Unit 8+9 Flashcards
most single participant researchers use a research strategy popularized by skinner, what is it?
experimental analysis of behavior
single participant researchers work is often published in two highly respected journals, what are they?
the journal of the experimental analysis of behavior
the journal of applied behavior analysis
what are advantages of single participant research?
unique individuals
unique treatments
looking for large effects
freedom of design
ethical and moral dilemmas
reduces cost and time
why are unique individuals apart of single participant research?
because it would be impossible to find a population that is similar to them
how are unique individuals an advantage?
could spur future research ideas that may be generalizable to the larger population
studying the individual would provide valuable information that would, in a group design be averaged out
what is the problem with the unique individuals advantage?
there is probably no population to which you could generalize your results
how are unique treatments an advantage to single participant designs?
allows researchers to try out new drugs or treatment techniques on willing individuals when all traditional treatments have failed. the new treatments can pave way for better understanding of the affliction in question
how is looking for large effects an advantage to single participant designs?
small effects are unlikely to be found (ex. in a study with a large sample the drug might work for 85% of the participants but the placebo might work for 80% of the participants, this is considered significant and would not happen in single participant designs)
its more likely that only independent variables that have large effects on the dependent variable will surface
how do single participant designs have freedom of design?
the experimenter has freedom to modify the experiment on the fly because its only one participant (with group designs, all the participants experiences need to be altered to better fit the study… way more work)
how are ethical and moral dilemmas an advantage to single participant designs?
single participant designs avoid ethical and moral concerns involving control groups
what are some disadvantages to single participant designs?
between subjects designs are necessary
cannot draw cause and effect conclusions
bias
limited external validity
why are between subject designs necessary?
with single participant designs, the single participant cant be in two different experimental conditions which is a huge issue cause then theres no control
why is it a disadvantage that we cannot draw cause and effect conclusions with single participant designs?
cant control extraneous variables, this introduces error or excess amounts of variability into the design which makes it difficult to observe real effects in single participant designs (especially if effects are small), it also makes any observed effect open to many other plausible interpretations or explanations
how are single participant designs subject to more bias than group designs?
more difficult to reduce experimenter effects because single and double blind studies are difficult to implement, thus participants and experimenters might interpret good about the success of the treatment (ex. participant knows the study so they lie about results in order for the study to be proven true OR the experimenter avoids cues in order to skew results)
do single participant designs have internal validity?
its very limited
is one participant ever representative of everyone? chapter: for an example involving the visual system*
normally not, but in some cases they do represent the population if the variability of the population is limited
what are necessary control techniques in single participant ABAB designs?
baseline or base rate (A)
independent variable or treatment (B)
withdrawing the treatment (A)
reintroduction of the treatment (B)
what is a baseline?
a measure of the behaviour in question prior to manipulation of the independent variable or treatment, serves as a reference point (similar to control group)
how long should behaviour be measured in order to establish an accurate baseline? the a in Abab
when you believe you have shown stability in the behaviour you are measuring (if you can show a stable level of responding over 10-20 trials, then you can make a relatively strong case for stability)
ex. studying continuous handwashing, to find the baseline see how much harold washes his hands, if his hand washing continues to increase or stay remotely the same, this would be valid (because any baseline is that the behaviour will continue in the same fashion predicted by the baseline)
with harold washing his hands, what is an independent variable that may be established? the B in aBab
give him a drug (THC) and observe his hand washing, it was seen that the drug decreased the amount of handwashing harold did
why dont we just conduct an AB design?
would never know for sure if it was the independent variable (B) that caused a different behaviour
ex. wouldnt know if it was the THC that decreased the amount of handwashing
what is an ABA design?
a simple reversal design/ withdrawal design (sometimes treatments produce an immediate change in behaviour)
adding more reversals (AB) can address two issues with interpreting the results from an ABA reversal design, what are they?
- a single reversal could be caused by a random fluctuation in the persons behaviour
- ethical concern (might not be beneficial for the participant for the design to end with a withdrawal of the treatment)
what do we do in the A part of abAb?
withdraw the independent variable (THC) from the participants system (maybe give placebo) and observe handwashing. if the THC was the main reason for the decrease in handwashing, then removing it from his system should bring his handwashing levels back up
what are some interesting factors to withdrawing the treatment? abAb
- sometimes treatment is permanent and not reversible (can be bad (might never know if it was your treatment that cause permanent change in behaviour) or good (want to produce permanent positive changes in a participants behaviour))
- ethical issues (should you stop treatment and return a participant to a previous dysfunctional level? is it right to stop treatment and return harold to a dysfunctional state?)
what occurs in the B in abaB?
second treatment phase
if the first two phases of an experiment (AB) are similar to the results observed in the following two phases (abAB), what conclusion might you draw from the remainder of the results in the ABAB design?
the treatment is working
ex. when THC is given, handwashing decreased, when going through withdrawal his handwashing increased again, when getting the THC treatment again his handwashing decreased
what is something you should always remember when conducting an ABAB design?
only manipulate or vary one variable at a time or else you wont know which one the treatment is
if you want to use two independent variables at a time you will need to get creative
ex. A= baseline, B= independent variable 1, C=independent variable 2
A-B-A-B-BC-B-BC
what is the A-B-A-B-BC-B-BC design often called?
the interaction design for single participant research
how does the the interaction design for single participant research work?
purpose is to determine if C has an effect in addition to B alone A-B-A-B-BC-B-BC
what does this interaction design for single participant research design tell you? A-B-A-B-BC-B-BC-C-BC
want to determine the effect of each variable alone
what is an across situations multiple baseline design?
same behavior is measured in different settings (ex. at home or school) at different times in each setting
what is an across behaviours multiple baseline design?
several different behaviours of a single participant are measured over time. at different times, an intervention is introduced to target each of these behaviors
ex. reward system introduced to increase reading behavior of a child
what is an across subjects multiple baseline design?
behavior of several participants is measured over time (i.e. across different participants or settings)
ex. preschool children (kate, ben, tyler, jake) were introduced to play intervention at different times and it did not have the exact same effect on compliance across children
when can multiple baseline designs be done?
with or without a withdrawal treatment condition and a reintroduction of treatment (can be performed using only an AB type of design or by the more typical ABAB type of design)
what are quasi experiments?
designs that address the need to study independent variables in settings in which true experimental designs are not possible (they are used out of necessity)
what does quasi mean in quasi experiments?
“as if”
can quasi experiments make causal inferences?
no, but they can indicate how variables are related
what are the two key factors that distinguish true experiments from quasi experiments?
- quasi experimental designs typically involve the selection of participants from groups that already exist
- extent to which the experimenter has control over the variables in the study
what does “quasi experimental designs typically involve the selection of participants from groups that already exist”, mean?
that true experiments randomly assign participants to conditions and quasi experiments randomly select participants for different research conditions from groups that already exist
ex. most research that looks at differences between males and females will be quasi, since there cant be groups with both males and females
why are quasi experiments sometimes called ex post facto designs?
ex post facto= after the fact
because the experiment is carried out only after the formation of the different groups
how is the “extent to which the experimenter has control over the variables in the study” different between true and quasi experiments?
true experiment: experimenter has the luxury of ruling out alternative explanations by controlling the what, when, where and how of most true experimental designs
quasi: experimenter rarely has such control and will never be able to rule out the possibility that the differences found resulted from a pre- existing difference between the selected groups of participants
ex. study the difference between in class learning and online learning in two preexisting groups, we couldnt be sure that any observed learning differences resulted from the teaching methods (in a true experiment confounds that may have pre existed would be equally dispersed among groups through random assignment, thus eliminated)
which experiment type is better, true experiment or quasi experiment?
true experiment (BUT if complete control of some variable is impossible, dont give up, instead use a quasi experimental design)
what is a common type of quasi experimental design?
non- equivalent group design
what does the non- equivalent group design consist of?
a control and experimental group (but lacks random assignment of participants to each condition)
a pre test measure and a post test measure of some dependent variable
what are selection differences?
pre existing differences between control and experimental groups that are confounded with the independent variable, these differences provide an alternative explanation for the results
how do non equivalent group designs work?
psychologists use intact groups that they assume are similar and assign one group to the control condition and the other group to the experimental condition
ex. psychologist might use two grade 4 classes from the same school and assign one class as the experimental group and the other as the control
what is the main problem with non equivalent group designs?
the control and experimental groups are never as similar as they would be if random assignment took place (must be careful when interpreting the results)
in non equivalent group designs, the control and experimental groups are never as similar as they would be if random assignment took place, what is an example of this?
you are interested in determining how effective a new, interactive and online study tool is in helping students learn the concepts of research methods. one professor that teaches research methods courses agrees to let their students use the interactive online site and they teach two sections. the professors morning class becomes the experimental group and the afternoon class becomes the control. you give the morning class access to the online interactive study tool two weeks before the final test and just give a pep talk to the afternoon class. the dependent variable is research methods knowledge (measured in two ways: 1. recording students current mark on the last day of class and 2. post test score)
what would a non equivalent control group pretest- posttest design look like?
not a true experiment (but is the most useful quasi experimental design)
adding a pretest gives an advantage (know pretest scores so can see whether the groups were the same or different on the pretest)
what is the history threat to internal validity?
occurs when some event (i.e. increased access to computers) occurs either before or during the course of treatment
when does a cross over pattern on a graph appear?
when the control condition appears to remain the same over time, but the treatment condition starts out lower and ends higher (or starts higher and ends lower)
what is regression toward the mean/ regression effect?
tendency of participants or groups of participants with extreme scores on a first measure to score closer to the population mean on a second testing (the scores regress toward the population mean)
what are four not as common quasi experimental methods?
one group posttest- only design
one group pretest-posttest design
interrupted time series design
control series design
longitudinal
why is the one group posttest-only design considered a quasi experimental design?
lacks a control group or other source of comparison
cant make causal inferences about the effect of the independent variable on the dependent variable
lacks internal validity
what is a one group pretest-posttest design?
measures participants before the manipulation (pretest) and again afterwards (posttest)
it is an improvement from the one group posttest only design
why is the one group pretest-posttest design considered a quasi experimental design?
threats to internal validity
what is internal validity?
ability to claim based on a true experiment that the independent variable causes changes in the dependent variable
when is something considered a threat to internal validity?
when it allows for some other reasonable alternative explanation for changes to the dependent variable, other than the manipulation of the independent variable (when a true experiment is well designed it doesnt have threats to internal validity)
what are the 8 threats to internal validty?
history
maturation
testing
instrument decay
regression toward the mean
attrition
selection effects
cohort effects
how is history a threat to internal validity?
historical event that affects all or most participants and that is not of interest to the researcher
how is maturation a threat to internal validity?
natural changes to participants short term states (i.e. fatigue) or long term development (i.e. education) not of interest to the researcher
how is testing a threat to internal validity?
simply taking the pretest influences peoples responses to the posttest
how is instrument decay a threat to internal validity?
characteristics of the measurement instrument changes with repeated use
how is regression toward the mean a threat to internal validity?
may occur when participants are chosen or groups are divided based on extreme scores on the pretest, because extreme scores tend to become less extreme on repeated measurement
how is attrition a threat to internal validity?
participants leave the study
how are selection effects a threat to internal validity?
groups are divided based on any reason other than random assignment
how are cohort effects a threat to internal validity?
groups are divided by age. a special type of selection effect
how can threats to internal validity affect true experimental designs that use random assignment?
threat affects one group differently than the other (i.e. history)
if the threat affects both groups equally it (i.e. if the dependent measure becomes less reliable between the pretest and the posttest (ex. instrumental decay) for all groups, any genuine effect of the manipulation might be masked by this increase in measurement error)
how do you overcome threats to internal validity?
by using appropriate control groups (it can help rule out the effects of history, regression toward the mean…)
what is a interrupted time series design?
design similar to one group pretest-posttest design and the non equivalent group pretest-posttest design but instead have multiple pretests and multiple posttests instead of just one
ex. study: in 1996 driving drunk was legal, in 1997 it was illegal, instead of just observing 1996 and 1997 you might observe 1990 to 1998 (interrupted time series design)
what is a control series design?
the interrupted time series design with a control group
ex. in the drunk driving experiment, for the control, use other provinces
what is developmental psychology?
new field of study that is the scientific study of change and/or continuity from conception to death (sheds light on issues like nature vs nurture and if “you cant teach an old dog new tricks” is true or not)
what three scientific methods do developmental psychologists employ to study how we change?
the longitudinal method
the cross sectional method
sequential method
what are longitudinal designs/ methods?
methods that collect data on the same people as they grow older (shares some similarities with within subjects design)
ex. class takes an IQ test in grade 1,3,5,7,11…
why is the data gathered in longitudinal designs invaluable?
because they assess developmental changes that occur over time as a result of aging and avoid some of the confounding cohort effects observed in other developmental designs
when do confounding cohort effects happen?
happen when differences in a study’s results are due to the unique experiences of a specific group of people rather than the factors the study is trying to investigate
what are the drawbacks of longitudinal designs?
time consuming
expensive
many people drop out of the study
what occurs in a cross sectional study?
studys people of different ages at a single point in time (shares some similarities with between subjects design)
ex. all the grade 1s, 2s, 3s,4s… from a school district are given personality tests at the same time and developmental changes in personality are derived by looking at changes between the different groups
advantages to cross sectional designs
good at showing similarities and differences among different age groups of people
fast
easy
inexpensive to run
dont have the problem of attrition and repeated testing
disadvantages to cross sectional designs
cannot be sure that differences among age groups are due only to age (developmental change is not observed directly with the group of pepple but is based on comparisons among different cohorts (groups of people born at around the same time))
what method is more common, longitudinal method or cross sectional method?
cross sectional method (as its inexpensive and yields immediate results)
ex. to study a group of people from age 20 to 50 it would take 30 years to conduct a longitudinal design, but a lot less for a cross sectional design
what is a sequential design?
powerful method that combines the methods of both the longitudinal and cross sectional methods
how does the sequential design work?
ex. you use the cross sectional method to compare 55 year olds and 65 year olds, all participants are then studied using the longitudinal method
what is program evaluation research?
research on programs that are proposed and implemented to achieve some positive effect on a group of people. such programs may be implemented in schools, work settings or entire communities
ex. program designed to reduce obesity by teaching elementary school children to make healthy choices about diet and activity
what are the 5 types of questions that might be emphasized for a evaluation research?
evaluation of need
program theory assessment
process evaluation/ program monitoring
outcome evaluation/ impact assessment
efficiency assessment
what types of questions are asked in evaluation needs?
asking whether there are problems that need to be addressed in a target population
ex. what services do homeless people need most?
what is in program theory?
after identifying needs, this program addresses them
may involve researchers, service providers and prospective program clients collaborating to ensure that the program addresses the target populations needs
what does assessing program theory include?
articulating the rationale for how members will benefit from the program (including how they will access and use the programs services), this rationale can then be evaluated (will this program actually reach the target population as intended?)
what occurs in process evaluation?
when the program is under way, the researcher monitors whether it is reaching the target population, attracting enough clients and whether the staff is providing the planned services
(the researcher seeks evidence that the program is doing what it is supposed to do)
involves questionnaires and interviews, observational studies, and analysis of records kept by program staff
which research is important to avoid concluding that a program is ineffective, when really it is not being implemented properly?
process evaluation
what is outcome evaluation?
are intended outcomes of the program being realized?
to determine this, the evaluation researcher must devise a way of measuring the outcome and then study the impact of the program on the outcome measure
what is the efficiency essessment?
once the program has an intended effect, researchers determine whether the benefits are worth the programs cost or if the resources used to implement the program might be put to some better use
