Midterm - Flashcards

« Previous card

Next card »

INCORRECT CORRECT

Click Card to flip

what is research

a critical process for asking and attempting to answer questions about the world

Click Card to flip

false consensus effect

a person tends to overestimate the degree of agreement that others have with them. tendency for ppl to assume that their own opinions, beliefs, preferences etc are 'normal' and that others think the same way ***
this bias tends to create a consensus that doesn't exist (thus false consesus)

Click Card to flip

steps in the scientific method

ask a ? ***
do background research ***
construct a hypothesis ***
develop a method ***
make predictions ***
test hypothesis ***
analyze data ***
communicate results ***

Click Card to flip

why research skills matter

understand and interpret info ***
maximize effectiveness of our practice ***
discover and validate scientific concepts ***

Click Card to flip

goals of research

exploration ***
description ***
prediction ***
explanation ***
action ***

Click Card to flip

being a critical consumer of research

the who? - who's doing the research ***
the where? - where is it being done; lab v natural setting ***
the how? ***
the why? ***

Click Card to flip

methods --> ______

outcomes ***
methods used affect your outcomes! ***
if you cut corners on methods, you will get poor results ***

Click Card to flip

ways of gaining knowledge

custom and tradition - info passed down ***
authority - i.e. govt provided info ***
personal experience ***
pseudo science - looks like science but its not; i.e. phrenology

Click Card to flip

the scientific approach

empiricism - only OBSERVABLE data; NOT OPINION ***
falsifiability - hypothesis must be able to be supported or not ***
peer review

Click Card to flip

research process: begin with a _____ topic, then get specific

general

Click Card to flip

define the research question

is there a relationship? ***
how many people do this? **
what causes this condition? ***
when is this occurring?

Click Card to flip

poor scientific question

is exercise healthy?

Click Card to flip

good scientific question

can brisk walking 3x per week for 6 mos improves CV health in overwt adult? ***
SPECIFIC

Click Card to flip

types of questions

nonexperimental - descriptive, correlational ****
experimental

Click Card to flip

independent variable

variable which is manipulated

Click Card to flip

dependent variable

variable which is measured

Click Card to flip

basic research

research for the sake of knowledge ***
tries to answer fundamental questions ***
designed to address theoretical issues concerning phenomena such as cognition, emotion, learning etc

Click Card to flip

applied research

applying knowledge to some condition ***
conducted to address issues in which there are practical problems and potential solutions

Click Card to flip

example of a descriptive research question

how many uc students smoke? ***
how many have tried to quit in the past year? ***
what strategies have they used to quit?

Click Card to flip

example of a correlational research question

is smoking frequency related to drinking more alcohol?

Click Card to flip

ex of an experimental research question

does running 3x a week for 6 mos have an effect on smoking among uc students?

Click Card to flip

example of a hypothesis

uconn students who smoke and who start a running program will smoke less after 6 mos compared to controls

Click Card to flip

purpose of your research proposal

represents your understanding of the topic ***
demonstrates appropriateness of methods to study the topic ***
guide for data collection ***
to obtain grant funding ***
IRB review ***
communicate/demonstrate knowledge

Click Card to flip

steps in research process - funnel shaped

1. initiate an idea - general ***
2. select a specific question ***
3. review the literature ***
4. determine variables of interest ***
5. define problem specifically ***
6. re-review literature

Click Card to flip

hypothesis

formal statement asserting what you believe is true in a particular situation; statement asserting that 2 or more variables are related to one another

Click Card to flip

title

usually written last ****
concise but include variables of interest; includes important key wrds

Click Card to flip

name and title of investigators

use full names with abbreviations for titles ***
principle investigators (PIs) and Project Directors ***
full names of insitution(s) ****
i.e. Uconn Health Center, Farmington, CT

Click Card to flip

abstract

written last ***
SUMMARY OF PROJECT (~120 words) ***
intro to problem ***
what is known/not known ***
purpose and objectives ***
methods and overall predictions ***
NO references!

Click Card to flip

intro to lit review (____ tense)

past ***
brief description of the overall problem ***
cite key studies, using reputable sources - give stats and facts; discuss if studies to date conflict/agree ***
end with transition to lit review

Click Card to flip

lit review (____ tense)

past ***
purpose - explain rationale for your study ***
provide foundation for your research q - what is known? do studies agree/disagree? what's next? ***
use subheadings ***
end section with conclusion ***
use primary references v secondary references ***
refer to each study by author and yr ***
end with objectives - list by #, measurable, hypotheses following each objective

Click Card to flip

methods (____ tense)

future ***
dividied into sections - design, participants ***
explain measures and procedure in detail ***
data analysis should be linked with hypotheses and stats to be used must match hypotheses ***
timeline

Click Card to flip

predictions

what outcomes do you expect? why? ***
mirrors intro section ***
sums up implications ***
what you expect to find

Click Card to flip

discussion

generalizability of the findings ***
variables not controlled for ***
other study limitations/strengths ***
how were weaknesses addressed? ***

Click Card to flip

references

ABC order ***
articles/books/disserations/reviews ***
use APA format ***
start on new page and double space ***
list all authors, page numbers ***

Click Card to flip

appendix

surveys, questionnaires, other tools etc - develop your own or reference ***
informed consent form

Click Card to flip

final proposal

review at least 3x using checklist: **
1 for grammar ***
1 for content ***
1 for necessary components ***
follow samples

Click Card to flip

milgram's: participants

40 males between 20-50; range in occupation and education ***
told paricipants they were testing memory - teaching and learning ***
given word pairs - had to recall 2nd half or would get shocked

Click Card to flip

what happened at 150V?

learner would say he wanted to stop but experimenter would urge teacher to continue

Click Card to flip

______ laughter occurred in 14/40 teachers

nervous

Click Card to flip

learner was an ________ of the experimenter

accomplice ***
provided many wrong responses on purpose ***
teacher was the one being studied

Click Card to flip

___% of teachers obeyed the experimenter's commands fully

50

Click Card to flip

obedience was _____ as physical closeness to learner increased.

reduced

Click Card to flip

obedience was ______ when experimenter was physically removed

reduced

Click Card to flip

survey for milgrams

people would disobey by 135V ***
none beyond 300V ***
psychiatrists said only 1/1000 would go to 450V

Click Card to flip

reality for milgrams

50% of teachers went all the way! ***
why? who's responsible for the learner - experimenter

Click Card to flip

distance between teacher and learner

teacher was more obedient when learner was farther away - depersonalized

Click Card to flip

Definition

Click Card to flip

closeness of authority

teacher more obedient when authority was close by

Click Card to flip

perception of institution

i.e. Yale ***
more likely to go along with a reputable institution

Click Card to flip

sampling

goal = represent target population ***
identify target pop then identify exp pop (accessible population) ***
choose a sampling procedure ***
define inclusion/exclusion criteria

Click Card to flip

inclusion/exclusion criteria

determine criteria for which to include and exclude people from your study ***
i.e. exclude people on certain medications or with a fam hx

Click Card to flip

sampling bias

particular attributes of a chosen sample arent representative ***
leads to sampling error

Click Card to flip

probability sampling

simple random **
disproportional **
cluster **
stratified

Click Card to flip

simple random sampling

every member of the population has an equal probability of being selected for the sample. (i.e. if the population has 1000 members, each person has a 1/1000 chance of being selected) ***
randomly draw people

Click Card to flip

disproportional sampling

uneven numbers in each group **
i.e. uneven genders

Click Card to flip

cluster sampling

randomly selecting clusters **
i.e. randomly select retirement homes and then randomly assign the homes

Click Card to flip

stratified sampling

population is divided into subgroups and random sampling techniques are then used to select sample members from each stratum **
i.e. sexual attitudes survey may stratify on the basis of age or gender as these attributes may affect sexual attitudes

Click Card to flip

advantages of probability sampling

representative of the population **
for cluster - researcher does not have to sample from lists of individuals in order to get a truly random sample

Click Card to flip

disadvantages of probability sampling

may cost more; may be difficult to get full list of all members of any population of interest/any cluster

Click Card to flip

nonprobability sampling

probability sampling not feasible; each person does NOT have an equal chance of being selected **
used if studying something very specific **
convenience **
quota **
purposive **
snowball

Click Card to flip

convenience sampling

select a sample in any way that is convenient **
use convenient pool - i.e. volunteers ***
ex. stand outside union on tuesday at 9, ask people in your class **
be aware of biases - be sure to randomize

Click Card to flip

quota sampling

sample conveniently until you reach a certain number

Click Card to flip

purposive sampling

purposefully go after a certain type of person - i.e. people under age 30, people with children

Click Card to flip

snowball sampling

recruit 1 person and have them recruit others and have those people recruit others and so on **
good for hard to reach populations - i.e. drug users

Click Card to flip

advantages of nonprobability sampling

inexpensive, efficient, convenient - convenience **
sample includes only those youre interested in - purposive **
inexpensive, efficient, convenient, slightly more sophisticated than convenience - quota

Click Card to flip

sample size

needs to be representative and large enough **
larger samples are more likely to yield data that accurately reflect the true population value

Click Card to flip

sampling frame problems

sampling frame = ACTUAL population of individuals from which a random sample will be drawn ***
i.e. if you only take people out of the phonebook- your results will be biased (won't include people unlisted or without phones)

Click Card to flip

response rate issues

you send out 1000 surveys but maybe only 500 come back **
important because response rate indicates how much bias there might be in the final sample of respondents - nonrespondents may differ in many ways (age, income, education etc)

Click Card to flip

experimental control

primary goal - show CAUSE/EFFECT between IV and DV

Click Card to flip

confounding variables

any variables that can effect your outcome ***
extrinsic - outside person; i.e. environment **
intrinsic - in person; i.e. age

Click Card to flip

how to control for confounds

eliminate (exlusion criteria) i.e. age, gender ***
control environment ***
sample to make subjs equivalent ***
minimize impact of researcher/staff - how they interact w patients can be a confound; train and monitor data collectors ***
use same equipment ***
use random assignment ***
id confounds ***
control subjects activities ***
reduce obs bias - use single or double blind

Click Card to flip

reduce inter-subject differences

select homogenous subjects - i.e. aqe, gender, race ***
matching - match people in 1 group with those in the other ***
subjects as their own control ***
statistical techniques

Click Card to flip

nonexperimental designs: descriptive

describes data and characteristics about the population or phenomenon being studied ***
does NOT determine cause/effect

Click Card to flip

nonexperimental designs: correlational

describes a relationship between 2 variables ***
does NOT determine cause/effect ***
degree of association (-1 to 1) ***
positive, negative, zero, curvilinear ***

Click Card to flip

experimental designs

determine cause/effect

Click Card to flip

key uses of nonexperimental designs

describes co-occurence of events ***
generate hypotheses for exp research ***
compare 2 methods or tests ***
prediction

Click Card to flip

key limitations of nonexperimental research:

cant draw cause/effect

Click Card to flip

within-subjects design

ONE GROUP ***
pretest - intervention - posttest ***
subjects act as THEIR OWN CONTROL GROUP! ***
prescores (control) and postscores (exp)

Click Card to flip

quasi experimental

randomly assigning SITES ***
i.e. 10 clinic sites as control and 10 clinic sites as experimental
limitations: those who know theyre getting control may not go; sites could differ (i.e. inner city v suburb)

Click Card to flip

true experimental

true control group **
true random assignment **
groups are statistically equal **
blind **
sample size is relatively large **
control confounds **
participants randomly assigned to diff conditions

Click Card to flip

factorial design

at least 2 IVs with each having 2 levels **
want to see different interactions/combinations

Click Card to flip

repeated measures design

same individuals participate in all/both conditions ***
participants are repeatedly measured on the DV after being in each condition of the experiment ***
take measures repeatedly thru experiment

Click Card to flip

ABA/ABAC design

A = baseline *
B = intervention *
C = another intervention *

Click Card to flip

internal validity

how well the test measures what its intended to measure **
refers to the ability to draw conclusions about causal relationships from your data ***
design of study, measurement of DV

Click Card to flip

External validity

extent to which the results can be generalized to the population youre studying

Click Card to flip

history threats

events outside the study; factors that happen over a period of time; occur between your 1st and 2nd measurements **
problem with longitudinal studies **
solutions: ***
large sample sizes will wash out some influence ***
random assignment to avoid unequal influences on 1 group over the other

Click Card to flip

maturation threats

changes in subjects over time; get older, wiser etc. ***
ex. as they get older, may decrease smoking because they are more concerned about their health not bc of your intervention ***
solutions: **
random assignment

Click Card to flip

repeated measurement threats

practice effects **
participants may develop practice effect; may figure out what youre studying **
solution: **
change format

Click Card to flip

attrition threats

dropouts from the study; final sample not representative **
problem with longitudinal studies **
solutions: **
statistically control for differences in sample at the beginning and end

Click Card to flip

testing effects threats

differences may arise from the way measurements are taking place **
solutions: **
measurements must be standardized! **
same person measuring **
same tool used

Click Card to flip

hawthorne effect

impact of someone being observed; mat affect how participant acts ***
ex. productivity at work **
solution: **
keep researcher hidden from sight

Click Card to flip

placebo effect

just being a part of a study may change people's bx ***
solution: ***
double or single blind

Click Card to flip

halo effect

giving experimental group incentive because you want to see a specific outcome

Click Card to flip

measurement threats to internal validity

accuracy of instrument **
researcher training and skill **
floor/ceiling effects **
improper measurement of variables - nominal, ordinal, interval, ratio **
construct validity of instrument **

Click Card to flip

ceiling effect

only measures up to a certain pt; not capturing whole range of possibilites **
IV appears to have no effect on DV measure only because participants quickly reach the max performance level

Click Card to flip

floor effect

occurs when a task is so difficult that hardly anyone can perform well

Click Card to flip

hierarchy of the levels of measurement

nominal, ordinal, interval, ratio

Click Card to flip

nominal

arbitrary ***
categories with NO NUMERICAL SCALES ***
ex. male/female; introvert/extrovert ***
impossible to define any quantitative values and/or differences between/across categories

Click Card to flip

ordinal

rank order; numerical values limited ***
ex. 2, 3, and 4 star restaurants ***
ex. A B C D F grades ***
intervals between items unknown

Click Card to flip

interval

numerical properties are literal ***
ex. intelligence, temperature, apt test score ***
NO TRUE ZERO

Click Card to flip

ratio

TRUE ZERO ***
weight, age ***
zero indicates absence of value measured ***
can form ratios (someone weights twice as much as someone else)

Click Card to flip

construct validity of the instrument

refers to the adequacy of the operational definition of variables; does the operational definition of a variable actually reflect the true theoretical meaning of the variable?

Click Card to flip

criterion related validity

used to demonstrate the accuracy of a measure or procedure by comparing it with another measure or procedure which has been demonstrated to be valid

Click Card to flip

predictive validity

instrument does a good job at predicting actual bx; research that uses a measure to predict some future bx

Click Card to flip

concurrent validity

instrument gives outcome similar to an existing measurement

Click Card to flip

convergent validity

Convergent validity, is the degree to which an operation is similar to (converges on) other operations that it theoretically should also be similar to. ***

extent to which scores on the measure in question are related to scores on other measures of the same construct or similar
constructs
should converge

Click Card to flip

discriminant validity

Discriminant validity describes the degree to which the operationalization is not similar to (diverges from) other operationalizations that it theoretically should not be similar to.
***
measure should discriminate between construct being measured and other unrelated constructs

Click Card to flip

threats to external validity

sample characteristics - must be similar to target pop characteristics! ***
subject/tx interaction - only certain participants do well - i.e. youngest segment ***
artificial research setting - may not generalize well to actual environment

      Mode:   CARDS LIST       ? pages   PRINT EXIT

what is research	a critical process for asking and attempting to answer questions about the world
false consensus effect	a person tends to overestimate the degree of agreement that others have with them. tendency for ppl to assume that their own opinions, beliefs, preferences etc are 'normal' and that others think the same way *** this bias tends to create a consensus that doesn't exist (thus false consesus)
steps in the scientific method	ask a ? *** do background research *** construct a hypothesis *** develop a method *** make predictions *** test hypothesis *** analyze data *** communicate results ***
why research skills matter	understand and interpret info *** maximize effectiveness of our practice *** discover and validate scientific concepts ***

Generated by Koofers.com

goals of research	exploration *** description *** prediction *** explanation *** action ***
being a critical consumer of research	the who? - who's doing the research *** the where? - where is it being done; lab v natural setting *** the how? *** the why? ***
methods --> ______	outcomes *** methods used affect your outcomes! *** if you cut corners on methods, you will get poor results ***
ways of gaining knowledge	custom and tradition - info passed down *** authority - i.e. govt provided info *** personal experience *** pseudo science - looks like science but its not; i.e. phrenology

Generated by Koofers.com

the scientific approach	empiricism - only OBSERVABLE data; NOT OPINION *** falsifiability - hypothesis must be able to be supported or not *** peer review
research process: begin with a _____ topic, then get specific	general
define the research question	is there a relationship? *** how many people do this? ** what causes this condition? *** when is this occurring?
poor scientific question	is exercise healthy?

Generated by Koofers.com

good scientific question	can brisk walking 3x per week for 6 mos improves CV health in overwt adult? *** SPECIFIC
types of questions	nonexperimental - descriptive, correlational **** experimental
independent variable	variable which is manipulated
dependent variable	variable which is measured

Generated by Koofers.com

basic research	research for the sake of knowledge *** tries to answer fundamental questions *** designed to address theoretical issues concerning phenomena such as cognition, emotion, learning etc
applied research	applying knowledge to some condition *** conducted to address issues in which there are practical problems and potential solutions
example of a descriptive research question	how many uc students smoke? *** how many have tried to quit in the past year? *** what strategies have they used to quit?
example of a correlational research question	is smoking frequency related to drinking more alcohol?

Generated by Koofers.com

ex of an experimental research question	does running 3x a week for 6 mos have an effect on smoking among uc students?
example of a hypothesis	uconn students who smoke and who start a running program will smoke less after 6 mos compared to controls
purpose of your research proposal	represents your understanding of the topic *** demonstrates appropriateness of methods to study the topic *** guide for data collection *** to obtain grant funding *** IRB review *** communicate/demonstrate knowledge
steps in research process - funnel shaped	1. initiate an idea - general *** 2. select a specific question *** 3. review the literature *** 4. determine variables of interest *** 5. define problem specifically *** 6. re-review literature

Generated by Koofers.com

hypothesis	formal statement asserting what you believe is true in a particular situation; statement asserting that 2 or more variables are related to one another
title	usually written last **** concise but include variables of interest; includes important key wrds
name and title of investigators	use full names with abbreviations for titles *** principle investigators (PIs) and Project Directors *** full names of insitution(s) **** i.e. Uconn Health Center, Farmington, CT
abstract	written last *** SUMMARY OF PROJECT (~120 words) *** intro to problem *** what is known/not known *** purpose and objectives *** methods and overall predictions *** NO references!

Generated by Koofers.com

intro to lit review (____ tense)	past *** brief description of the overall problem *** cite key studies, using reputable sources - give stats and facts; discuss if studies to date conflict/agree *** end with transition to lit review
lit review (____ tense)	past *** purpose - explain rationale for your study *** provide foundation for your research q - what is known? do studies agree/disagree? what's next? *** use subheadings *** end section with conclusion *** use primary references v secondary references *** refer to each study by author and yr *** end with objectives - list by #, measurable, hypotheses following each objective
methods (____ tense)	future *** dividied into sections - design, participants *** explain measures and procedure in detail *** data analysis should be linked with hypotheses and stats to be used must match hypotheses *** timeline
predictions	what outcomes do you expect? why? *** mirrors intro section *** sums up implications *** what you expect to find

Generated by Koofers.com

discussion	generalizability of the findings *** variables not controlled for *** other study limitations/strengths *** how were weaknesses addressed? ***
references	ABC order *** articles/books/disserations/reviews *** use APA format *** start on new page and double space *** list all authors, page numbers ***
appendix	surveys, questionnaires, other tools etc - develop your own or reference *** informed consent form
final proposal	review at least 3x using checklist: ** 1 for grammar *** 1 for content *** 1 for necessary components *** follow samples

Generated by Koofers.com

milgram's: participants	40 males between 20-50; range in occupation and education *** told paricipants they were testing memory - teaching and learning *** given word pairs - had to recall 2nd half or would get shocked
what happened at 150V?	learner would say he wanted to stop but experimenter would urge teacher to continue
______ laughter occurred in 14/40 teachers	nervous
learner was an ________ of the experimenter	accomplice *** provided many wrong responses on purpose *** teacher was the one being studied

Generated by Koofers.com

___% of teachers obeyed the experimenter's commands fully	50
obedience was _____ as physical closeness to learner increased.	reduced
obedience was ______ when experimenter was physically removed	reduced
survey for milgrams	people would disobey by 135V *** none beyond 300V *** psychiatrists said only 1/1000 would go to 450V

Generated by Koofers.com

reality for milgrams	50% of teachers went all the way! *** why? who's responsible for the learner - experimenter
distance between teacher and learner	teacher was more obedient when learner was farther away - depersonalized
	Definition
closeness of authority	teacher more obedient when authority was close by

Generated by Koofers.com

perception of institution	i.e. Yale *** more likely to go along with a reputable institution
sampling	goal = represent target population *** identify target pop then identify exp pop (accessible population) *** choose a sampling procedure *** define inclusion/exclusion criteria
inclusion/exclusion criteria	determine criteria for which to include and exclude people from your study *** i.e. exclude people on certain medications or with a fam hx
sampling bias	particular attributes of a chosen sample arent representative *** leads to sampling error

Generated by Koofers.com

probability sampling	simple random ** disproportional ** cluster ** stratified
simple random sampling	every member of the population has an equal probability of being selected for the sample. (i.e. if the population has 1000 members, each person has a 1/1000 chance of being selected) *** randomly draw people
disproportional sampling	uneven numbers in each group ** i.e. uneven genders
cluster sampling	randomly selecting clusters ** i.e. randomly select retirement homes and then randomly assign the homes

Generated by Koofers.com

stratified sampling	population is divided into subgroups and random sampling techniques are then used to select sample members from each stratum ** i.e. sexual attitudes survey may stratify on the basis of age or gender as these attributes may affect sexual attitudes
advantages of probability sampling	representative of the population ** for cluster - researcher does not have to sample from lists of individuals in order to get a truly random sample
disadvantages of probability sampling	may cost more; may be difficult to get full list of all members of any population of interest/any cluster
nonprobability sampling	probability sampling not feasible; each person does NOT have an equal chance of being selected ** used if studying something very specific ** convenience ** quota ** purposive ** snowball

Generated by Koofers.com

convenience sampling	select a sample in any way that is convenient ** use convenient pool - i.e. volunteers *** ex. stand outside union on tuesday at 9, ask people in your class ** be aware of biases - be sure to randomize
quota sampling	sample conveniently until you reach a certain number
purposive sampling	purposefully go after a certain type of person - i.e. people under age 30, people with children
snowball sampling	recruit 1 person and have them recruit others and have those people recruit others and so on ** good for hard to reach populations - i.e. drug users

Generated by Koofers.com

advantages of nonprobability sampling	inexpensive, efficient, convenient - convenience ** sample includes only those youre interested in - purposive ** inexpensive, efficient, convenient, slightly more sophisticated than convenience - quota
sample size	needs to be representative and large enough ** larger samples are more likely to yield data that accurately reflect the true population value
sampling frame problems	sampling frame = ACTUAL population of individuals from which a random sample will be drawn *** i.e. if you only take people out of the phonebook- your results will be biased (won't include people unlisted or without phones)
response rate issues	you send out 1000 surveys but maybe only 500 come back ** important because response rate indicates how much bias there might be in the final sample of respondents - nonrespondents may differ in many ways (age, income, education etc)

Generated by Koofers.com

experimental control	primary goal - show CAUSE/EFFECT between IV and DV
confounding variables	any variables that can effect your outcome *** extrinsic - outside person; i.e. environment ** intrinsic - in person; i.e. age
how to control for confounds	eliminate (exlusion criteria) i.e. age, gender *** control environment *** sample to make subjs equivalent *** minimize impact of researcher/staff - how they interact w patients can be a confound; train and monitor data collectors *** use same equipment *** use random assignment *** id confounds *** control subjects activities *** reduce obs bias - use single or double blind
reduce inter-subject differences	select homogenous subjects - i.e. aqe, gender, race *** matching - match people in 1 group with those in the other *** subjects as their own control *** statistical techniques

Generated by Koofers.com

nonexperimental designs: descriptive	describes data and characteristics about the population or phenomenon being studied *** does NOT determine cause/effect
nonexperimental designs: correlational	describes a relationship between 2 variables *** does NOT determine cause/effect *** degree of association (-1 to 1) *** positive, negative, zero, curvilinear ***
experimental designs	determine cause/effect
key uses of nonexperimental designs	describes co-occurence of events *** generate hypotheses for exp research *** compare 2 methods or tests *** prediction

Generated by Koofers.com

key limitations of nonexperimental research:	cant draw cause/effect
within-subjects design	ONE GROUP *** pretest - intervention - posttest *** subjects act as THEIR OWN CONTROL GROUP! *** prescores (control) and postscores (exp)
quasi experimental	randomly assigning SITES *** i.e. 10 clinic sites as control and 10 clinic sites as experimental limitations: those who know theyre getting control may not go; sites could differ (i.e. inner city v suburb)
true experimental	true control group ** true random assignment ** groups are statistically equal ** blind ** sample size is relatively large ** control confounds ** participants randomly assigned to diff conditions

Generated by Koofers.com

factorial design	at least 2 IVs with each having 2 levels ** want to see different interactions/combinations
repeated measures design	same individuals participate in all/both conditions *** participants are repeatedly measured on the DV after being in each condition of the experiment *** take measures repeatedly thru experiment
ABA/ABAC design	A = baseline * B = intervention * C = another intervention *
internal validity	how well the test measures what its intended to measure ** refers to the ability to draw conclusions about causal relationships from your data *** design of study, measurement of DV

Generated by Koofers.com

External validity	extent to which the results can be generalized to the population youre studying
history threats	events outside the study; factors that happen over a period of time; occur between your 1st and 2nd measurements ** problem with longitudinal studies ** solutions: *** large sample sizes will wash out some influence *** random assignment to avoid unequal influences on 1 group over the other
maturation threats	changes in subjects over time; get older, wiser etc. *** ex. as they get older, may decrease smoking because they are more concerned about their health not bc of your intervention *** solutions: ** random assignment
repeated measurement threats	practice effects ** participants may develop practice effect; may figure out what youre studying ** solution: ** change format

Generated by Koofers.com

attrition threats	dropouts from the study; final sample not representative ** problem with longitudinal studies ** solutions: ** statistically control for differences in sample at the beginning and end
testing effects threats	differences may arise from the way measurements are taking place ** solutions: ** measurements must be standardized! ** same person measuring ** same tool used
hawthorne effect	impact of someone being observed; mat affect how participant acts *** ex. productivity at work ** solution: ** keep researcher hidden from sight
placebo effect	just being a part of a study may change people's bx *** solution: *** double or single blind

Generated by Koofers.com

halo effect	giving experimental group incentive because you want to see a specific outcome
measurement threats to internal validity	accuracy of instrument ** researcher training and skill ** floor/ceiling effects ** improper measurement of variables - nominal, ordinal, interval, ratio ** construct validity of instrument **
ceiling effect	only measures up to a certain pt; not capturing whole range of possibilites ** IV appears to have no effect on DV measure only because participants quickly reach the max performance level
floor effect	occurs when a task is so difficult that hardly anyone can perform well

Generated by Koofers.com

hierarchy of the levels of measurement	nominal, ordinal, interval, ratio
nominal	arbitrary *** categories with NO NUMERICAL SCALES *** ex. male/female; introvert/extrovert *** impossible to define any quantitative values and/or differences between/across categories
ordinal	rank order; numerical values limited *** ex. 2, 3, and 4 star restaurants *** ex. A B C D F grades *** intervals between items unknown
interval	numerical properties are literal *** ex. intelligence, temperature, apt test score *** NO TRUE ZERO

Generated by Koofers.com

ratio	TRUE ZERO *** weight, age *** zero indicates absence of value measured *** can form ratios (someone weights twice as much as someone else)
construct validity of the instrument	refers to the adequacy of the operational definition of variables; does the operational definition of a variable actually reflect the true theoretical meaning of the variable?
criterion related validity	used to demonstrate the accuracy of a measure or procedure by comparing it with another measure or procedure which has been demonstrated to be valid
predictive validity	instrument does a good job at predicting actual bx; research that uses a measure to predict some future bx

Generated by Koofers.com

concurrent validity	instrument gives outcome similar to an existing measurement
convergent validity	Convergent validity, is the degree to which an operation is similar to (converges on) other operations that it theoretically should also be similar to. *** extent to which scores on the measure in question are related to scores on other measures of the same construct or similar constructs should converge
discriminant validity	Discriminant validity describes the degree to which the operationalization is not similar to (diverges from) other operationalizations that it theoretically should not be similar to. *** measure should discriminate between construct being measured and other unrelated constructs
threats to external validity	sample characteristics - must be similar to target pop characteristics! *** subject/tx interaction - only certain participants do well - i.e. youngest segment *** artificial research setting - may not generalize well to actual environment

Generated by Koofers.com

© Copyright 2012 , Koofers, Inc. All rights reserved.

The information provided on this site is protected by U.S. and International copyright law, and other applicable intellectual property laws, including laws covering data access and data compilations. This information is provided exclusively for the personal and academic use of students, instructors and other university personnel. Use of this information for any commercial purpose, or by any commercial entity, is expressly prohibited. This information may not, under any circumstances, be copied, modified, reused, or incorporated into any derivative works or compilations, without the prior written approval of Koofers, Inc.