Research Exam 2

research design

specific set of tactics used to carry out a plan of attack

group research designs

one or more groups are exposed to one or more levels of the IV. The average performance of the group on the DV is obtained to see/test if a relationship exists

between-subject designs

performances of separate groups of subjects are measured and compared

needs to have subject matching and subject randomizatiion

exerimental group

groups that receive the treatment

control group

group that does not recieve the treatment

within subject designs

studying the effects on the same group

performance of the same subject compared in different conditions

single subject designs

can be across subjects or across behavior

begins with baseline, but then one level of intervention is applied until stability, then a second level is applied, etc

focus on the behavior of indiviual subjects, rather than the average of a group

internal validity

degree to which a design meets it dual purpose within a study-did it answer the research Qs and provide credible evidence

external validity

degree to which generalizations can be made or transferred outside the confines of the study

threats to internal validity on a single study

history

maturation

instrumentation

testing

threats to internal validity on a multi group

selection bias

regression to the mean

social interaction

attrition

What are the weaknesses of the single subject design compared to the group designs?

single subject design has a lower external validity bc of the smaller number of participants

order bias, in group designs only one group recieves treatment

hierarchy of evidence

highest: systematic review and meta analysis

second highest: randomized control trials

third: cohort studies

lowest: expert opinion

quasi expiremental study

no randomization, has a control group

what would be the reason the preexpiremental design considered weak design

it has no control group and no randomization, giving it a low internal and external validity.

preexpiremental design

exploratory, weak in validity

no random assignment or a control

X to O

randomized control trial/ true experimental design

has a strong internal validity

randomized pre and post test with a control group design

meta analysis

It combines and analyzes quantitative data from multiple independent studies addressing the same research question to produce a single, pooled summary effect size

on the top of the evidence pyramid

treatment efficacy

Does one treatment work better than another?

treatment effectiveness

Does the treatment work?

Clinically significant change in the skill/ability/etc that was targeted for improvement

treatment effect

in what way does the treatment alter the behavior

therapeutic effect

something improved that you didn’t intend

mixed design

combines quanitative and qualitative data

how to solve internal validity threats

add a control group

larger sample size counters testing

filler tasks hides the purpose of the study

external validity threats

selectiion bias

regression to the mean- if you score extremely low or high on the test you will score closer to the middle the second time

social interaction

attrition

how to combat external validdity threats

random assignment

blinding

generalizability

Can this effect be generalized beyond the subjects in the study?

random assignment

an experimental technique where study participants are placed into different groups (such as a treatment group or a control group) by chance

subject matching

experimenter attempts to match pairs of participants on all extraneous variables possible

to avoid confounding variables

sequence randomization

randomizing the order of events, a solution to sequencing effects

counterbalancing

a solution to sequencing effects, an experimental design technique used to control for order and sequence effects in repeated-measures or within-subjects studies, usability testing, and clinical trials

experences in regular order and then reverse order

order effects

the performance earlier in an experiment has an effect on later portions (positive or negative)

carryover effect

the influence of a specific treatment condition on the performance of the next condition

baseline A in a single subject design

baseline data, pre test

treatment intervention in a single study design

B, treatment

withdrawal design (ABA design)

pretest A, have treatment B, then withdraw treatment and see if client gets worse

ABAB design

ABA design except you reintroduce the treatment

preffered because it allows for replication, treatment can be at the end of the study

reliability

the consistency of a measure

validity

extent to which the scores from a measure represent the variable they are intended to

institutional review board

informed consent

privacy and confidentiality

test-retest reliability

measuring a property that you expect to stay the same overtime

interrater

multiple researchers making observations or ratings about the same topic

parallel forms

using two different tests to measure the same thing

internal conssitency

using a multiterm test where all items are intended to measure the same variable

face validity

the experiment has a good representation of what you want to test

construct validity

Does the test measure the concept that it’s intended to measure?

criterion validity

Do the results correspond to a different test of the same thing?

different tests to test the same thig to see if the tests are valid

content validity

Is the test fully representative of what it aims to measure?