quasi experimental designs

why do researchers sometimes prefer to study behaviour in a more “natural setting”

bc although experiments w experimental control are high in internal validity, they can also produce artificial behaviours

what are some trade-offs to studying behaviours “in the field”

• the loss of control over extraneous variables

• researcher can’t randomly assign participants to conditions → groups in the study are non-equivalent from the start

you no longer have a true experiment when … → you have a …

• full experimental control of the IV & random assignment to conditions isn’t possible

• quasi-experiment

a quasi-experiment attempts to…

examine cause-and-effect relationships in spite of not having random assignment or proper manipulation of IV

with non-equivalent groups, there’s always the possibility that differences observed are caused by …

initial group differences

how are quasi-experiments helpful

• can provide valuable insights in spite of not being true experiments

• useful for investigating situations where it’s impossible/unethical for researchers to assign participants to groups

• i.e. natural experiments such as observations of prosocial behaviour before & after 9/11

LOF studies are used to assess … in neuroscience

necessity

natural experiments

• the experimental & control conditions are determined by a natural/historical event

• usually before vs after comparisons → but can also be between subjects

examples of natural experiments

• ecological, social & clinical research before, during & after COVID contingencies

• people suffering from specific strokes vs matched controls

interrupted time series design

• DV is measured @ several points in time (i.e. traffic accidents b4 & after the introduction of breathalyzer tests in the UK)

• interruption refers to introduction of an IV that separates the multiple measures of the DV

how is an interrupted time series design helpful

• having pre- and post-data may allow researchers to rule out the influence of confounding variables → to a certain extent (i.e. people driving less, less drinking)

• i.e. comparing social development before & after COVID restrictions

all studies can fall into one of three classifications:

• true experimental

• quasi-experimental

• non-experimental

true experimentation demands … and this is best achieved in …

• careful control of all variables

• a laboratory setting

criticisms of laboratory work (control & validity)

• narrowness of IV & DV

• inability to generalize

• artificiality

  • what’s observed in lab setting may not necessarily reflect what happens outside the lab → results in laboratory artifact → may look like true expt but is in fact an unintentional effect caused by expt setting

• bc of strict control required → may only be able to study a very narrow aspect of phenomenon in a given study

• effects discovered in lab may be even more dramatic irl

pros of lab experimentations (control & validity)

• extremely useful/effective esp. where basic research is concerned

• expt control allows researchers to rule out alt. hypotheses. that might account for observed results

• findings from highly controlled experiments can help refine theories & point to further hypotheses to be explored

• if theory is valid → it should also apply outside of lab settings → this further validates it

mundane realism

expt resembles everyday life → not always necessary in research

experimental realism

study evokes experiences/behaviours one would see in everyday life → is necessary in research

why should a good lab expt strive for experimental realism

it will reduce artificiality & demand characteristics → participant might not start thinking abt the objectives of the study

three major types of comparison studies

• cross-sectional studies

• longitudinal studies

• cross-cultural studies

pros of cross-sectional studies

• quick to do

• low risk of attrition

cons of cross-sectional studies

• cohort effects

• cannot observe how people change overtime → bc groups are made up of different members

cohort effect

confounding effect in cross-sectional studies that occurs when differences across sections are large enough that effects observed by researchers may be due to diff experiences (social, political, etc.)

cross-sectional short-term longitudinal study

researchers can study cross-section of groups @ diff intervals over short period of time

cross-cultural research

• to determine if findings that emerge can extend to other cultures

• involves comparing two or more different societies or social/ethnic subgroups on diff variables of interest

• can shed light on universality of psychological concepts → while highlighting cultural factors that moderate diff phenomena

• have to mindful of ethnocentrism

ethnocentrism

• biased view of another cultural in comparison to one’s own

• can result in false interpretations of behaviour

biological determinism & research bias

• positions biological explanations as inherently more valid than other explanations

• fails to account for social, economic, and cultural factors as intervening variables

• has historically resulted in erroneous conclusions with potentially dangerous consequences

broca’s problematic claim

• brain volume determines intelligence

• biased questions:

  • do diff races have diff brain volumes?

  • do poor ppl have smaller brains than rich ppl?

• underlying assumptions:

  • some races are smarter than others

  • economic success is solely determined by intelligence

recent cross-cultural studies have:

• moved away from ethnocentrism

• adopted stance of cultural relativity

• led to an appreciation of individualistic & collectivist dimensions as a means of framing cultural differences

cultural relativity

idea that a person’s behaviour & characteristics can only be understood through that person’s own cultural environment

laboratory artifact

may look like true expt but is in fact an unintentional effect caused by expt setting