Research methodologies can be broadly categorized into experimental and non-experimental methods. Non-experimental methods, often qualitative, include observation, surveys, and interviews.
Surveys are used when you want to question a large group of people without high cost in a short time.
Difficult to document
May be difficult to make ethical
Multiple participants interacting
Difficult to analyze
Social desirability bias
Gain personal information
More relaxed
Easy to analyze
Field experiments are conducted when you want to have a good ecological value. That is, when you want to make sure that people would actually react a certain way. (Otherwise, it is usually better to do a laboratory experiment) (You could also say that it is when the researcher comes to you, and not the other way around)
Natural experiments occur, they are usually conducted if possible.
Difficult to make good questions
Response bias
Often easy to compare and read information
Cost and time effective
Large domain
Interviews are like surveys, but a lot more detailed and in depth, but with fewer participants.
Case studies are a lot like natural experiments only that the participant is questioned more than made to go through experiments, and listened to rather than observed.
Low ecological value
Easy to replicate
A lot of controll
Finds cause and effect
Laboratory experiments should be used when you want to find the cause and effect relationship.
Difficult to replicate
Little control
Good ecological value
Observation is basically what you do in field experiments. These two overlap all the time and are difficult to separate.
Weaknesses
Difficult to extract useful, reliable information
No control
Researcher bias
Strenghts
No demand characteristics
Examples of studies
Loftus and Palmer's car crash study.
Milgram’s experiment on obedience
Breuer & Freud - Anna O
Harlow - Phineas Gage
Jacobsen, Alice Juel
Ethnography and Education
Grunschel, Carola; Patrzek, Justine; Fries, Stefan
European Journal of Psychology of Education