IQ, mood, education
order effects
when performance as measured by the DV are influenced by the specific order that participants complete both conditions
type of order effect
practice effects
They have more experience from repeating the task and do better, or they may get fatigued
type of order effect
carry over effects
- The influences that a particular task has on performance in the task that follows
E.g. drink driving tests not having dropped in blood alcohol before doing the ‘not drunk’ test
demand characteristics
participant’s interpretation of the aim or purpose of an experiment causes them to make changes in their behavior (consciously or unconsciously) to fit their interpretation
experimenter effects
the researcher’s expectations, actions or biases influence the participant’s behavior in the experiment
Two types
1. Experimenter expectancy - they provide clues or hints towards the responses participants should make in the experiment
2. Experimenter bias - unintentional biases in the collection or treatment of data
non-stadardised instructions and procedures
the research procedures (including instructions) are not uniform, or the same for all participants
placebo effects
a change in behavior or response of participants due to their expectations and beliefs that they are receiving some treatment
types of sampling
random
- Every member of the population of research interest has an equal chance of being a part of the sample
types of sampling
stratified
- members of the population are divided into their sub-groups based on the characteristics they share
The participants are randomly selected from each sub group in the same proportion that exists in the population
types of sampling
convenience
- Obtain a sample of participants that are readily available without worrying about making the sample representative of the population
- Based on accessibility
types of sampling
vollunteer
- People recruiting themselves to participate
Often done through an advertisement in a newspaper
types of sampling
snowball
- Used in qualitative research
- Find a good participant, ask them to recruit others
- Used when the group is hard to access or might not respond to an advert or have a very specific characteristic
internal reliability
how consistently a method measures within itself
external reliability
how consistently a method measures over time when repeated
construct validity
- Can the test/method be used to support the underlying theory, to what extent do the operationalization reflect the construct?
Linking back to the theory
internal validity
How accurate is the methodology and procedures? Is the study actually testing the hypothesis? Is there a cause and effect relationship between variables? Is the sample a true representation of the population?
external validity
- Do the findings apply to other people, settings, situation and time periods?
ecological validity
Does the environment in which the study was carried out affect the results, making it less representative of behavior in uncontrolled environments?
population validity
Is the sample somehow biased? If so, how might this effect the results of the study
