Lecture 12: Quasi Experimental Flashcards
what is a quasi-experimental design?
a research design where researchers cannot assign participants to conditions/ manipulate the independent variable; instead comparisons are made between preexisting groups or within groups before and after the quasi treatment occured
what is a quasi-independent variable
the variable that participants experience for other reasons other than the researcher manipulation bc they cannot manipulate it
in quasi experiments, there are threats to the experiments _____ _______-
internal validity
the quality of a quasi experimental design depends on how many ______ to ______ ________ it successfully eliminates; designs that eliminate ________
threats to internal validity…
elimination of more threats to internal validity - preferable
Pre-test posttest design–one group
awful- fails to eliminate most threats to internal validity
what is the central threat to internal validity with a simple interuprted time series deisgn
contemporary history - develops when another event occurs at the same time as a quasi independent variable
in a simple interrupted time series design, ___________
the dependent variable is measured multiple times
one group pretest post test design study described
one group is test before and after the introduction of the quasi experimental IV
–> fails to eliminate threats to internal validity
What are threats of the internal validity in one group pretest post test designs?
extraneous factors
regression to the mean: tendency of the scores in a data set to move or regress toward the mean of the distribution with repeated testing= random fluctuations in participants scores from one measurement to the next, thus is the change due to this or the IV?
nonequivalent control group design
researcher looks for one or more groups of participants that appear to be reasonably similar to the group that received the quasi independent variable
nonequivalent groups post test only design
whats a threat that presents itself with this form of design?
two preexisting groups are studied one that experienced the quasi IV one that did not
–> threat = selection bias unequal groups
to avoid threats to internal validity it is important that researchers______ and ______
select groups that are as similar as possible in as many characteristics as possible
collect as much information as possible from the two groups so that any differences can be identified and potentially controlled for
non equivalent groups pre-test post-test design
whats a threat that may occur?
two preexisting groups are studied one that experienced the quasi IV one that did not BUT groups are tested twice once before the study once afterwards
local history effect extraneous event happens to one experimental group but not the other
time series design
quasi experimental design where participants are tested several times throughout the experiment both before and after the introduction of the IV is introduced
simple interrupted time series design
whats the rational behind it?
several pretest and post test measures
examine the effects of the quasi independent variable against the backdrop of other changes that may be occurring in the DV
what is a central threat to internal validity with simple interrupted time series design?
contemporary history
when an event occurs at the same time as the quasi experimental variable
interrupted time series design with reversal
DV is measured then the IV is introduced then the DV is measured then the IV is removed then the DV is measured again
control group interrupted time series design
perform interrupted time series analysis on the group that received the IV and on the non equivalent control group
helps rule our certain history effects
comparative time series design
examines two or more variables overtime to understand how changes in one variable lead to changes in another variable
Longitudinal designs
when are they frequently used
a single group of participants is studied over time
frequently used when development psychologists want to explore age related changes in a whole host of outcomes
age is the IV
what are the three major drawbacks of longitudinal studies?
difficult to obtain participants who agree to be studied again and again over a long period of time
often have trouble keeping track of participants many of whom invariably move, some may die
time effort and money!!! so much of it!
Longitudinal > cross sectional designs
cross sectional designs cannot distinguish between age related changes and generational effects
differences due to generation they group up in
thus longitudinal studies as they are all one generation allow researchers to examine how an individual changes with age
cross-sequential cohort designs
researchers follow two or more age groups across two or more time points
this allows researchers to tease apart age and cohort effects
can explore multiple age groups in a short period of time as compared to longitudinal designs
disadvantages to cross sequential cohort designs?
attrition
findings equivalent measures
to infer a cause and effect relation three requirements need to be met: what are they
can you infer cause and effect with quasi experimental? why or why not?
the cause has to precede the effect
the cause and effect must covary
participants are randomly assigned to groups and any confounding variables are controlled for
you cannot infer a cause and effect relationship because random assignment and control for confounding variables cannot be met
