- experimental research
2. non-experimental research
- manipulating an variable and testing it in an experiment
2. no manipulation of independent variable –> observe behavior i its natural manner (= representive observation)
non-experimental research
- ethnograpy
- sociometry
- archival research
- case history
- content analysis
- naturalistic observations
- = observation of a target audience; researcher becomes immersed in behavioral/ social system being studied (used to study cultures)
- -> observing as non-/participant (act as a functioning member of a group/ observe as a non-member) - =measure social relations; identifying, measuring interpersonal relationships within a group
- =research based on other literature, recordings, studies
- =descriptive technique; observe, report on a specific case (individual person/ time event); no manipulation, determination of cause
- = deals with written/ spoken data
- =observing subjects in their natural environment without any attempt to control/ manipulate variables (high external validity, can’t be used to determine a causal relationship)
meta analysis
= includes amount of studies that could find statistical relevant infos and looks for similar variables
range effect
- floor effect
- ceiling effect
= variables have an upper/lower limit
- = variable reaches lowest possible value
- = variable reaches its highest possible value
types of dependent variables:
- behavioral measure
- physiological measure
- self-report measure
- implicit measure
- = record actual behavior of subject
- = monitor bodily functions
- peripheral nervous system (heart rate, breathing, skin conductance,..)
- central nervous system (brain activity with fMRI, EEG, MEG,..) - = through ratings, questionnaires (subjective measures)
- retrospective verbal report (past experience)
- prospective verbal report (future situation) - = measures unconscious responses
(most popular: implicit association test = people more quickly associate positive characteristics with favored group)
experimenter bias
= researcher already expect certain outcome of the experiment, is therefor bias
–> expectancy effect
- single-blind technique
- double-blind technique
pilot study
= test/ study before your real study/ experiment to prevent any problems that could occur
reliability
= ability to produce similar results, when measure is repeated under identical conditions
- test-repeated reliability; split-half test; parallel-form reliability
validity of measure
- internal validity
- external validity
- face validity
- content validity
- criterion-related validity
- concurrent validity
- predictive validity
- construct validity
= extent to which it measures what you intend it to
- =extend to which experiment is free from errors
- =need to be able to get generalized to whole population)
- =how well measurement instrument measures what it was designed to
- =how adequate content of test samples the knowledge, skills, behavior that it is intended to
- =how adequately test score can be used to infer an individual’s value on some criterion measure
- =scores on test made ate the same time
- =comparing scores on test collected with value of criterion measure observed at a later time (really a delay between two tests)
- =test is designed to measure construct, which is a variable, not directly observable, that has been developed to explain behavior on the basis of some theory (can be reliable but inaccurate);(has some more theories behind it)
(criterion: has to be something that you really can measure)
